Sha if = i ; OY Na, Seertattees ay BEE HENGE Ea : ) Say i i nt tt Ate, Mhkle— C4 Oo 2ebe too TOEO O WVU A Sul! ~~ WI 6 | ANALYSIS OF. DEVELOPMENT EDITEDVBY BENJAMIN H. WILLIER Professor of Zoology, Johns Hopkins University, Baltimore, Maryland PAUL A. WEISS Member, Rockefeller Institute for Medical Research, New York, New York; formerly Professor of Zoology, University of Chicago, Chicago, Illinois VIKTOR HAMBURGER Professor of Zoology, Washington University, St. Louis, Missouri ILLUSTRATED W. B. SAUNDERS COMPANY Philadelphia & London - 1955 Copyright, 1955, by W. B. Saunders Company Copyright under the International Copyright Union All Rights Reserved. This book ts protected by copyright. No part of it may be duplicated or reproduced in any manner without written permission from the publisher. Made in United States of America Press of W. B. Saunders Company, Philadelphia Library of Congress Catalog Card Number: 55-5565 CONTRIBUTORS L. G. BARTH, Ph.D. Professor of Zoology, Columbia Univer- sity, New York, New York. N. J. BERRILL, Ph.D., D.Sc., F.R.S., F.R.S.C. Strathcona Professor of Zoology, McGill University, Montreal, Quebec. DIETRICH BODENSTEIN, Ph.D. Insect Physiologist, Medical Laborato- ries, Army Medical Center, Maryland. EDGAR J. BOELL, Ph.D., D.Sc. Ross Granville Harrison Professor of Ex- perimental Zoology, Yale University, New Haven, Connecticut. ROBERT KYLE BURNS, Ph.D., D.Sc. Staff Member, Department of Embryol- ogy, Carnegie Institution of Washington; Honorary Professor of Zoology, The Johns Hopkins University, Baltimore, Maryland. W. M. COPENHAVER, Ph.D. Professor of Anatomy, College of Physi- cians and Surgeons, Columbia Univer- sity, New York, New York. DONALD P. COSTELLO, Ph.D. Kenan Professor of Zoology, University of North Carolina, Chapel Hill. WILLIAM ETKIN, Ph.D. Associate Professor of Biology, The City College, New York, New York. GERHARD FANKHAUSER, Ph.D. Professor of Biology, Princeton Univer- sity, Princeton, New Jersey. VIKTOR HAMBURGER, Ph.D. Professor of Zoology, Washington Uni- versity, St. Louis, Missouri. JOHANNES HOLTFRETER, Ph.D. Professor of Zoology, The University of Rochester, Rochester, New York. FLORENCE MOOG, Ph.D. Associate Professor of Zoology, Washing- ton University, St. Louis, Missouri. J. S. NICHOLAS, Ph.D. Sterling Professor of Biology, Yale Uni- versity, New Haven, Connecticut. JANE M. OPPENHEIMER, Ph.D. Professor of Biology, Bryn Mawr College, Bryn Mawr, Pennsylvania. MARY E. RAWLES, Ph.D. Research Associate (Embryology), De- partment of Biology, The Johns Hopkins University, Baltimore, Maryland. HANS RIS, Ph.D. Professor of Zoology, University of Wis- consin, Madison. DOROTHEA RUDNICK, Ph.D. Professor of Biology, Albertus Magnus College, New Haven, Connecticut. 1V FRANCIS 0. SCHMITT, Ph.D., D.Sc. Professor of Biology, Massachusetts Insti- tute of Technology, Cambridge. ISAAC SCHOUR, D.D.S., Ph.D. Professor of Histology, University of Ili- nois, College of Dentistry, Chicago. H. B. STEINBACH, Ph.D. Professor of Zoology, University of Min- nesota, Minneapolis. CURT STERN, Ph.D. Professor of Zoology, University of Cal- ifornia, Berkeley. VICTOR CG. PWILLY, PhD, Professor of Biology, Stanford University, Stanford, California. ALBERT TYLER, Ph.D. Professor of Embryology, California In- stitute of Technology, Pasadena. CONTRIBUTORS RAY L. WATTERSON, Ph.D. Associate Professor of Biology, North- western University, Evanston, Illinois. PAUL A. WEISS, Ph.D., M.D. (hon.) Member, Rockefeller Institute for Med- ical Research, New York, New York; formerly Professor of Zoology, Univer- sity of Chicago. BENJAMIN H. WILLIER, Ph.D., Sc.D. Henry Walters Professor of Zoology, The Johns Hopkins University, Baltimore, Maryland. C. L. YNTEMA, Ph.D. Professor of Anatomy, State University of New York Medical Center at Syracuse. EDGAR ZWILLING, Ph.D. Associate Professor of Animal Genetics, Storrs Agricultural Experiment Station, University of Connecticut, Storrs. PREFACE THIS BOOK owes its inception to informal gatherings, seminar fashion, of a small group of embryologists who for several summers (1933-1940) periodically retired from the busy scene of the Marine Biological Labora- tory at Woods Hole to the peace and quiet of the sand dunes along the northern coast of Cape Cod near Barnstable. With the sea as background and the sand for a blackboard the “Sandpipers” (a name derived from our alert and ever-searching avian companions on the beach) discussed at length the problems of development and groped for a better under- standing of the mechanisms of embryogenesis. To those who took part in them, these group discussions were a valuable experience. The satisfaction that came from the exchange and conciliation of conflicting views aroused our urgent desire to broaden the experience and share it with a far wider circle of biologists. Jointly the hope was engendered that future accounts of embryological knowledge would emphasize the dynamic and causal aspects of embryogenesis rather than mere description and seriation of developmental stages, a prac- tice still too common in the lecture room and textbook. To transcend descriptive embryol- ogy and blend experimental data with “Beob- achtung und Reflexion” was clearly set as our goal. Only by such an account could younger students be challenged and _ influ- enced in their future research and teaching in this important field. Above all, the need was felt for helping to overcome the trends of over- specialization by encouraging a wider, inter- disciplinary perspective and by integrating the ever-growing volume of accumulated in- formation into a broad conceptual framework. The need for a well-balanced account of the developmental process was apparent. But how was such a plan to be translated into action? It was evident from the start that the sub- ject matter had grown in volume and intric- acy to the point where it seemed futile for any one individual to attempt to cope with such a task. The alternative was to call on many specialists for authoritative presentations of their respective subjects. We realized that by this procedure much of the desired unity and integration would be sacrificed, and the pres- ent volume bears plainly the stigmata of these imperfections. Yet, despite our hesita- tions on this score, the three of us, encouraged by the urging of many colleagues, outlined in 1947 a plan for a collaborative work on the analysis of the developmental process. The original blueprint contained an out- line and table of contents of the subject mat- ter to be covered in hierarchical divisions, as well as specifications for their serial order and relative proportions. For this basic pattern the three Editors take full responsibility. Yet, within that general frame, the individual con- tributors were given no more than a general topical guide that left full scope to their per- sonal preferences in the choice of samples, style, and manner of presentation, the only provision being that they conform to the gen- eral spirit and objectives of the undertaking. The guiding aims were expressed to them in the following commentary. The purpose of this book is to present a modern synthesis of our knowledge of the principles and mechanisms of development. In these days of rap- idly expanding information, it becomes increasingly difficult to keep perspective. It is urgent, therefore, that this book provide not just another source of information, but that it view the phenomena of development from a common perspective so that the reader may recognize the great main lines and inner coherence of the field above the multiplicity of often unrelated details of which the field seems composed when viewed too closely. There is perhaps need for a comprehensive compilation of all the experimental data that have been amassed in the field of Experimental Embryology in the past. However, this book is not intended to fill that need. It is not to be a handbook. It does not aim at a com- plete and exhaustive review of the field. Each con- tributor is asked to make a critical and, in a way, subjective selection of the special field to be covered in his article. He should give a clear outline of the general problems, concepts, and lines of imvestiga- tion of his topic and illustrate them with selected examples from experimental data. Only those ex- periments should be presented that are crucial and analytically strong and convincing. Repetitiveness V1 should be avoided. Use should be made of tabulations and graphs wherever possible. Since the book ad- dresses itself mainly to active or potential investi- gators (particularly in the experimental branches of embryology, pathology, histology, endocrinology, and developmental genetics), it would be of value to point out gaps in our knowledge, the lack of critical experimental data in unexplored or con- troversial fields, and lines of research which would deserve being followed up. In summary, the book has as its major objective the synthesis and evalua- tion of pertinent material selected from the whole field of animal growth and development, with emphasis upon recognized principles and mechan- isms as well as on unsolved and new problems. With these suggestions we approached twenty-five biologists prominent in the sub- ject areas to be covered in the volume. They readily accepted the invitation to collaborate despite the tribulations and obligations in- herent in such undertakings. The Editors are very grateful to all of them not only for their contributions to this book but also for the spirit of cooperation and patience which they exhibited during the years of arduous labor that went into its preparation. As in all con- certed creative efforts of this kind, progress in realization was slow and at times faltering. Contrary to the development of an organism, no forces were at work to coordinate the sep- arate creative efforts, and the Editors did not see fit to weld the different contributions into a uniform mold. Each contributor is finally responsible for the organization, scope, and content of his text. The Editors, on the other PREFACE hand, must bear the responsibility for the plan and the scope of the book, and assume the blame for any defects in its structure. Whatever its imperfections and limitations, the book represents a first-hand portrayal of present-day views of animal development. As such, we hope it may provide a basis of de- parture for future endeavors of this kind. The science of embryology, like the embryo, is governed by the principles of progressive dif- ferentiation, its present status only a transi- tory moment between past and future—its full potentialities yet to be realized. It is to the pioneering spirit of those students who here- after will enter the field of development and growth that this volume is primarily dedi- cated. In no lesser degree we inscribe these pages to students and investigators in other fields of the biological sciences, including medicine and agriculture, who are constantly confronted with problems of a developmental nature and must deal with them. The Editors have been fortunate indeed in the cordial relationship which has existed between them and the publishers from the beginning of this undertaking. We are most grateful to them for their unlimited patience, resourcefulness, and splendid cooperation in making a book such as this all that it should be in style and typography. B. H. WILLIER Paut WEIsS VikToR HAMBURGER CONTENTS SECTION I PROBEENMS: CONCEPTS AND HEIR HISTORY =) ete ee 1 BY JANE M. OPPENHEIMER The Early Embryology of the Greeks-eAristotlem nee 1 Embryology and the Renaissance: Habriciuss Hanveys -. ee eee: 3 Embryology and the New Micro- scope: Preformation and Mal- pighi; Epigenesis and Wolff .... 6 SECTION II METHODS AND TECHNIQUES BY JANE M. OPPENHEIMER Introduction: Some General Con- SIderatlOnsi ee eee 95 Observation vs. Interference as an Approach to Embryological Prob- VemiS.Secun ones hee ee eee 26 SECTION III Embryology and Naturphilosophie: Goethe and Von Baer ......... 8 Embryology and Evolution: Dar- Wwildean dlllaeccie lua see 13 Techniques of Interference with the Embryowge ono nce ete: 31 CEELUEARE Ss TRUGRURE- AND ACTIV TGS eits eee Meee ee 9 Chapter 1. CrLL ConstITuTION BY FRANCIS 0. SCHMITT Techniques of Analytical Cytology 39 The Problem of Fixation ........ 44 The Colloidal Organization of Pro- COD LASTINRN efit oes eR 41 Water and Dissolved Substances .. 42 Particulate Systems ............. 42 Lipochondria (Golgi System) .... 45 Fibrous Systems of Cytoplasm .... 46 518) The Paracrystalline (Mesomor- phic) State; Tactoids, Coacer- vates and Long-Range Forces .. 53 Sol-Gel Transformations, Contrac- tility and the Cell Cortex ...... 54: The Cell Membrane ............ 56 Some Physical Chemical Considera- tions of Morphogenetic Processes 59 Viil CONTENTS Chapter 2 Grr wrsn METABOLISM fae) 4): eee en eee wero. BY H. B. STEINBACH AND F. MOOG Glycolytic and Oxidative Mecha- The Regulation of Metabolism in TIA GTIAG IAA Ee ee: SE eee 71 Developmenteess =] one eee 83 The Control of Cellular Metabolism 79 Chapter Soe Cru. Wvision han 5 eer ea cE ee ee BY HANS RIS imtroductione eee eeter 91 Analysis Of IMMEOSIS . 2. 2<1<...-e 93 Description of Mitosis in the White- dichwblastular <<... ca . 4-4-5475: 161 Qualitative Variability of Genic Genetic Asymmetries .......... 161 Content eet Oe hoists 153 Difterentiatlonpee eee eee 163 Winneror Genre pAction.= =e eee 155 SECTION V EMBRYOGENESIS: PREPARATORY PHASES 5 2 3 = = « = 5 oem Chapter 1. GAMETOGENESIS, FERTILIZATION AND PARTHENOGENESIS . . . 170 BY ALBERT TYLER Ganietozenesisua ee eee eee 170 Artificial Parthenogenesis ...... 201 Rertilazatronge sein sete. 181 Chapier.2. CrLravaAce, BLASTULATION, AND GASTRUEATION |=) neue e220 BY DONALD P. COSTELLO Determinate and Indeterminate Cleavage without Nuclei ....... 216 Gleavages carer acetates a5 MAG Redistribution of Egg Substances CONTENTS and Their Relation to Cleavage Patterns Function of Partition Membranes in Cellular Differentiation .... SECTION VI am Factors Concerned in Shaping of the lastila . ...... 3s eee Gana ohstecdocoecocecce EMBRYOGENESIS: PROGRESSIVE DIFFERENTIATION Chapter 1. AMPHIBIANS . Chapter 4. BY J. HOLTFRETER AND V. HAMBURGER Egg Organization and Determina- TLOMEOE CAKES) eho tver, SESSA Gastrulation, Fate Maps ........ Significance of Gastrulation Move- ments and Induction as Organ- WAV? IEVONWS 350 agcocccccue$ State of Organization of the Early Gastrullay can eee Analysis of the “Organizer” .... The Organization of the Meso- derm Mantle and the Tail Bud im the INeurula: . - =. 222: =:- 470 Chapter 7. TEETH BY ISAAC SCHOUR nit oduction eee eee 492 Growth ks. 6 eye CONTENTS 346 346 4.02 415 429 440 462 492 CONTENTS Calcifications nee eee 497 iri EOMs ky, 5 oye ose os eae ea erin « A497 ATER IU OTM er Easier aeons ee 497 Chapter 8. SKIN AND ITs DERIVATIVES BY MARY E. RAWLES Imtroductionmen eee see 499 Omnpin' of the Skin. = ...-545-.2 499 Regional Specialization of the (ST te ote se geet gieetne 504 SECTION VIII ENERGY EXCHANGE AND ENZYME DEVELOPMENT DURING EMBRYOGENESIS BY E. J. BOELL Introduction: Energy Require- ments of the Embryo ........ 520 Energy Release during Periods of Reduced Oxygen Supply or An- ACTODIOSISHEA Ne ee ker 522 SECTION IX ONTOGENY OF IMMUNOLOGICAL PROPERTIES BY ALBERT TYLER Antigens in Development ....... 556 Development of Natural Anti- bodies and of Complement .... 564 SECTION X Integumentary Patterns ........ 512 Conclusion sess eee ee 516 IRGH OURO ooooocccecoouroonce 523 Energy Sources during Develop- TRIGTALL SY Cree era 533 Respiratory Mechanisms ....... 538 Gonclhusiont ose fo 548 Development of Antibody-Form- ine. Gapaci tye eee eee 564 The Concept of Natural Auto- Anitibodies: *y52rs's.eeecens ee 566 ONTOGENY OF ENDOCRINE CORRELATION BY B. H. WILLIER Anterior Pituitary—Thyroid Rela- EONS MWe nee BORA Wk A ee 574 Anterior Pituitary—Adrenal Rela- TLOTIS Meco oan cho Re Re era 585 SECTION XI THE DETERMINATION OF SIZE BY N. J. BERRILL Growth hater soe eee 623 Anterior Pituitary-Gonad Rela- TONS) s. «ta cas ieee 593 Anterior Pituitary and Growth .. 600 Islets of Langerhans Parathyroid Glands ........-..-: 611 Relative) Growth) 94.05.0004. 0 - 625 x1 499 520 556 574 620 Xii SECTION XII METAMORPHOSIS BY WILLIAM ETKIN Endocrine Control of Amphibian IMetamorphosisy =.) 4.4. 631 Controlling Factors in Insect Met- amOrphosisuce ob 1. Sees ee 637 Mode of Action of the Metamor- PhicwStimlwyS o. 22. ee eee 644: SECTION XIII REGENERATION Chapter 1. BY L. G. BARTH INVERTEBRATES The Polarity Problem in Regener- ACIOT ae re Fae ee 664 Mavs, Sirah. sogseboeaace oaae 666 iPotencies of Gellsisen ieee eee nae 667 Chapter 2. VERTEBRATES . BY J. S. NICHOLAS Introduction c..5 eee eS reyes 674 Amphioxus and Piscine Forms .. 675 (Rene NIB Me ink et 677 Endocrine Effects: & 29. ...7250%. 685 Reptilian Regeneration ........ 686 SECTION XIV TERATOGENESIS BY EDGAR ZWILLING Imtroductionwac. escent ie 699 The-Gausal Agents: .¢....-...- 700 How Agents Produce Their Ef- isle eh eee Rie eRe C ERS 703 INDEX CONTENTS 631 The Activation of the Metamor- phic Wechanisimh 45 55..400. : 653 Cessation of Metamorphic Trans- NOWIEIHO) geapeoccosocoonebe 656 Summniar yuk Ash ane + 657 664: 664 Correlative Differentiation ..... 668 Environmental Factors ........ 671 Metabolism and Regeneration ... 672 674 Physiological Regeneration ..... 688 General Eustolocyanre eee 690 ithe Hyerandléens aes ee 690 Conclusions. seer oe eee eee 693 699 A Classification of Terata Based One bimbryOlogy eee eee 706 Generalizationst ace eee eae 710 RGsume sc ,.4r aoe | ee 715 (OHS Section I PROBEEMS, "CONCEPTS ANDSPHETR HISTORY JANE M. OPPENHEIMER * “Is cell-differentiation inherent or induced? “A thoughtful and distinguished naturalist tells us that while the differentiation of the cells which arise from the egg is sometimes inherent in the egg, and sometimes induced by the conditions of develop- ment, it is more commonly mixed; but may it not be the mind of the embryologist, and not the natu- ral world, that is mixed? Science does not deal in compromises, but in discoveries. When we say the development of the egg is inherent, must we not also say what are the relations with reference to which it is inherent? When we say it is imduced, must we not also say what are the relations with reference to which it is induced? Is there any way to find this out except scientific discovery?” W. K. Brooks (’02, pp. 490-491) Ir is the self-imposed task of the present compendium to review and evaluate the past and present accomplishments of the science of embryology in order more intelligently to facilitate progress into its future. The sepa- rate contributions which make up the main body of the volume must necessarily concen- trate on particular fields of investigation. It is the purpose, therefore, of the first two chapters to provide a general background against which these more special subjects may be considered. Out of convenience, rather than from logical necessity, these two chapters will concern themselves first with concepts, and secondly with techniques, though the nature of the scientific method is such that these two aspects of the prob- lem are inextricably interrelated. Arbitrar- ily, too, the topics chosen for discussion will be selective rather than exhaustive; since it is not possible in a few pages to do justice to even a few of the great contributors of the past, only those have been chosen whose writings are most relevant to the sequel, and * The writing of Sections I and II was carried out both at the Osborn Zoological Laboratory, Yale University, and at Bryn Mawr College. I owe espe- cial gratitude to the Library of the College of Phy- sicians of Philadelphia for the use of their collec- tions and for generous assistance. even of these, many can enjoy only the bar- est mention. THE EARLY EMBRYOLOGY OF THE GREEKS: ARISTOTLE Since it was the Greeks who performed the great tour de force of freeing science from magic and elevating it into the realms of pure reason, it is sensible to begin by exam- ining a few of their contributions to embry- ology. They were early to develop an inter- est in beginnings; their very word for nature (dios, physis) according to some, including Aristotle (Parts of Animals, 1945 edition, pp. 74-75), implies growth, genesis or origin (ptecbar.), and Anaximander, who flour- ished in the sixth century B.c., spoke of the yov.ov, the germ or fetus of the world. They recognized early that change was an essence of existence, as we know from Hera- kleitos’ emphasis on flux, and as is evident from their mythological conception of cos- mos evolving from chaos. And from the be- ginning they compared cosmos to the organ- ism, witness Plato (Timaeus, [1944] edi- TLOTAS Oye): Its composing artificer constituted it from all fire, water, air, and earth; leaving no part of any one of these, nor any power external to the world. For by a reasoning process he concluded that it would thus be a whole animal, in the highest de- gree perfect from perfect parts. But more than this, perhaps even because of it, they were able even as early as the time of Anaximander to conceive of the or- ganism as emergent, and indeed of animals as related to man: a fragment concerning the teachings of Anaximander reads that “living creatures arose from the moist ele- ment, as it was evaporated by the sun. Man was like another animal, namely, a fish, in the beginning” (Burnet, 730, p. 70). No attempt can be made here to enumer- ate the many Greek philosophers to build upon these beginnings, or to evaluate the 2 PROBLEMS, CONCEPTS AND THEIR HISTORY contributions of those who did. It will have to be sufficient here to name a few, and the interested reader is referred to Balss (736) for additional details. Suffice it here to com- ment that theirs was the task of the first early and perhaps random collection of data, which must precede even the primitive classification which many consider to repre- sent the first stage of scientific inquiry. Of some, we know only from the meager extant fragments, that they recorded what they thought to be observed fact; for in- stance, from Parmenides a fragment remains implying that males are generated on the right and females on the left. In the case of others, even before Aristotle, it is clear that they believed that around the observed facts they could elaborate theory. Empedokles, for example, believed the fetus to arise partly from male and partly from female semen, the children resembling most the parent who contributed most to the offspring; he spoke of the influence of pictures, statues and so forth in modifying the appearance of the off- spring, of twins and triplets as due to “super- abundance and division of the semen” (Bur- net, ’30, p. 244) ; he knew there was a regular sequence of events in development and spoke of the heart as formed first in development, the nails last, sowing seeds of concepts, which, right or wrong, were destined often to recrudesce in subsequent ages. A Hippocratic treatise on generation went further in developing theories, formulating an early expression of the doctrine of pan- genesis, and, relating to it, what seems to be on post hoc reasoning a doctrine of the in- heritance of acquired characters. This trea- tise, before Aristotle, recognized the impor- tance of methodology, and advocated sys- tematic daily observation of chicken eggs: “Take twenty or more eggs and let them be incubated by two or more hens. Then each day from the second to that of hatching re- move an egg, break it, and examine it. You will find,’ continues the writer, illustrating an apparent dependence of concept on method and inferring the great generaliza- tion, “exactly as I say, for the nature of the bird can be likened to that of man” (Singer, opie). Aristotle’s own accomplishment was none the less impressive, for all he drew on his predecessors and contemporaries. “There was a wealth of natural history before his time; but it belonged to the farmer, the huntsman, and the fisherman—with something over (doubtless) for the schoolboy, the idler and the poet. But Aristotle made it a science, and won a place for it in Philosophy” (Thompson, 40, p. 47). And in establishing it as scientific, he set its standards higher than hitherto by far. He followed, in embryology, the method of the Hippocratic writer On Generation, to perform and record most of the available ob- servations, many in error but also many cor- rect, thus to constitute a collection of knowl- edge on the development of the chick which became the foundation on which all embry- ology was to build; and it has been said, with much justice, of his account that “al- most two thousand years were to roll by before it was to be equaled or surpassed” (Adelmann, ed., in Fabricius, 1942 edition, p. 38). He concerned himself not only with the development of the chick but also with the generation of many other forms, and elaborated a kind of classification (though not in the modern sense; cf. Thompson, ’40) of animal forms according to their mode of reproduction. By so doing, he both estab- lished embryology as an independent sci- ence, and he fitted embryological knowl- edge into a pattern larger than its own, with great clarity of vision and imagination. On the theoretical side, he followed his predecessors by adopting a modified view of pangenesis, and concurred with them in sup- porting the doctrine of the inheritance of acquired characters. He broke away from his predecessors, however, in developing a new and erroneous yet highly influential concept of the relative roles of male and fe- male in development, postulating the former as providing the form, at once formal, effi- cient and final cause, and the latter the sub- stance, the material cause, for the new organism. By thus undervaluing the egg, he paid embryology the obvious immediate disserv- ice; but in formulating his conception of biological form as inseparable from matter he laid the way open for ultimate progress in biological science. The argument is meta- physical to the taste of the modern scientist; but Aristotle will be found not to be the last embryologist to be so tainted. We concur with his intent, after all, every time we speak of “animal forms” as a euphemism for “animal species.” And Aristotle, with the natural historian’s innate feeling for natural form, by envisioning form as a part of actuality rather than something above it, brought bio- logical material to be directly investigable by the sense organs. His theories concerning special develop- mental phenomena, related to his primary PROBLEMS, CONCEPTS AND THEIR HISTORY 3 philosophy as they were, are deep in much of the embryological and indeed the wider biological thinking both of the past and the present. His description of the heart as the first organ of the embryo to be formed, both in time and in primacy, tied as it was to the conception of the soul as formal and final cause and of vital heat in the blood as the agent of the soul, dominated the notions not only of the developing but also of the adult circulation, and hence all physiology, through to the nineteenth century and the downfall of the phlogiston theory. His con- cept of organ as related to final cause epito- mizes teleology, and with all the weight of Galen’s authority in support still permeates much of the thought of modern biology. Matter with form inseparable from it as op- posed to the more material matter postulated by Leucippus’ and Democritus’ atomic the- ory, which implied preformation, in a way made possible the whole doctrine of epigene- sis, first clearly formulated by Aristotle and still central in all embryological thinking today. Form as inseparable from matter makes possible a conception of pattern emergent, an analogy of development and the process of plaiting a net or the process of painting a picture; for Plato, the Ideal mesh would have been already woven, the Ideal portrait previously complete. Aristotle (Generation of Animals, 1943 edition, pp. 147, 149, 225) could frame the modern ques- tion: How, then, are the other parts formed? Either they are all formed simultaneously—heart, lung, liver, eye, and the rest of them—or successively, as we read in the poems ascribed to Orpheus, where he says that the process by which an animal is formed resembles the plaiting of a net. As for simultaneous formation of the parts, our senses tell us plainly that this does not happen: some of the parts are clearly to be seen present in the embryo while others are not. . . . Since one part, then, comes earlier and another later, is it the case that A fashions B and that it is there on account of B which is next to it, or is it rather the case that B is formed after A?... In the early stages the parts are all traced out in outline; later on they get their various colours and softnesses and hardnesses, for all the world as if a painter were at work on them, the painter being Nature. Painters, as we know, first of all sketch in the figure of the animal in outline, and after that go on to apply the colours. The metaphor will speak for itself to mod- ern experimental embryologists. Aristotle, however, for all his natural acuity, was strangely double-minded. In his dynamic feeling for form, derived from direct study of living biological material, he was modern, and was to lead eventually straight to the inductive biology of modern times. But his conceptions of the wider Universe, based on pure reason, because statically and_ struc- turally interpreted and thus transmitted by medieval commentators, deluded posterity, and it was unfortunately the static Aristotle, the Aristotle of a sterile cosmogony, crystal clear but crystal rigid, who dominated the thought of the Middle Ages. So far as even the embryology was concerned, the Middle Ages transmitted his concepts, and occasion- ally amplified them, as in the case of Al- bertus Magnus, but devitalized them and thereby hardly improved them. Appreciation of their dynamic qualities awaited the Ren- aissance and later ages. EMBRYOLOGY AND THE RENAISSANCE: FABRICIUS, HARVEY When the Renaissance came under way it accelerated its course into the new thought by taking strength from the Greek past through all the resources of Humanism; and a “reconstruction of the Greek spirit” (cf. Singer, [’41], p. 166) was an essential part of the rebirth. Even Galileo has been called a “typical Paduan Aristotelian” in method and philosophy at least, if not in physics (Randall, cited by Adelmann, ed., in Fabri- cius, 1942 edition, p. 55), and Whitehead (25, p. 17) reminds us that Galileo “owes more to Aristotle than appears on the sur- face of his Dialogues: he owes to him his clear head and his analytic mind.” Vesalius’ interpretations of his observations were as teleological as those of Galen after which they were modelled (cf. Singer, ’44, p. 81, who called him “a disciple of Galen by training, by inclination, and by his whole cast of thought”); his method, however, was also in part that of Aristotle. Copernicus, who was accused by Kepler of interpreting Ptolemy, not nature, at least challenged the Aristotelian cosmogony; Vesalius imitated the method of the Aristotle who is so rarely remembered as having written about an embryological problem (Generation of Ani- mals, 1943 edition, pp. 345, 347): This, then, appears to be the state of affairs . . . so far as theory can take us, supplemented by what are thought to be the facts about their behaviour. But the facts have not been sufficiently ascertained; and if at any future time they are ascertained, then credence must be given to the direct evidence of the senses more than to theories. The scientist, who customarily characterizes 4 PROBLEMS, CONCEPTS AND THEIR HISTORY the Renaissance as a movement for freedom with respect to authority, often neglects to remember that it was in part from “author- ity” that the inspiration to achieve freedom derived. It was Fabricius, student of Fallopius, himself a student of Vesalius, who first ex- haustively applied the rigorous “new” Ve- salian method of direct observation to the study of embryos, though he had many predecessors who had made isolated obser- vations on embryonic material (among them Columbus, Fallopius, Eustachius, Arantius, Aldrovandus, Coiter et al. Cf. Needham, ’34, and Adelmann, ed., in Fabricius, 1942 edi- tion, for full discussion; see also Adelmann for full critical treatment of Fabricius him- self). On the observational side, he was the first to publish illustrations based on systematic study of the development of the chick, and this, though he neglected to describe them in detail, was probably his most significant contribution. He made the way easier for the later preformationists by drawing the supposed three and four day chicks much too advanced for their normal chronological age; among his other fallacies, the most notable was his ascription to the chalazae of the role of forming the embryo. Among his improve- ments to the existing embryological knowl- edge was his emphasis that the carina (whose metaphysics he discussed more com- pletely than its embryological fate) is formed before the heart, controverting Aris- totle, and before the liver, taking issue with Galen in both fact and philosophy. He stud- ied the fetal anatomy of various vertebrates, that of many mammals, including man, and presented illustrations of the comparative anatomy of the placenta, showing his spe- cial interest in the umbilical and the fetal circulation, though he devoted himself to Galenic principles in his interpretations of these. Even Fabricius, then, as late as the sixteenth century was exemplifying the con- flict of the Renaissance between allegiance to authority and confidence in direct per- sonal observations. But though in one sense his position represents an inevitable retreat, even behind the position of Aristotle, in that he emphasized the anatomy of embryos rather than the process of development, yet his work looked forward to the new embry- ology in the influence it exerted on William Harvey. Fabricius’ name, as Adelmann points out (op. cit., p. 115) begins the first sentence of Harvey’s text on generation; and Harvey, too, like his preceptor, looked back to Aris- totle in his interpretations, for all that his demonstration of the circulation in method, fact, and conception, was to lead to the whole experimental and analytical biology of the future. Harvey followed Bacon’s prin- ciple of explaining nature by observation and experiment, and Galileo’s of measuring what is measurable and making measurable what is not. Harvey’s contemporaries be- lieved, with Fracastorius, that “the motion of the heart was to be understood by God alone” (Harvey, De motu, 1931 edition, p. 25). Harvey proved it to be a mechanical function. Yet he could speak of the motion of the blood, after Copernicus, Kepler and Galileo, as “circular in the way that Aris- totle says air and rain follow the circular motion of the stars” (ibid., p. 70) and, like a good Aristotelian, he left the vital spirits remaining in the blood. “Whether or not the heart,” he wrote, “besides transferring, dis- tributing and giving motion to the blood, adds anything else to it, as heat, spirits, or perfection, may be discussed later and de- termined on other grounds” (ibid., p. 49). Harvey may have surmised how to treat the organ as a machine, but he was in some ways too Aristotelian to appreciate the im- plications of his own advanced experiment. He was not so bound by authority, how- ever, as to be unable to free himself from some of the old embryological errors. He refuted on an observational basis, for in- stance, the notion that right and left repre- sent maleness and femaleness, and he cor- rected the idea of Fabricius concerning the role of the chalazae by demonstrating the cicatricula (our blastoderm) as the source of the embryo; he corrected, too, various specific observational errors of Aristotle. Most important, he abolished for all time the Aristotelian conception of female as sub- stance and male as form. Galen to be sure had seemed to localize both material and efficient causes in both male and female semen, as had Fabricius after him in a con- fused sort of way; but it was Harvey, for all his fanciful speculation concerning the sig- nificance of fertilization, who finally ele- vated the egg to its full and ultimate dig- nity. The processes of development can obviously hardly be investigated before the object that is developing is at least defined as their residence, and Harvey’s contribu- tion here was therefore a significant one. It is abundantly clear, however, that by egg Harvey meant something different than we do. He knew there was necessary for de- PROBLEMS, CONCEPTS AND THEIR HISTORY 5 velopment a double contribution, deriving from both male and female parent: The egge is a certain Conception proceeding from Male and Female, qualified with the power of both: and out of it being One, one Animal is con- stituted. ... An egge can no more be made without the assistance of the Cock and Henne, then the fruit can be made without the Trees aid. . . . For without a Cock it cannot be fruitfull, without a Henne it cannot be at all (1653, pp. 136, 157, 155). Yet even in the case of the chick this is not the egg to Harvey that is the visible entity of the laboratory or kitchen: And though it be a known thing, subscribed by all, that the foetus assumes its original and birth from the Male and Female, and consequently that the Egge is produced by the Cock and Henne, and the Chicken out of the Egge: yet neither the Schools of Physitians, nor Aristotles discerning Brain, have disclosed the manner, how the Cock and its seed, doth mint and coine the Chicken out of the Egge. ... But that neither the Hen doth emit any Seed in coition, nor poure forth any blood at that time into the cavity of the Uterus; as also that the egge is not formed after Aristotles way; nor yet (as Physitians suppose) by the commixture of Seeds, and likewise that the Cocks seed doth not penetrate into the hol- low of the womb, nor yet is attracted thither, is most manifest, from this one Observation, namely, That after coition there is nothing at all to be found in the Uterus, more then there was before (ibid., pp. 250, 199). He met the same failure in a vain exam- ination of the uterus of the mammal, and was driven therefore to resolve his dilemma by a poor analogy: The Egge is... a kind of an exposed Womb, and placed where the Foetus is formed: for it executes the office of the Matrix, and shelters the Chicken till its just time of Birth. . . . Oviparous creatures are therefore not distinguished from Viviparous, in this, that these bring forth their Foetus alive, but they do not; ... but their maine difference consists in the manner of Generation; namely, in that Viviparous creatures continue their Womb within them, in which the Foetus is fashioned, cherished, and com- pleated: but Oviparous expose their Egge or Matrix without: yet nevertheless they do ripen and cherish it as much by Incubation, as if they did reserve it within their bowels (ibid., p. 127). Martin Llewellyn put it more succinctly, if less elegantly, in his poem “To the In- comparable Dr. Harvey, On his Books of the Motion of the Heart and Blood, and of the Generation of Animals,” when he wrote (ibid., n.p.): That both the Hen and Houswife are so matcht, That her Son Born, is only her Son Hatcht. Harvey began his embryology from an Aristotelian metaphysical preconception: “All perfect science depends upon the knowledge of all causes: and therefore to the plenary comprehension of Generation, we must ascend from the last and lowest efficient to the very first and most supreme, and know them all” (ibid., p. 259). Frus- trated by the inability of his own senses to find the physical reality he sought, he took solace in a metaphysical conception of his own, at a different level, and envisioned a metaphysical egg: The Egge . . . seemes to be a kinde of Medium; not onely as it is the Principium, and the Finis, but as it is the Common work or production of both Sexes, and compounded of both. . . . It is also a Medium, or thing between an Animate and an In- animate creature; being neither absolutely impow- ered with life, nor absolutely without it. It is a Mid- way or Passage between the Parents and the Chil- dren; between those that were, and those that are to come. .. . It is the Terminus a quo, the Point or Original from which all the Cocks and Hennes in the world do arise and spring: and it is also the Terminus ad quem, the Aim and End proposed by nature, to which they direct themselves all their life long. By which it comes to pass, that all Individ- uals, while to supply their Species they beget their Like, do continue and perpetuate their duration. The Egge is at were [sic] the Period of this Eternity (ibid., p. 137). But though Harvey necessarily ended as he began in metaphysics, he had shown the embryologists to follow him where to begin their physical investigations. His transmis- sion, therefore, of Aristotle’s notions of epi- genesis takes on a new meaning, since his epigenesis takes place in an egg which to embryologists succeeding him was the visible ego of reality, the egg which he searched for even though he failed to find it. His de- scription of epigenesis, in which All... parts are not constituted at once, but suc- cessively, & in Order. . . . Nature doth feed and enlarge all the Parts, out of the self same Nutri- ment, whereof the [sic] first did frame them... and like a potter, first she divides her materials, and she allots to the Trunk, the Head, and the Limbs, every one their share or cantlin: as Painters do, who first draw the Lineaments, and then lay on the Colours (ibid., pp. 225, 331), is a description significant for the modern embryologist in more ways than by the repe- tition of Aristotle’s metaphor. The fairly common delusion, however, that Harvey championed the cause of epi- genesis to the exclusion of others has little basis in fact. Harvey was more cautious than many more modern investigators in empha- sizing that “the principles of divers Animals being also diverse . . . the manner of the 6 PROBLEMS, CONCEPTS AND THEIR HISTORY generation of Animals is diverse likewise” (ibid., p. 384), and while he considered some animals to be “perfected by a succes- sion of parts’ (ibid., p. 344), he knew others to be “made intire at once” (loc. cit.), “formed and transfigured, out of matter al- ready concocted and grown” (ibid., p. 222). “The form ariseth ex potentiad materiae prae- existentis, out of the power or potentiality of the pre-existent matter; and the matter is rather the first cause of the Generation, then any external Efficient” (ibid., p. 223). This resounds strongly of preformation- ism, and indeed in the modern rather than the old-fashioned sense. There are those who claim that Harvey’s work on generation was of little historical moment because of its rela- tive obscurity at the time of its publication. Malpighi, however, knew it, and he knew because of it to start his studies with the blastoderm; indeed, Harvey is mentioned on the first page of Malpighi’s text, a notation which may bear witness to the fact that the ideas of preformation may themselves have been fostered at least in part by the inade- quacies of the early epigenetic postulate. EMBRYOLOGY AND THE NEW MICROSCOPE: PREFORMATION AND MALPIGHI; EPIGENESIS AND WOLFF Harvey’s failure which drove him back to the metaphysics from which he started we have called a physical one related to the inadequacy of his senses. Malpighi here had the advantage over him, with the use of the microscope as a new tool, and with it he overstepped the old limitations to enter what might seem in some ways a new conceptual realm, namely that of preformationism. This doctrine of preformation, however, was no clear and strong new reply to an old question by a new science. It was a principle deeply intrenched in ancient philosophy and destined to outlast for many years the valid- ity of the scientific evidence once seeming to favor it. It remained, indeed, a philosoph- ical dogma rather than a scientific principle even after long being discussed on a scien- tific basis: its most ardent biological cham- pion, Bonnet, was to betray the preponder- ance of its philosophical over its scientific weight by calling it ‘fone of the greatest triumphs of rational over sensual conviction” (cited by Needham, ’34, p. 191). The concept roots, on the philosophical side, at least as remotely into antiquity as the times of Leucippus and Democritus, whom Aristotle so strongly opposed, and the implications of preformation inherent in the ancient materialistic doctrines were clearly realized by Lucretius (De rerum natura, Bk. I, lines 159-214). Seneca wrote as early as the first century (cited by Needham, °34, p. 48): In the seed are enclosed all the parts of the body of the man that shall be formed. The infant that is borne in his mother’s wombe hath the rootes of the beard and hair that he shall weare one day. In this little masse likewise are all the lineaments of the bodie and all that which Posterity shall discover in him. When the formed element is present ab initio, the end and the beginning are the same, and ihe principle of emboitement be- comes difficult to escape. It too was recog- nized early; a theory of emboitement ex- pressed by Saint Augustine is quoted by Wheeler (1898). Nearer to the time of Mal- pighi (for the early and intervening develop- ment of the concepts see Cole, ’30; Meyer, 39; and Needham, ’34), Joseph of Aroma- tari (1625) reiterated an old idea of Empe- dokles that the plant is present in the un- germinated seed and said that “the chick is formed before the hen broods upon it” (Meyer, 739, p. 63). It was Malpighi, however, who in 1673 re- ported the observations which were to endow the theory with new vigor. He studied with the new microscope what he thought to be the unincubated egg, to see in its blastoderm the structures so magnificently portrayed in the familiar plates, and to interpret them to signify that the parts of the animal may pre-exist in the ege. He indulged in less dogmatism in his claims, however, than pos- terity customarily attributes to him, as is emphasized in analysis of his work by Adel- mann currently in progress. Malpighi or- ganized no formal and systematic theory of development; he did not himself use the word preformation, and there is some ques- tion, according to Adelmann, as to what he meant by the pre-existence of the animal in the egg. He expressed his notions only ten- tatively, and he was, in fact, uncertain whether new structures existed before he observed them: “Nam primum ortum non assequuti, emergentem successivé partium manifestionem expectare cogimur” (1685 edition, p. 577). In sum, according to Adel- mann’s interpretation, he can justly be called a preformationist only with consider- able qualification. As Maitre-Jan was to point out and ex- plain in 1722, the egg examined by Mal- pighi was not what he had considered it—an PROBLEMS, CONCEPTS AND THEIR HISTORY 7, egg studied after exposure to the heat of the August sun in Bologna is “unincubated” only with reference to the hen—but the work was no less influential than had it been founded on a different premise. Mal- pighi’s primary contribution was his success- ful presentation for the first time of visible evidence on the detailed constitution of the young embryo. And evidence adduced by one of the new tools was as certain in the seventeenth as in the twentieth century to draw popular enthusiasm. It was a function of his times that such evidence could be construed as support for embryological theory. What Malpighi saw and figured could be interpreted according to postulates compatible with Descartes’ hypothesis of the infinite divisibility of matter; what he figured could be general- ized into theories implying the embryo to be the same kind of “earthly machine” as Descartes’ adult, and the concepts which were to incorporate his observations into the doctrine of preformation were crystallizing in many minds. Malpighi was no lone prophet of the new embryology. His con- temporaries were going far to cry physical facts to fit a philosophical pattern. Croone, at much the same time, was making some- what similar claims for the pre-existence of the chick in the egg, on the basis of a fan- tastically egregious error, mistaking a frag- ment of vitelline membrane for the embryo (Cole, *44). Swammerdam had in 1672 ex- pressed a somewhat comparable concept for the ege of the frog, and Grew an analogous one for plants the same year. Malebranche, on the basis of observations as well as specu- lation, was expressing similar conclusions and generalizing the doctrine for plants and animals on a strong philosophical founda- tion (Schrecker, ’°38). For Leibniz, preforma- tion was not only a metaphysical but also a strictly biological postulate which he related to his concept of the fixity of species. Bonnet, after his discovery of parthenogenesis in aphids, made preformation the basis for all his biological and philosophical specula- tions; the theory was supported by all the weighty authority of Haller, and even by such advanced experimentalists as Spallan- zani and Réaumur. Vallisnieri’s speculations on the possibility that not only the whole human race but also all human parasites were represented in the ovaries of Eve, and Hartsoeker’s calculation of the necessary size of a rabbit large enough to enclose all rabbits from the beginning of time; Dalenpatius’ absurd claims to have seen the homunculus in the spermatozoon, and all the foolish arguments between ovists and animalculists exemplified the extremes to which the doctrine was led; and such ridiculous claims served primarily to over- burden it until it was close to collapse under its own weight. But it also fell, as it rose, on the basis of more serious philosophical prin- ciples; again a philosophical need had cre- ated a demand which again an observational embryologist—this time Caspar Friedrich Wolff—was to fulfill. Wolff started out with a full apprecia- tion of the philosophical limitations to em- bryological progress implied by the pre- formation doctrine: “Qui igitur systemata praedelineationis tradunt, generationem non explicant, sed, eam non dari, affirmant” (1759; cited from 1774 edition, p. xii); those who adopt the systems of predelineation do not explain generation but affirm that it does not occur. He was to launch his own attack from two sides, from the philosopher’s and the observer’s, but he started from the former’s position: ““Verum explicat genera- tionem, qui ex traditis principiis & legibus partes corporis & modum compositionis de- ducit. .. . Et absoluit theoriam generationis, qui totum corpus generatum ex principus & legibus illis eo modo deducit” (ibid., pp. xii, xii). Deducing the body from principles and laws is the philosopher’s way, not the em- bryologist’s; but Wolff’s virtue was that he felt compelled to supplement his abstruse reasoning by examination of his material and he was thus able to substantiate his theory. Aristotle, as a Greek, had experi- enced no such compulsion; Harvey had had the will but not the way; Wolff had not only the desire, but also the good fortune and the good skill to be both philosophically and observationally accurate within closer limits than his predecessors, and posterity concurs in von Baer’s evaluation of some of his work as at that time “die grésste Meis- terarbeit, die wir aus dem Felde der be- obachtenden Naturwissenschaften kennen” (iss 7s 42). Starting from highly abstract speculations concerning growth and nourishment in their relation to what we should call differentia- tion, he took up in particular detail (to be sure, some years before Goethe, but also, well over a millennium later than Theophras- tus) the metamorphosis of plants, pointing out that the rudiments of leaves are basically similar to those of the parts of the flower and that the rudiments of both alike are derived 8 PROBLEMS, CONCEPTS AND THEIR HISTORY from essentially undifferentiated tissue. This was to lead to his fundamental premise that in animals as well as plants development proceeds by gradual differentiation of orig- inally homogeneous material. There is importance in the fact that he considered the undifferentiated material to be comparable in plants and animals. But though his emphasis on this similarity of construction of material in plant and animal may, as Sachs and Huxley realized though later generations have forgotten, have had effect on the development of the doctrines implying universality of cell and_proto- plasm, it was primarily his emphasis on its early undifferentiatedness that was of more immediate import. It has been said of Wolff, as of Harvey, that he was without influence in his own day, his writings neglected and without effect until after their translation into Ger- man by Meckel in 1812. This is inaccurate. Haller knew his work, and certainly the biological world was kept plentifully and liberally informed of what Haller was think- ing; Kant knew of the concept via Blumen- bach (cf. Critique of Judgment, §81) and from the beginning Kant had the attention of the scientists. Diderot could affirm with confi- dence, as early as 1769, when writing about ‘“‘germes préexistants,” that Cela est contre l’expérience et la raison: contre Vexpérience qui chercherait inutilement ces germes dans l’oeuf et dans la plupart des animaux avant un certain age; contre la raison qui nous apprend que la divisibilité de la matiére a un terme dans la nature, quoiqu’elle n’en ait aucun dans |’entende- ment (Entretien entre D’Alembert et Diderot; writ- ten 1769, first published 1830; cited from 1875 edi- tion, IT, 110). Diderot was no technical embryologist; while it is possible, it is hardly probable that he had read the Theoria generationis. It is far more likely that he was expressing for political reasons an appropriate scien- tific concept which was already sufficiently widely disseminated to have reached his ad- mittedly universal ear. May not this be the clue to Wolff’s suc- cess where Malpighi had failed? The new century had brought new thinking with it, a new thinking in terms of change. Social and political change were soon to grow out of it: revolution and evolution had a com- mon philosophical background; it was even- tually, with Hegel, to reach full fruition as the central doctrine of a specific philosoph- ical system. Wolff’s work was an early ex- pression of this tendency. It is no accidental coincidence that the Christian Wolff who taught philosophy to Caspar Friedrich Wolff, the originator of a biology of change, was a popularizer of Leibniz who had invented a calculus as a mathematics of change. Without this background, it is as unlikely that Wolff would have found a homogeneous blastoderm under his microscope as it was inevitable that Malpighi should have denied one a century before. But Wolff's thinking typified too another kind of thought that was soon to broaden more generally. Wolff was prone to general- ize from plant to animal (cf. “a bat is a perfect leaf . . . for the mode of origin of the two is the same’; from Theorie der Gen- eration, 1764, §64, unavailable to me; cited by Huxley, 1853b, p. 293). His first proof of epigenesis in the chick came from his demonstration that the blood vessels of the chick blastoderm are not present from the beginning; he was probably led to the in- vestigation by his false and far-fetched analogy between the vessels of the animal and those of the plant whose development he had already studied. Certainly if seman- tics gives any indication, his preoccupation with plants colored his later interpretations of observation on animal development. When he demonstrated that animal organs—the in- testine and probably the central nervous sys- tem—are formed by the folding of homoge- neous layers into tubes he called them by the name for leaf. This is a strong hint, as is his “tracing of the body from principles and laws,” of the Naturphilosophie to come, and it is curious, from his own point of view, that the em- phasis laid by history on his accomplishment is centered so strongly on his microscopic discovery of what was not there. In his time, what influence he had probably spoke more positively in the direction of transcendental- ism. Upon his concept of epigenesis and change and upon his intimations of layering in the embryo—both concepts to which Wolff was led by his tendencies towards Naturphilosophie—embryology was to fol- low with its whole momentous sequel, but only after a serious delay during which con- cepts were to arise which in many ways negated the concept of change which Wolff originated, concepts paradoxically enough also derived on a basis of Naturphilosophie. EMBRYOLOGY AND NATURPHILOSOPHIE: GOETHE AND VON BAER But why now Naturphilosophie, whose in- fluence on embryology was to grow so strong PROBLEMS, CONCEPTS AND THEIR HISTORY 9 that its domination is not yet now com- pletely outworn? It was certainly at least in part the clear and inevitable reaction against Cartesianism and against the instillation of the analytical order and system of seven- teenth century mechanics into the study of animate nature. Goethe, one of its warmest partisans, has spoken specifically to this point in his Geschichte meines botanischen Studiums (Gedanken und Aufsdtze, 1944 edition, XII, 314): Vorlaufig . . . will ich bekennen, dass nach Shake- speare und Spinoza auf mich die grésste Wirkung von Linné ausgegangen, und zwar gerade durch den Widerstreit, zu welchem er mich aufforderte. Denn indem ich sein scharfes, geistreiches Abson- dern, seine treffenden, zweckmassigen, oft aber willkiirlichen Gesetze in mich aufzunehmen ver- suchte, ging in meinem Innern ein Zwiespalt vor: Das, was er mit Gewalt auseinanderzuhalten suchte, musste, nach dem innersten Bediirfnis meines Wes- ens, zur Vereinigung anstreben. Goethe is as good an example as any with whom to continue the discussion, not only because he originated the concept of mor- phology in our modern and dynamic sense, but also especially because he was so vividly articulate in describing what went on in his own mind during the process of it. His own studies on the metamorphosis of plants and on the vertebral constitution of the skull, emphasizing the unity of type, and what he thought was his discovery of the intermaxil- lary bone in the human fetus, suggesting that the uniformity of anatomical plan is based on the existence of a developmental archetype, typify the new Naturphilosophie. Natural phenomena represent modifications of an Idea in the Mind of the Creator: here is a new Idealism, less important in that it revivified Plato than that it again lost sight of Aristotle, with as disastrous delaying con- sequences as in the Middle Ages: Agassiz, as late as 1857, could still answer with an unequivocal affirmative his self-addressed question as to whether the taxonomic divi- sions of the animal kingdom have ‘been in- stituted by the Divine Intelligence as the categories of his mode of thinking” (1857, p. 8). This must inevitably appeal, with all of its implications of beauty in nature, to Goethe the poet, who is said to have soothed himself to sleep visualizing a seed growing into a plant. Its mysticism, quite in the neo- Platonic tradition, should have been oppro- brious to the scientist; but this was the mo- ment in history when the scientist turned romantic, to his own loss. The Middle Ages, for all the weaknesses of scholasticism, maintained the firm conviction that the Uni- verse was capable of being understood by human reason; and, as Whitehead (725) re- minds us, it is to medieval scholasticism that we are indebted for our habits of exact thought. The WNaturphilosophen, at their most emotional extremes, grew away from reason in its best sense, and their thought was hardly precise in the sense of modern science. Whitehead stresses too the high standard of objectivity set by the ancient and medieval worlds, with its obvious advantage for science. Its loss was part of the price paid for the developing individuality emerging from the philosophy of the late eighteenth and early nineteenth centuries, and this was a debt whose payment nearly bankrupted the intellectual economy of the Naturphi- losophen. The movement had its philosophical sup- port from Kant, who, like Leibniz before him, laid emphasis on the metaphysical, and who put a premium on transcendentalism; Kant’s categories, after all, were given in advance of experience and the Ding-an-sich was beyond it. Goethe, however, was inde- pendent of Kant. “Meine ‘Metamorphose der Pflanzen,” he told Eckermann, “habe ich geschrieben, ehe ich etwas von Kant wusste, und doch ist sie ganz im Sinne seiner Lehre” (Eckermann, 1905 edition, I, 310). It might have been better if he, and the other Natur- philosophen, had known Kant better. Kant, as Radl (’30, p. 369) has pointed out, “had declared that the Absolute is never known and can never be known; yet his followers,” to continue with Radl, “—the Romantic Philosophers—made this Absolute the basis of their philosophy, the only real thing left in the Universe.” For biology, it was this confusion between the Idea and its representation in the or- ganism, the Absolute and the knowable, that was dangerous. Goethe typifies this, too. He could coolly dictate the rules for observing scientific objectivity (Gedanken und Auf- sdtze, 1944 edition, XII, 93): Jeder Forscher muss sich durchaus ansehen als einer, der zu einer Jury berufen ist. Er hat nur darauf zu achten, inwiefern der Vortrag vollstandig sei und durch klare Belege auseinandergesetzt. Er fasst hiernach seine Ueberzeugung zusammen und gibt seine Stimme, es sei nun, dass seine Meinung mit der des Referenten iibereintreffe oder nicht. “Sobald man in der Wissenschaft einer gewissen beschrankten Konfession angehort,”’ he said to Eckermann, “ist sogleich jede un- befangene treue Auffassung dahin. .. . Es 10 PROBLEMS, CONCEPTS AND THEIR HISTORY gehort zur Naturbeobachtung eine gewisse ruhige Reinheit des Innern, das von gar nichts gestort und praokkupiert ist” (Ecker- mann, 1905 edition, II, 218, 220). He could, however, as sublimely ignore his own precepts. Where was his inner pur- ity without preoccupation, where was his in- dependence of a particular confession, when Eckermann came in to him with news of the July Revolution to have him cry (ibid., I, 473): Nun... was denken Sie von dieser grossen Bege- benheit? Der Vulkan ist zum Ausbruch gekommen; alles steht in Flammen, und es ist nicht ferner eine Verhandlung bei geschlossenen Thiiren, and heard him reply, when Eckermann spoke of ministers and royal family (ibid., II, 474475): Ich rede gar nicht von jenen Leuten; es handelt sich bei mir um ganz andere Dinge. Ich rede von dem in der Akademie zum 6ffentlichen Ausbruch gekommenen, fiir die Wissenschaft so héchst bedeu- tenden Streit zwischen Cuvier und Geoffroy de Saint-Hilaire! . . . Die Sache ist von der héchsten Bedeutung. . . . Wir haben jetzt an Geoffroy de Saint Hilaire einen machtigen Alliierten auf die Dauer. ... Das Beste . . . ist, dass die von Geoffroy in Frankreich eingefiihrte synthetische Behand- lungsweise der Natur jetzt nicht mehr riickgangig zu machen ist.... Von nun an wird auch in Frank- reich bei der Naturforschung der Geist herrschen und iiber die Materie Herr sein. Man wird Blicke in grosse Schépfungsmaximen thun, in die geheim- nisvolle Werkstatt Gottes!—Was ist auch im Grunde aller Verkehr mit der Natur, wenn wir auf analyti- schem Wege bloss mit einzelnen materiellen Teilen uns zu schaffen machen, und wir nicht das Atmen des Geistes empfinden, der jedem Teile die Rich- tung vorschreibt und jede Ausschweifung durch ein innewohnendes Gesetz bandigt oder sanktioniert! Here is the romantic fallacy that lies at the hollow core of Naturphilosophie: here it is that the Naturphilosophen separate from Kant. Kant did not question the valid- ity of natural science in its own realm: in- deed, he justified it. He simply defined the regions in which it could operate, while the Naturphilosophen with their zeal for syn- thesis and their preoccupation with the spirit as the synthesizing element related the real to the transcendent in such a confused way that they could think clearly on neither. Idealism for the philosopher is one thing: Kant felt that science could be accurate only when mathematically expressed, which is one kind of idealism. Huxley had the same intuition; in a paper on the Mollusca he wrote (1853a, p. 50): From all that has been stated, I think that it is now possible to form a notion of the archetype of the Cephalous Mollusca, and I beg it to be under- stood that in using this term, I make no reference to any real or imaginary “ideas” upon which animal forms are modelled. All that I mean is the concep- tion of a form embodying the most general proposi- tions that can be affirmed respecting the Cephalous Mollusca, standing in the same relation to them as the diagram to a geometrical theorem, and like it, at once imaginary and true. Boyle had presented the problem earlier to the physical scientist. His law was set for the ideal gas, and it became the task of the scientist to check experimentally the be- havior of the real gas against that postu- lated for the ideal. Such a conception lacks meaning to the biologist; no such experi- ment is possible for him in relating the real to the ideal set up by the Naturphilosophen. Neither Boyle’s kind of idealism, nor Hux- ley’s, is that of the MNaturphilosophen, whose weakness was not so much that it left no room for the experiment as that it closed their minds to whole systems of possible in- terpretations of the observed phenomena which they collected to gain credence for their fancies. The weakness of the Naturphilosophen by and large was that they tried to force a rigid and fixed and obvious structure out of Spi- noza’s deeper and more fluid and subtle pan- theism. Goethe, with more strength and with more sensitivity, could like Herder pass be- yond them to be carried away by the dynamic wholeness of nature which to him was alive in the sense of the new morphology which was to follow later. Goethe, too, could grow beyond the romantic in other realms of thought than the scientific; but the profes- sional biologists largely lacked his profundity and maturity and remained at the static phase too long. While Goethe’s significance as a prophet for Naturphilosophie is hardly to be minimized, there were others who were to bear the responsibility for working out the biological details and who carried the doc- trine to the illogical extremes which were so to retard the progress of biology proper: Goethe’s friend, Nees von Esenbeck, who con- sidered the entire vegetable world a leaf; Goethe’s hero Etienne Geoffroy Saint-Hilaire; and Serres, who said, as Oken was saying too, that the entire animal kingdom was a single organism; Oken was going so far as to com- pare the parts of a plant to fire, water, earth and air. Yet it was against this dark background that the students of the natural philosopher Dollinger (cf. Temkin, *50) at Wiurzburg PROBLEMS, CONCEPTS AND THEIR HISTORY sas were to begin to build constructively upon Wolff’s concepts; and the fact that they could start to do so is related probably to Wolff's own compatibility with Naturphilosophie. Wolff's epigenesis had started as conceptual; his concept, and the results of his own micro scopic examinations, led those who were to follow him to the material where they could build upon what he had postulated and dem. onstrate a mechanism of the process of change he had postulated. The group consisted of Pander, who first demonstrated the existence of the three pri- mary germ layers in the embryo of the chick; of Goethe’s friend D’Alton, who acted as artist; and of von Baer, who generalized Pander’s germ layers for other animals and who in so doing generalized the science of embryology itself. Pander’s advance was a great one, in a Way, in terms of independence of thought; and his achievement, in an environment of overgeneralization, in being able to concen- trate on describing specific processes of de- velopment in a single form without drawing far-fetched analogies, was considerable. But Pander could not, or did not, carry through, and it was left to von Baer, or rather, von Baer took it upon himself, to broaden the base by the examination of more varied material. With his inspiration from the romantics, he looked at the diverse material with a question in his mind as to its comparability; and he came away from it with the convic- tion that the comparability was there, in terms of origin (hence the discovery of the mammalian egg), and in terms of process in the similarity of the formation of the germ layers and in the derivation of similar or- gans from comparable layers in the different vertebrate forms. He demonstrated develop- ment to be at once from homogeneous to heterogeneous, from general to special, in all the forms that he studied. Though his feat was an overwhelmingly intellectual, not technical, achievement, his great advance was the extent to which he based his con- clusions on the zealous and accurate and untiringly meticulous microscopic observa- tions on a wide variety of animal material: Beobachtung preceded Reflexion in his title. His emphasis on comparability involved, to be sure, as did that of the other Natur- philosophen, an emphasis on Type: “Zufrie- den wiirde ich seyn,” he wrote, “wenn man es als meinen Antheil betrachtet, nachge- wiesen zu haben, dass der Typus der Orga- nisation die Entwickelungsweise bedingt” (1828, I, xxi1). But he meant by Type some- thing different than the others: Vor allen Dingen mache ich darauf aufmerksam, dass man den Grad der Ausbildung des thierischen Korpers und den Typus der Organisation unter- scheiden muss. Der Grad der Ausbildung des thieri- schen K6érpers besteht in einem groéssern oder gerin- gern Maasse der Heterogenitat der Elementartheile und der einzelnen Abschnitte eines zusammenge- setzten Apparates, mit eimem Worte, im der grossern histologischen und morphologischen Son- derung. . . . Typus nenne ich das Lagerungsver- haltniss der organischen Elemente und der Organe (1828, I, 207-208). Our persuasion, that the grades of development must be distinguished from the types of organiza- tion, is founded upon the following considerations: —We know that all the functions of the perfect animal body contribute to a general result,—to the life of the animal; but also that the general mass manifests the total life (for animal life is always a totality)... . With a greater separation and more complete independence of these functions is com- bined a greater differentiation of the body into or- ganic systems, and of these systems again into sep- arate more individualized sections. In this consists the higher development of the animal body. But the mode in which these organs of the animal body are united together, is a wholly distinct mat- ter. And it is to this manner in which the organic elements are combined that we give the name of Type. Every type may be manifested in higher and lower degrees of organization; the type and the grade of development together determine the spe- cial forms (1826; cited from 1853 edition, pp. 178- 179). Here there is an implication still of the Type and Archetype of the Naturphiloso- phen, but it is becoming more Type in com- mon with Aristotle’s form in the sense of potentiality. While von Baer has adopted a concept from Naturphilosophie, he has de- veloped it further; for von Baer, it is the embryo, not the Idea, that is becoming the type. It is irrelevant for our purposes that he considered the primary types to be those of the vertebrate, the annulate, the radiate and the mollusk (the double symmetrical, the longitudinal, the radiate and the spiral; polarity and symmetry were a central idea, both problem and metaphysical reply to it, for the Naturphilosophen as for modern bi- ologists). What is significant is that he could regard them, rightly or wrongly, as separate types of extant, visible, dissectable and ob- servable animals perceived by his sense or- gans. This is phenomenological type, type not in an Idea but present as structure in an adult organism, and if masked there, some- times discernible in the structure of the embryo; and thereby the relationship of £2 PROBLEMS, CONCEPTS AND THEIR HISTORY embryology to comparative anatomy be- comes fixed for all time. The extent to which he could emancipate himself from the tendency to overgeneralize of the Naturphilosophen is probably no- where made clearer than in his refusal to accept the “law” of parallelism most clearly expressed before his time by Meckel (for dis- cussion of earlier contributors to the develop- ment of this concept see especially Needham, ’34, and Meyer, ’35; there are those who con- sider the doctrine to be foreshadowed even by Aristotle and Harvey, an interpretation which Meyer quite justly disputes): Dass der Embryo hoéherer Thiere, ehe er seine vollkommne Ausbildung erreicht, mehrere Stufen durchlauft, wurde schon oben bemerkt; hier ist nachzuweisen, dass diese verschiednen Stufen denen entsprechen, auf welchen tiefer stehende Thiere das ganze Leben hindurch gehemmt_ erscheinen (Meckel, 1821, I, 396-397). Von Baer, in contrast, with his emphasis on difference in adult type, denies the validity of Meckel’s “Law” (1828, I, 220): Dadurch ist aber nicht erwiesen, dass jeder Em- bryo einer héhern Thierform allmahlig die niedern Thierformen durchlaufe. Vielmehr scheint sich der Typus jedes Thiers gleich anfangs im Embryo zu fixiren und die ganze Entwickelung zu _beherr- schen ... Der Embryo des Wirbelthiers ist schon anfangs ein Wirbelthier. . . . Mithin durchlaufen die Embryonen der Wirbelthiere in threr Entwicke- lung gar keine (bekannten) bleibenden Thierfor- men. Frequent misconceptions have been ex- pressed concerning von Baer’s relationship to the formulation of the biogenetic law which have been well clarified by Meyer (35); misconceptions which have arisen probably at least in part because of Darwin’s quotation, in later editions of the Origin of Species, of von Baer’s passage accentuating, in line with his stress on development from the general to the special, the likeness of young vertebrate embryos: The embryos of mammalia, of birds, lizards, and snakes, probably also of chelonia are in their earliest states exceedingly like one another, both as a whole and in the mode of development of their parts; so much s0, in fact, that we can often distinguish the embryos only by their size. In my possession are two little embryos in spirit, whose names I have omitted to attach, and at present I am quite unable to say to what class they belong (Darwin, 1902 edition, II, 241), and his stress was exclusively on resem- blances between embryos rather than be- tween adults of one group and embryos of another. Darwin might have made von Baer’s posi- tion clearer had he cited the delightful passage from the Fifth Scholion (1828, I, 203-204): Denke man sich nur, die Vogel hatten ihre Ent- wickelungsgeschichte studirt, und sie waren es, welche nun den Bau des ausgewachsenen Sauge- thiers und des Menschen untersuchten. Wirden nicht ihre physiologischen Lehrbiicher Folgendes lehren kénnen? “Jene vier- und zweibeinigen Thiere haben viele Embryonenahnlichkeit, denn ihre Schadelknochen sind getrennt, sie haben keinen Schnabel, wie wir in den fiinf oder sechs ersten Tagen der Bebriitung; ihre Extremitaten sind ziem- lich gleich unter sich, wie die unsrigen ungefahr eben so lange; nicht eine einzige wahre Feder sitzt auf ihrem Leibe, sondern nur diimne Federschafte, so dass wir schon im Neste weiter sind, als sie jemals kommen, ihre Knochen sind wenig spréde und ent- halten, wie die unsrigen in der Jugend gar keine Luft; iberhaupt fehlen ihnen die Luftsacke und die Lungen sind nicht angewachsen, wie die unsrigen in frithester Zeit; ein Kropf fehlt ihnen ganz; Vor- magen und Muskelmagen sind mehr oder weniger in Einen [sic] Sack verflossen; lauter Verhaltnisse, die bei uns rasch voriibergehen, und die Nagel sind bei den meisten so ungeschickt breit, wie bei uns vor dem Auskriechen; an der Fahigkeit zu fliegen haben allein die Fledermause, die die vollkommen- sten scheinen, Theil, die iibrigen nicht. Und diese Saugethiere, die so lange nach der Geburt ihr Futter nicht selbst suchen kénnen, nie sich frei vom Erd- boden erheben, wollen héher organisirt seyn, als wir? It must be granted that von Baer himself sometimes indulged in flights of fancy com- parable to those of the other Naturphiloso- phen. “Dass namlich Kiefern und Extremi- taten Modificationen eines Grundtypus sind,” he wrote, “ist augenscheinlich . . . Die Kiefern aber nahern sich so sehr der Natur der Rippen, dass man von ihnen einen Grund hernehmen kann, auch die Extremitat des Rumpfes fiir verstarkte Rippen anzuse- hen” (ibid., I, 192). His theories of fertili- zation, and of the significance of symmetry with regard to type, were as foolish as those of his contemporaries; but his strength in being so frequently able to overcome such temptation was more remarkable than the occasional symptoms of his succumbing to it. He had on the whole a particularly clear appreciation of his contemporaries’ confu- sion between fact and idea: he stated it ex- plicitly in one of his criticisms of the theory of parallelism: “Man lernte allmahlig die verschiedenen Thierformen als aus einander entwickelt sich denken—und schien dann, von einigen Seiten wenigstens, vergessen zu wollen, dass die Metamorphose nur eine Vorstellungsart sey” (ibid., I, 200). PROBLEMS, CONCEPTS AND THEIR HISTORY 13 Von Baer’s own embryology, for the first time, for all of its emphasis on the relation- ship of the special to the general, was an embryology in which the metaphysical be- came subordinate to the biological in the sense of modern embryology, and became an embryology which proceeded from embry- ological facts and phenomena towards em- bryological concepts, rather than in the reverse direction; and von Baer, in accom- plishing this feat, made one of the greatest advances in all biological history. His force of intellect, his consequent self-mastery and ability to free himself to develop beyond the thought in which he was trained, are un- matched in biological progress. He could emancipate himself from the thinking of his times more than Vesalius, more than Harvey before him; more than Darwin after him, and though perhaps not in an analytical sense, yet in a synthetic sense more than Mendel to follow. His courage to maintain his own inde- pendence of thought may have been fed by his century’s new kind of awareness of the individual, with its philosophical back- ground from Leibniz and Kant, developed by Fichte and Schelling and Hegel, and its translation during the eighteenth century into more widespread acceptance than ever before of the significance of freedom of indi- vidual action. Von Baer absorbed the con- cept of independence: “Deswegen ist auch das wesentlichste Resultat der Entwickelung, wenn wir sie im Ganzen tibersehen, die zunehmende Selbststindigkeit des werden- den Thiers” (zbid., I, 148), he wrote, and he concentrated on the individual. Wesen- heit was all: “Die Wesenheit des Thiers beherrscht die Ausbildung. . . . Die Wesenheit ... der zeugenden Thierform beherrscht die Entwickelung der Frucht” (ibid., I, 147-148). But like Aristotle and Goethe, von Baer was obsessed with the dynamic and func- tional qualities of the organism as a whole. “We know that all the functions of the per- fect animal body contribute to a general re- sult,— to the life of the animal,’ we have already quoted, “but also that the general mass manifests the total life (for animal life is always a totality).” “Die Entwickelungs- geschichte des Individuums ist die Ge- schichte der wachsenden Individualitat in jeglicher Beziehung” (ibid., I, 263), he wrote too; and he thought of the growing individual too, like Goethe, in respect to a larger whole; to him the palm, “dem es vor- behalten ist, die bildenden Krafte des thier- ischen Kérpers auf die allgemeinen Krafte oder Lebensrichtungen des Weltganzen zuriickzufuuhren” (ibid., I, xxii). Like Goethe, he could say, and in almost the same figure of speech, that “die Ge- schichte der Natur ist nur die Geschichte fortschreitender Siege des Geistes tiber den Stoff’ (1864, pp. 71-72), but he meant it in a different context (1828, I, 263-264): Hat aber das eben ausgesprochene allgemeinste Resultat Wahrheit und Inhalt, so ist es Ein Grund- gedanke, der durch alle Formen und Stufen der thierischen Entwickelung geht und alle einzelnen Verhaltnisse beherrscht. Derselbe Gedanke ist es, der im Weltraume die vertheilte Masse in Spharen sammelte und diese zu Sonnensystemen verband, derselbe, der den verwitterten Staub an der Ober- flache des metallischen Planeten in lebendigen For- men hervorwachsen liess. Dieser Gedanke ist aber nichts als das Leben selbst, und die Worte und Syl- ben, in welchen er sich ausspricht, sind die verschie- denen Formen des Lebendigen. But if von Baer’s outlook, like Goethe’s, was cosmic in scope, his inspiration was the detailed and specific study of the developing form of the individual living embryo. Like Aristotle, he accepted form as inseparable from the formed. The advantage of refine- ment bestowed by time to his philosophical and technical method, as compared with Aristotle’s, enabled him to concentrate more on the formed, which was biological ma- terial, than on form as such and metaphys- ical; and in so advancing he became the true synthetic genius which Goethe had aspired to be, “der grésste unter uns in Vergangen- heit, Gegenwart und weiter Zukunft” (Roux, 1889; cited from Roux, 1895b, IT, 25). EMBRYOLOGY AND EVOLUTION: DARWIN AND HAECKEL It has been said that all biology since Darwin has been a commentary on the Origin of Species. Embryology would be in a more advanced position than its present one if one could claim that all embryology after von Baer represented a commentary on his great treatise. It is true that immedi- ately following his time, even during it, great strides were made in the amplification and refinement of his teachings. Rathke, in particular, whose quality of mind was in many ways like that of von Baer, made par- ticular advances in demonstrating the exis- tence of the germ layers in invertebrates as well as vertebrates and in discovering the presence of gill slits in the mammalian em- bryo. The universality of the germ layers was given new meaning with the enuncia- tion of the doctrines of universality of proto- 14 PROBLEMS, CONCEPTS AND THEIR HISTORY plasm and of cells when von Kolliker and Remak and others brought together the re- sults of these with the results of the germ layer doctrine. Had these continued as the main trends of embryology, von Baer’s syn- thetic scheme would have been broadened in the fashion it deserved. Instead, a return to the overgeneralization of Naturphiloso- phie once more delayed its progress into the future. One symptom of this was the continued emphasis, in spite of von Baer’s warnings, on the comparability of embryos of “higher” forms to the adults of “lower,” and such false analogy was carried over even into the germ layers, von Baer’s own territory, thus seem- ing to be supported by his facts though he had so explicitly denied the concept. Huxley, in 1849 (p. 426), described the Medusae as constructed of two membranes “which ap- pear to bear the same physiological relation to one another as do the serous and mucous layer of the germ,” opening the way for the ultimate generalization. When, therefore, shortly after the publication of the Origin of Species, Kowalewski found that invertebrate and vertebrate embryos alike formed from a bilaminar sac, that in the most varied material—Psolinus, Amphioxus, Phoronis, Limnaeus, Ophiura, Echinus, Asteracanthion, Sagitta, Ascidia, Escholtzia, Sepiola as well as birds, mammals and turtles— . . . bei allen von mir hier erwahnten Embryonen geht die Bildung der beiden erwahnten Schichten oder Blatter (der ausseren und inneren) ganz auf dieselbe Weise vor sich. . . . Also wire die erste Bildung des Embryo fiir alle diese verschiedenen Thiere ganz iibereinstimmend; nur in den weiteren Veranderungen sehen wir die Unterschiede auftre- ten, welche jeden einzelnen Typus bezeichnen (1867, p. 5), the decision was sealed. He ended on the same note of caution as von Baer, but his voice too was drowned out by the clamor originating from a new cry of transcenden- talism that surpassed anything the earlier Naturphilosophen would have dreamed pos- sible. The difference, of course, was that meantime the Origin of Species had ap- peared. The compulsion to synthesize all of ani- mate nature into a single grandiose scheme which the false analogies of the earlier tran- scendentalism had formerly satisfied was now to be assuaged by the evolution doctrine, which represented a synthesis on another basis; in the new scheme, common descent replaced the archetype as the primary syn- thesizing factor. It has been said that it was Darwin who “dragged [organisms] down from. . . meta- physical regions into daily life, and ex- amined their immediate purpose in relation to the whole environment of the living or- ganism” (Radl, ’30, p. 381). But Darwin in some ways advanced no more abruptly in respect to the descent from metaphysics than had von Baer, and the structure of thought he created was in many ways as metaphysical as that of his predecessors. The influence of the environment alone was no new concept; philosophy had been worrying about this problem at least since the time of Leibniz. The species concept, from which Darwin started, was so highly metaphysical that even now the term defies adequate biological definition. Darwin’s system was a metaphysical one, too, in that his concern for the individual organism was subordinate to his interest in the interdependence of or- ganisms; his clue to the nature of their re- lationships came equally from the organisms themselves and from wider areas of thought: from generalizations invented for the fields of geology and economics, and indeed, in a way from his whole century’s mood for “Progress.” One of the primary contributions of Dar- win, however—indeed of all those concerned with the new doctrine of evolution: Buffon, Erasmus Darwin, Lamarck, John Hunter, Wallace and the many others—was the re- focussing once more of attention on the or- ganism as a whole. The key to evolution had come from consideration of the whole living animal, not its parts; evolution was inferred from natural history, not deduced from the preparations in the cabinets of the anatomists where the evidence had been awaiting for centuries the interpretations which the fixity and the tenacity of the no- tions of unity of type had excluded from coming into being. It is one of the more curious ironies of history that while before Darwin, transcen- dentalism had closed the minds of investiga- tors to the possibility of explanation of re- semblances between parts of organisms and between whole organisms on the basis of common descent, yet after him the combina- tion of the doctrines was to lead to extremes of exaggeration that were attained sena- rately by neither. There were many who were to contribute to this: Kleinenberg in Germany and Lankester in England made an early start by relating phylogeny to ontogeny on the basis of comparability of the germ layers (for fuller treatment see PROBLEMS, CONCEPTS AND THEIR HISTORY 1) Oppenheimer, *40), but their views were relatively mild compared to those of many who followed them. The culmination was the work of Haeckel, the greatest revisionist of them all. The most extreme example of his immod- eration was perhaps his naming, on the basis of the “similarity, or homology, of the gas- trula in all classes of compound animals” (concept originally expressed in monograph on the Kalkschwaémme, 1872; cited here from Haeckel, [1900], p. 61), in the stead of the coelenterates formerly nominated for the position by Kleinenberg, the imaginary gas- traea as the progenitor of all multicellular forms. His figuring of a section through an animal that never existed on the same page (1891, p. 161) that illustrates Kowalewski’s gastrulae of Sagitta and Amphioxus and Carl Rabl’s of Limnaeus, with no comment in the label to signify that the “Gastrula eines einfachsten Pflanzenthieres, einer Gastraeade (Gastrophysema, Haeckel)” is any less real than the others—where is there a handsomer example in all biological or scientific his- tory of what Whitehead has called the “Fallacy of Misplaced Concreteness”’? Such a silly invention as the gastraea, that ‘““magere Tiergespenst,’ as Kleinenberg (1886, p. 2) called it, as an isolated case might probably have proved of little influ- ence; and its significance is as a symptom (a word used advisedly for its pathological con- notations) of Haeckel’s basic trouble. What was damaging to science was Haeckel’s fer- vency to oversystematize all morphology through his biogenetic law that “die Onto- genie ist eine Recapitulation der Phylogenie” S91. 7). In formulating it, he returned to the law of parallelism of Tiedemann, Meckel and Serres, quite by-passing von Baer’s more temperate statements. He was influenced to do so of course in part by Fritz Miiller, who had earlier (Fiir Darwin, 1864) pointed out on the basis of the study of crustacean larvae that individual development provides a clue to ancestral history. But he misinter- preted Miller, as have many more modern readers (cf. Meyer, ’35). While Miller was supporting Darwin—indeed, as Meyer says (35, p. 392), “his main conclusion was that his studies on the development of crustacea confirmed Darwin’s idea of evolution”—he yet was formulating no such dogma as Haeckel’s concerning the causal relation- ships between evolution and individual de- velopment. Sir John Lubbock, too, was early consider- ing the relationships of evolution and indi- vidual development in support of Darwin; but when he questioned whether insects dur- ing the course of metamorphosis pass through their ancestral stages he felt forced to a negative reply in the absence of evidence that a caterpillar ever existed as a fully de- veloped organism. Radl has commented (’30, p. 140) concerning the biogenetic law that “everything important that has ever been cited against the theory was known when the theory was first put forward; neverthe- less it was widely accepted.” Lubbock’s res- ervation is an example. But his exception, and all the other exceptions, seemed to lack the dramatic appeal of the false generaliza- tion, and the biogenetic law was acclaimed with the same rapt enthusiasm that had ereeted the earlier theories of preformation and of unity of type. Investigators in widely varied fields of interest rapidly carried over the theory into their own territories. Bunge applied it to physiology: The amount of common salt in the organism cor- responds with the amount in the environment... . Many plants contain only traces of sodium; those which are rich in it are only the sea-weeds and the plants which grow on the sea-shore, and on the salt-steppes which are dried-up sea-basins. .. . This is also the case with invertebrate animals; only those which live in the sea, and those nearest allied to them on land, contain much salt... . The land vertebrates are all remarkably rich in salt, in spite of the scanty supply around them. But even these are only apparent exceptions. We need but remember the fact that the first vertebrates on our planet all lived in the sea. Is not the large amount of chlorid of sodium found in the present inhabitants of dry land another proof of the gene- alogical connection we are forced to accept from morphological facts? ... If this interpretation is correct, we should expect that the younger the vertebrates are in their in- dividual development, the more salt they would possess. This is in fact the case. I have convinced myself by numerous experiments that an embryo of a mammal contains more salt than a new-born animal, and that it gradually becomes, after birth, poorer in chlorin and sodium as it develops. Cartil- age contains the most sodium of any tissue in our bodies, besides being also the tissue of greatest an- tiquity. . . . This phenomenon . . . can only be explained by the theory of evolution” (’02, pp. 101- 103). Workers in other fields than biology, too, adopted the theory with as much warmth. Preyer, a colleague of Haeckel’s at Jena, for- mulated his conceptions of child psychology with reference to Haeckel’s law; Herbart and Ziller before him had held that the in- 16 PROBLEMS, CONCEPTS AND THEIR HISTORY dividual repeats in his development the stages of cultural development through which the human race has passed. And in our own times, Jung, following Nietzsche and Freud among others, with all his im- measurable influence on modern psychology and literature, has erected the superstruc- ture of his Psychology of the Unconscious on the acceptance of Haeckel’s premise as though this were the immutable truth that Haeckel in his own day had hoped it: All this experience suggests to us that we draw a parallel between the phantastical, mythological thinking of antiquity and the similar thinking of children, between the lower human races and dreams. This train of thought is not a strange one for us, but quite familiar through our knowledge of comparative anatomy and the history of develop- ment, which show us how the structure and func- tion of the human body are the results of a series of embryonic changes which correspond to similar changes in the history of the race. Therefore, the supposition is justified that ontogenesis corresponds in psychology to phylogenesis. Consequently, it would be true, as well, that the state of infantile thinking in the child’s psychic life, as well as in dreams, is nothing but a re-echo of the prehistoric and the ancient” (’27, pp. 27-28). But the blind adoption of Haeckel’s doc- trines by such workers in bordering fields, and their infection with his faith that ‘‘de- velopment is now the magic word by means of which we shall solve the riddles by which we are surrounded” (cited from Radl, ’30, pp. 126-127), is less reprehensible than their uncritical acceptance by the professional embryologists, who swallowed them with as much gullibility, and who remained utterly unperturbed by the fact that Haeckel him- self was never in any sense a professional embryologist. The seduction of embryology by a fanatic who expressed himself even metaphorically in terms of magic represents a darker chapter in its history than any of its earlier or later retreats to mere meta- physics lacking such taint of the mystic. Deplorably enough, the record of many of our “modern” textbooks is none too pure with respect to the biogenetic law. But there is no space here for a modern critiqué of the doctrine (for brief statements of the modern position see Shumway, ’32, and de Beer, 51); what is relevant here at the moment is not so much Haeckel’s rightness or wrongness as the magnitude of his influence. It was considerable, and acted as a delaying rather than an activating force; and it was stifling to immediate progress, since embryologists were for many years after to examine em- bryos primarily to establish evidence of phy- logenetic relationship. This was not wholly detrimental, of course; like the earlier tran- scendentalism this gave a strong incentive for looking at embryos, and many accurate observational data were collected which were later to stand embryology in good stead; but progress in terms of new concepts was neces- sarily impeded. Balfour specified the task, prescribed the fashion, set the standard (1880, I, 4-5): To test how far Comparative Embryology brings to light ancestral forms common to the whole of the Metazoa. . How far .. . larval forms may be interpreted as the ancestral type. . How far such forms agree with living or fossil forms in the adult state... . How far organs appear in the embryo or larva which either atrophy or become functionless in the adult state, and which persist permanently in mem- bers of some other group or in lower members of the same group... . How far organs pass in the course of their devel- opment through a condition permanent in some lower form... . Balfour himself acknowledged another de- partment of embryology concerned with the origin of organs and germ layers and tissues, but to this he devoted only a quarter of his great treatise; many of his contemporaries more fully ignored it, and advance had to wait until the furor over Darwin and re- capitulation had subsided. The degree to which evolutionary relationships dominated embryology is nowhere better shown than by the results of the few cases where investi- gators attempted to pursue other paths, and failed in influence. Leuckart and Bergmann had, in fact, sev- eral years before the publication of the Origin of Species, already set the programme for a new embryology (1851; cited from 185557p:736): Ebenso wie man gegenwéartig strebt, die Com- bination von Wirkungen zu ermitteln, auf welcher eine bestimmte Krystallform oder die Bildung und Umbildung der Zelle beruht, so wird man sich auch Wege zu er6dffnen suchen, um die bewirkenden Ursachen der Anordnung der Organe zu ermitteln: man wird eine Physiologie der Plastik dereinst anstreben. But the few who had the originality, during the nineteenth century, to attempt to work out a “physiology of the plastic” were doomed to failure. Lereboullet made an at- tempt to do so, in France (cf. Oppenheimer, °36), where he could do so in part because of the dominating spirit of Cuvier, who, like von Baer, emphasized animals rather than PROBLEMS, CONCEPTS AND THEIR HISTORY NY) relationships and their structure rather than their metaphysics. In part because of Cuvier’s strength of mind, Darwin never attained the same heights of scientific popularity in France as in Germany and England; and if Lereboullet was not of such intellectual cali- bre as to take full advantage of this for em- bryology, French science still benefited from it in the persons of Pasteur and Claude Ber- nard. In Germany, Wilhelm His in 1874, Goette in 1875, and Rauber in 1880 attempted mechanical explanations of development, working towards a “physiology of the plas- tic’; they attracted a few strong disciples such as von Koélliker, but on the whole they cried in the wilderness. His’ cry was the most explicit, perhaps, certainly one of the most violent against Haeckel; and quite the obverse, in many ways, of what Balfour was later to declaim (1874, pp. 161, 171-172, 174-175): Gegenstand und Methode der phylogenetischen Forschung . . . sind durchaus andere als diejenigen der von mir bearbeiteten physiologischen Entwick- lungsgeschichte des Individuums . . . Das nachste Interesse fiir uns liegt in der... . formulirten Frage: in wie weit die phylogenetische Geschichte einer Form zugleich als deren Erklarung gelten darf, und wie sich ihre eventuelle Erklarung verhalt zur physiologischen Erklarung? .. . Ich behaupte nun, die Kérperform ist eine unmit- telbare Folge des Keimwachsthums, und bei gege- bener Anfangsform des Keimes aus dem Gesetze des Wachsthums abzuleiten .. . Weiterhin ist aber das Keimwachsthum eine Folge der Eigenschaften des eben befruchteten Keimpro- toplasmas. Diese sind eine Folge von den Eigen- schaften der elterlichen Keimstoffe und der Art ihres Zusammentreffens u.s.w. Wir bekommen somit folgende Reihenfolge zu leistender Erklarun- gen: 1) Erklarung der Koérperform aus dem Wachs- thum des Keimes; 2) Erklarung des Keimswachsthums aus den Eigenschaften des befruchteten Keimprotoplasmas und aus den Bedingungen seiner Entwickelung (Temperatur, Ernahrungsbedingungen u.s.w.). 3) Erklarung der Eigenschaften des befruchteten Keimprotoplasmas aus den Eigenschaften der elter- lichen Keimstoffe und der besonderen Bedingungen ihres Zusammentreffens; 4) Erklarung der Eigenschaften der Keimstoffe aus dem Gange der elterlichen Kérperentwickelung; 5) Erklarung der besonderen Bedingungen der Befruchtung aus den Lebensverhaltnissen der beiden Erzeuger und so fort. Erst mit Nr. 5 der obigen Kette beginnt das Gebiet der phylogenetischen Erklarung, und es erstreckt sich von da in periodischer Wiederkehr ins Uner- messliche nach riickwarts. But to no immediate avail did His attempt to remove explanation from the level of the transcendental and ideal, and from the level of metaphysical relationships between or- ganisms, to the level of the embryo itself; Haeckel spoke too strongly in opposition. Haeckel’s greatest disservice, after all, was not his simple ignorance of the morpholog- ical exceptions to his law as a descriptive statement, but his emphasis on it as an irre- futable explanation of causal relationships. Transcendentalism of Haeckel’s variety was as fundamentally incompatible as had been the other kind with any concept of process. “Die Phylogenie,” he insisted, “ist die me- chanische Ursache der Ontogenese” (1891, p. 7), not only distracting to other areas the many who might have otherwise become in- terested in true mechanical explanations, but refuting, as he thought, irrevocably those who were already involved in develop- ing such interests. He felt His a particular foe, as well he might have; and his polemics against him, since they were inadequate to combat His on his own grounds, descended to ridicule of the most inane sort: ... [Es] lasst sich aus dem Studium der ontogeneti- schen Arbeiten von His bald erkennen, dass in seiner Vorstellung die bildende “Mutter Natur” weiter Nichts als eine geschickte Kleidermacherin ist. Durch verschiedenartiges Zuschneiden der Keim- blatter, Kriimmen und Falten, Zerren und Spalten derselben gelingt es der genialen Schneiderin leicht, alle die mannichfaltigen Formen der Thier- arten durch “Entwickelung” (!) zu Stande zu bringen. Vor Allem spielen die Kriimmungen und Faltungen in dieser Schneider-Theorie die wich- stigste Rolle. . . . Am possirlichsten ist, wie die Schneiderin bei Fabrication der zwei Paar Glied- maassen verfahrt . . . . Doch wird diese herrliche “Briefcouvert-Theorie” der Wirbelthier-Beine noch ubertroffen durch die ‘Héllenlappen-Theorie,” welche His von der Entstehung der rudimentdren Organe giebt. . . . Hier wirft also die schneidernde Natur die iiberfliissigen Gewebslappen hinter den Ofen, in die “H@lle” ! (1891, pp. 53-54). Haeckel, however, can hardly have been expected to accept His’ whole cloth. White- head has spoken, in connection with his Fallacy of Misplaced Concreteness, and thinking surely of precisely such mentalities as Haeckel’s, of those “clear-cut trenchant intellects, immovably encased in a hard shell of abstractions [who] hold you to their ab- stractions by the sheer grip of personality” (725, p. 82). But it was not mere personality that won for Haeckel his felicitous reception. His personality may have made him think as he did; but it was the eagerness of his contemporaries for oversystematization, in terms of just such abstractions as his, that was responsible for the success of his doc- 18 PROBLEMS, CONCEPTS AND THEIR HISTORY trine. Its supplanting had therefore to wait until a new period demanded a different kind of thought, and until Wilhelm Roux could succeed where Leuckart and Bergmann and His had apparently failed. The attempts at mechanical explanations of development begun by His and the others in the 1870’s were surely an outgrowth of the whole philosophy of materialism that pervaded the thinking of the 19th century. But the 19th century was also a strongly romantic century in many ways, perhaps again specifically in reaction against mate- rialism. It is a curious paradox that the Haeckelian doctrines, steeped in romantic idealism, emanated from the impersonal and objective doctrines of Darwin. But they did, and in doing so, flourished with sufficient strength to repress the mechanical theories of development really so much more com- patible with the doctrines of evolution; and it is ironical that therefore the success of the new embryological theories had to await a waning of interest in Darwinism. It must be admitted, however, that al- though evolution had delayed the new move- ment in one way, it is also true, in another, that it fostered it. The new interest in the new embryology came again from a new at- tack on the problem of epigenesis versus preformation, and this derived its origin from the absorption of the evolution doctrine into everyday thinking. The process of evo- lution implies epigenesis, in that change is the essence of both, and a gradual process from step to step. A generation habituated to thinking about change building on change in evolution could more easily than its fathers accept the concept of epigenesis with its causal connotations. Embryologists have been familiar with chain reactions for cen- turies; the significance of progressive dif- ferentiation in ontogeny had been made ex- plicit in the modern sense by Leuckart and Bergmann, and through His, who though his “organbildende Keimbezirke” are usually accredited for heralding neo-preformation, yet inferred the causal relationships implicit in neo-epigenesis (1874, p. 2): Die Entwicklungsgeschichte ist ihrem Wesen nach eine physiologische Wissenschaft, sie hat den Aufbau jeder einzelnen Form aus dem Ei nach den verschiedenen Phasen nicht allein zu beschreiben, sondern derart abzuleiten, dass jede Entwicklungs- stufe mit allen ihren Besonderheiten als noth- wendige Folge der unmittelbar vorangegangenen erscheint. Roux himself, who was a_ student of Haeckel’s, and felt himself his disciple, very definitely acted as intermediator between Darwinism and the new causal analytical embryology. His “Kampf der Theile im Or- ganismen,” though written before his great concentration of interest on embryological problems, was consummately important in this respect. By striking an analogy between the struggle for existence among organisms on the one hand, and that between the parts of an organism on the other, Roux pointed up for the first time in a new way the pos- sible significance for differentiation of inter- relationships between tissues, and suggested the possibility already in that communica- tion of the “Hervorbildung des chemisch und morphologisch Differenzirteren aus dem Einfacheren ohne differenzirende Aussere Einwirkungen,” as opposed to the conditions where “andere Gewebe . . . secundar durch Einwirkung seitens der ersteren aus dem embryonalen Blastem differenzirt werden” (1881; cited from Roux, 1895a, I, 332-333). But are not these clearly Haeckel’s innere and aussere Bildungstriebe—words and con- cepts borrowed by him deliberately from Goethe? And Haeckel, for all his faults, was infected by Goethe with an enthusiasm for the dynamic wholeness of the organism and its environment, which he passed on to Roux who made use of it as Haeckel never could. Spemann was to be moved by it too. He mentioned in his autobiography having known Martin Donndorf who had seen Ecker- mann, and commented that “man _ wird nicht mehr leicht jemand begegnen, der in Augen geblickt hat, welche Goethe gesehen haben” ([’43], p. 86). His own intellectual distance from Goethe was, like Roux’, dimin- ished by the intermediacy of Haeckel: Im Lager fiel mir das Buch von Wilhelm Preyer liber die Seele des Kindes in die Hand; mit schlech- tem Gewissen, wie ein Schuljunge mit einem Buch unter der Bank, sass ich damit in einer dunklen Ecke. Das kam aus der Gegend von Ernst Haeckel, der so manchen jungen Mann meiner und der vor- hergehenden Generation zur Biologie gefiihrt hatte. Dort begegnete ich auch zum erstenmal. soviel ich mich erinnere. dem Begriff der Biologie als einer umfassenden Wissenschaft vom Leben, mit all ihren aufwiihlenden Lehren tiber seine letzte Tiefe (zbid., p. 116). But if Roux began his work against the natural philosophical background of Haeckel, from whom he also inherited his predilection for setting up his concepts in a strongly theoretical framework, he was later to grow far beyond Haeckel’s romanticism. Haeckel’s own strong predilection for monism may have exerted its influence in this respect. But Roux had studied also with Goette. Particu- PROBLEMS, CONCEPTS AND THEIR HISTORY 19 larly, also, he was influenced in the direction of the new preformationism by Weismann. For the new embryology, strangely enough, received a strong impetus from the old pre- formation as this was revived in a new form. Radl (730, p. 263) has made the clever com- ment, and justly so, that “Haeckel silenced von Baer, the embryologist, and returned to the ideas of Meckel. Weismann ignored [von] Baer, the epigeneticist, and went back to the idea of the tedious and insipid Bonnet.” Roux derived his parentage from them both; and Weismann himself had been, after all, a stu- dent of Leuckart’s. It is always tempting to contrast Roux’ mechanistic preformationist tendencies with the vitalistic and epigenetic interpretations of Driesch, and to assume the enthusiasm for the new experimental embryology as arising out of the clear-cut difference between them, and as demanding the collection of new evi- dence to justify the choice between them. But what was more influential in starting the new experimental embryology on its way was not the opposition of results and inter- pretations of Roux and Driesch so much as the desire to verify or refute the Roux- Weismann hypothesis that the qualitative distribution of nuclear material is responsi- ble for the mosaic sort of differentiation which Roux thought he could demonstrate; and the attack begun on it by Driesch and Hertwig in 1892 not only represented the beginning of all the active experimentation to follow, but led directly to the constriction experiments of Spemann which were destined to have such momentous results. In the same measure as Weismann was in- directly responsible for the new embryology to follow, nineteenth century materialism lay behind it. Weismann was no freer than any other from the influence of his times; his idioplasm is a modification of that of Nageli which was “like a microscopic picture of the macroscopic individual” (cited from Radl, 30, p. 226). Atomicity, as elucidated by Dalton for chemistry, was implicit in the biological ideas of Mendel and Pasteur, and equally in the determinants postulated by Weismann. Weismann’s contribution, like that of so many influential figures in the history of science, was the expression of his doctrine to a century philosophically ripe for its acceptance. And once more a new tool could be ex- ploited to make visibly manifest concrete evidence of a theory bound to become popu- lar because of its appropriateness to the de- mands of its times. Now, it was the improved achromatic lens, which brought out the cyto- logical details of nuclear behavior during mitosis and meiosis and fertilization, and which gave new meaning to the relationship between cellular inclusions and cellular differentiation, and thus evoked a new in- terest in the old preformation. This had its effect not only on Roux among the em- bryologists. In a way, the work of Whitman and all his disciples who devoted themselves to the study of cell lineage in eggs with determinate cleavage grew out of the same background as Weismann’s, and came in a different way to setting up the premises for a new kind of preformation (see especially Whitman, 1894a, 1894b, and Wheeler, 1898, for brilliant expositions of the fin de siécle position against the older historical back- ground). It led too to the magnificent work of Boveri, who on a purely embryological basis established as soundly as the geneticists later the role of qualitative and quantita- tive distribution of the chromosomes. Weismann, however, and the new mate- rialism and the new preformationism, and Goette and His and Haeckel were not the only influences to culminate in Roux. He had his ideas from the botanists, too, and not only indirectly as in the case of Nageli’s idioplasm adopted by Weismann. Sachs, who was also greatly influenced by Nageli, had been experimentalizing plant biology as Roux was to do with embryology; with the physi- ologists’ primary interest in irritability he worked out the basis of what was later to culminate with Loeb in the theory of trop- isms, a concept which the embryologists adopted as soon as the physiologists, and perhaps with even greater fruit. Roux, in the first paper, after the introduction, published in the new Archiv (1895d), took it over in his use of “cytotaxis” and “‘cytotropism” as ex- planatory of the relationships of amphibian blastomeres to one another. And Roux was not the only one to employ it: Driesch called in “taktische Reizbarkeit” to explain certain behavior of the mesenchyme cells in the echinoderm embryo. Herbst elaborated it further for embryology, even going so far, in his Formative Reize in der tierischen Onto- genese, as to postulate on a theoretical basis the dependence of the development of the vertebrate lens on the optic cup, the verv year that Spemann (who had himself learned botany from Sachs) was to perform his first experiment demonstrating it (thouch inde- pendent of Spemann; cf. Spemann, ’03, p 566). The ideas of progressive differentiation which had been developing since Aristotle 20 PROBLEMS, CONCEPTS AND THEIR HISTORY were thus to come to final fruition; Spe- mann’s precision of thought and perform- ance, and his supreme intellectual power, instigated their analysis in a new way, doing more than full justice to the causal analytical motive of Roux. For if in Roux many influences converge and if from him many new trends begin, and if in embryology after his time as directly before it, it is not always easy to follow a single guiding motive, yet there stems from him the single modern approach, the ez- periment, and this we owe to him alone. It is true, as so often, that before he was to perform the experiment which was to start the new trend, there had been a gradual preparation for its acceptance on the con- ceptual side: biological science, like physical, had finally become generally experimental- ized; it was no accident that Pasteur and Claude Bernard preceded Wilhelm Roux in time. But Chun had previously performed a similar experiment with as striking results if not as momentous influence; and Roux not only performed the experiment but general- ized its significance. His ability to mobilize thought around it was due to his own quali- ties of mind and person. Driesch, in a different way than Chun, represents the contrast against which the contribution of Roux becomes more capable of evaluation. Roux had the perspicacity to appreciate that the embryo could be grappled with experimentally; Driesch, though he made a great experimental contribution to embryology, lacked it, and was so steeped in metaphysics that he finally made his option for philosophy proper. The comparison of their interpretations of what they thought a single comparable ex- periment illustrates the strength and weak- ness of both. Roux believed he freed one blastomere in the two-celled amphibian egg from the influence of another by killing the latter, and thought he demonstrated thereby the independent differentiation of the sur- viving cell. Driesch found a single isolated blastomere of the two-cell echinoderm egg able to form a whole embryo and believed he could prove the differentiation of a cell to be dependent on its position with respect to the whole. Both experiments were subject to critical errors, as we now know, and for both eggs both explanations are partially correct and partially inadequate. But this was not the only issue between them, nor the fundamental one. Roux inter- preted the egg for the first time as a mech- anism mechanically analyzable by outside interference; Driesch envisioned it as ruled by an entelechy as spiritual as any deus ex machina must be. This difference was within them, and not dictated by their times. Though the spirit of particular times may facilitate choosing one view or the other, there have been mechanists and vitalists at all ages. But those differences in Roux’ and Driesch’s interpretations were here deter- mined by their own casts of mind, and the fact is that Roux, by his choice, brought the embryo to become experimentally attackable by exact investigation. Roux set the whole program for experimental embryology, and this is his importance, not the fact that he performed an experiment which by 1910 had been proved to be erroneously conceived and interpreted. Roux’ importance, however, is not only in terms of presenting a method of solving problems, but of setting them, and this is pertinent. He could perform an experiment on an embryo because he could ask a ques- tion of the embryo that was experimentally answerable, at least within limits. His oppo- sition of differentiatio sui and differentiatio ex alio answers the warning raised by Brooks (cf. rubric heading this section) that points of reference must be specified. The choice between them could be made in a limited way by his own method of isolation; the necessity of a more crucial method to certify the choice led to the development of the transplantation technique by Born, Harrison and Spemann. And Roux’ experimental pro- gram, carried to its logical outcome by the addition of the transplantation methods, in a way implied by it, has led straight to the modern embryology which incorporates the valid features of both epigenetic and pre- formationistic concepts. Every embryologist, whether concerned with the development of enzyme systems, or the cleavage of so-called determinative eggs, or with fields and gradi- ents in a regulative egg, or whatever, is still concerning himself with the degree to which his material at a particular moment in de- velopment is answering to Roux’ description of differentiatio sui or differentiatio ex alio. Roux’ own primary interest was in a the- oretical and philosophical problem, that he called the Causalnexus of events. But his gift was that he could address his problem in such a way to the embryo that the embryo reacted in what Roux believed an intelligible wav to laboratory interference. The degree of intelligibility of the reac- PROBLEMS, CONCEPTS AND THEIR HISTORY 21 tion of the embryo to interference is a prob- lem to which we shall return in the next section. But it may be commented here that Roux, like von Baer, has made a greater ad- vance than Darwin in the descent from metaphysics to the level of the organism; as a result of Roux’ program, after all, hy- potheses can be tested by exact investigation —and the investigator need only to read any single paper by Spemann or Harrison to know how exact—in an experimental labora- tory. * * * * * * * This section was started, it may be re- membered, from the premise that a_ back- ward look might give a clue as to how to proceed in the future for the best progress of embryology. The primary trend that emerges from the survey has seemed the grad- ual transition from the metaphysical to the physical that characterizes the progress of all developing science. What help may we de- rive, then, from the embryologists who have made the greatest progress in this respect? The one valid generalization that can be drawn is that the great progressive minds of embryology—those of Aristotle, Wolff, von Baer, and Roux—and in our own times those of Spemann and Harrison—have been those of the investigators who have learned to ad- dress the embryo by the right question; and these are the men who have derived their intuitions primarily from the study of the embryo itself. The investigators who have derived their ideas from the philosophical side, and examined their embryos to fit their observations into philosophical patterns al- ready set and rigid—Goethe, St. Hilaire, Haeckel, Driesch—were the minds whose philosophical patterns delayed rather than accelerated the course of embryological progress. Aristotle and Wolff and von Baer and Roux started out too from philosophical and theoretical premises, but in such a way that they relegated the initiative of answer- ing their problems to the embryo itself; they could do so only because it was the embryo that gave them their clues as to how to ask their questions. The others had been more interested in the ideas than the embryos, and had become captured by them to the detri- ment both of themselves and science. Examining the problem from another as- pect, we may say that the greatest delaying influences on embryology have been first the acceptance of the seventeenth century preformation doctrine, then the doctrines of Unity of Type, later the recapitulation doc- trine, all concepts whose philosophical rather than their embryological content insured their success. Are we in the course of under- going a similar delay? Are we too indulging in too high a degree of metaphysical specu- lation, pushing back what we cannot under- stand into concepts of fields and gradients which still have only metaphysical reality and into invisible realms what visible struc- ture only inadequately explains? Is our pres- ent emphasis on the biochemical and _ bio- physical constitution of the embryo a reac- tion against this? It is now again beginning to be admitted (cf. many authors in Parpart, ed., °49) that the new biochemistry is insuf- ficient to answer our fundamental problem of organization. We can not safely under- estimate the complexity of the problem we attempt to solve, to borrow phraseology from Wheeler (1898). In biological science, struc- ture is inadequate to explain process. As biol- ogists, we are bound to fail when we use methods applicable only to the study of structure for the analysis of processes func- tioning in time. The problem of modern em- bryology as stated above was crudely summed up as the problem of determining the degree to which particular material an- swers to Roux’ description of differentiatio sui or differentiatio ex alio at any one mo- ment. We have no methods as yet to deal with analyzing the transition from one mo- ment to the next. Here the problem is sim- pler for the classical physiologist, and it is because of this that the progress of the em- bryologist has followed rather than preceded his. Another trouble has been, historically speaking, our constant opposition of the metaphysical to the physical; there may be a biological level, too, at which one might work without retreating to the camp of the spiritualists and vitalists, and this is where our imagination has been and still is at its weakest. Roux saw the dilemma, as have so many others (1895c, p. 23): Fiir den Forscher auf dem Gebiete der Entwicke- lungsmechanik gilt in hohem Masse das Wort: “Tncidit in scyllam, qui vult vitare charybdim.” Die zu einfach mechanische und die metaphysische Auffassung reprasentiren die Scylla und die Cha- rybdis, zwischen welchen dahin zu segeln in der That schwer und bis jetzt nur Wenigen gelungen ist. Some of the limitations of the too simply 22 PROBLEMS, CONCEPTS AND THEIR HISTORY mechanical conception will be entered into further in the section to follow; suffice it to conclude here with a few more words about the metaphysical. Modern science considers respect for it a fault, which in many ways it is. But modern science must remember, as Whitehead reminds us, that all thought is abstract, and that intellectual induction at least presupposes metaphysics. And primary success in dealing with the embryo, because of its complexity, must derive from the in- ductive rather than the deductive process. No scientific progress has ever been made with- out reflection and speculation; and it must be remembered that both imply holding a mirror to nature, and that the surfaces must be held true. Woodger (48), as a matter of fact, in a most interesting theoretical paper, ‘Obser- vations on the present state of embryology,” presented recently at the Second Growth Symposium in England, has concerned him- self with the necessary conditions for imme- diate embryological progress, to conclude that what we need is to concentrate our attention on a few key data in order to de- rive the key hypotheses we require to pro- ceed. There is a fallacy here for embryology. The greatest progressive minds of embryol- ogy have not searched for hypotheses; they have looked at embryos. How they have looked, and how they are looking now, is the burden of the section to follow. REFERENCES Agassiz, L. 1857 Essay on classification; in Con- tributions to the Natural History of the United States of America, Vol. I, pp. 1-232. Little, Brown and Company, Boston. Aristotle 1943 Generation of Animals. With an English translation by A. L. Peck. Harvard Uni- versity Press, Cambridge, Massachusetts. 1945 Parts of Animals. With an English translation by A. L. Peck and a foreword by F. H. A. Marshall. Movement of Animals. Pro- gression of Animals. With an English translation by E. S. Forster. Harvard University Press, Cam- bridge, Massachusetts. Baer, K. E. von 1828, 1837 Ueber Entwicke- lungsgeschichte der Thiere: Beobachtung und Reflexion. Gebriider Borntrager, K6nigsberg. 1853 Fragments relating to philosophical zoology. Selected from the works of K. E. von Baer; In Scientific Memoirs, Selected from the Transactions of Foreign Academies of Science, and from Foreign Journals: Natural History, edited by A. Henfrey and T. H. Huxley, pp. 176- 238. Taylor and Francis, London. 1864 Reden gehalten in wissenschaft- lichen Versammlungen und kleinere Aufsatze vermischten Inhalts, Part I: Reden. H. Schmitz- dorff, St. Petersburg. Balfour, F.M. 1880,1881 A Treatise on Compar- ative Embryology. Two vols. Macmillan and Co., London. Balss, H. 1936 Die Zeugungslehre und Embry- ologie in der Antike. Quellen u. Studien zur Ge- schichte der Naturw. u. der Med., 5:193-274. Beer, G. R. de 1951 Embryos and Ancestors. Clarendon Press, Oxford, England. Bergmann, C. and Leuckart, R. 1855 Anatom- isch-physiologische Uebersicht des Thierreichs. Vergleichende Anatomie und Physiologie. Em Lehrbuch fiir den Unterricht und zum Selbst- studium. New ed. J. B. Miller, Stuttgart. Brooks, W. K. 1902 The intellectual conditions for the science of embryology. Science, N.S., 15: 444454, 481-492. Bunge, G. 1902 Textbook of Physiological and Pathological Chemistry. 2d English ed. translated from the 4th German ed. by F. A. Starling and edited by E. H. Starling. P. Blakiston’s Son & Co., Philadelphia. Burnet, J. 1930 Early Greek Philosophy. 4th ed. A. & C. Black, Ltd., London. Cole, F. J. 1930 Early Theories of Sexual Differ- entiation. Clarendon Press, Oxford, England. 1944 Dr. William Croone on generation; in Studies and Essays in the History of Science and Learning Offered in Homage to George Sar- ton on the Occasion of his Sixtieth Birthday 31 August 1944, edited by M. F. Ashley Montagu, pp. 113-135. Henry Schuman, New York. Darwin, C. 1902 The Origin of Species by Means of Natural Selection or the Preservation of Fa- vored Races in the Struggle for Life. With addi- tions and corrections from 6th and last English edition. Two vols. in one. D. Appleton and Com- pany, New York. Diderot, D. 1875. Entretien entre D’Alembert et Diderot. Réve de D’Alembert. Suite de l’entre- tien; in Oeuvres completes de Diderot, edited by J. Assézat, Vol. II, pp. 101-191. Garnier, Paris. Eckermann, J. P. 1905 Gesprache mit Goethe in den letzten Jahren seines Lebens, edited by A. Bartels. Two vols. Eugen Diederichs, Jena. Fabricius, H. 1942 The Embryological Treatises of Hieronymus Fabricius of Aquapendente. The Formation of the Egg and of the Chick (De forma- tione ovi et pulli). The Formed Fetus (De formato foetu). A facsimile edition, with an introduction, a translation, and a commentary by H. B. Adel- mann. Cornell University Press, Ithaca, New York. Goethe, J. W. von 1944 Gedanken und Aufsatze; in Goethes Werke, edited by E. Merian-Genast, Vol. XII. Birkhauser, Basel. Haeckel, E. 1891 Anthropogenie oder Entwicke- lungsgeschichte des Menschen. Keimes- und Stammes-Geschichte. 4th rev. and enl. ed. Wil- helm Engelmann, Leipzig. [1900] The Riddle of the Universe at the Close of the Nineteenth Century. Translated PROBLEMS, CONCEPTS AND THEIR HISTORY 23 by J. McCabe. Harper & Brothers, New York. Harvey, W. 1653 Anatomical Exercitations, Concerning the Generation of Living Creatures: To Which Are Added Particular Discourses, of Births, and of Conceptions, &c. Octavian Pulleyn, London. 1931 Exercitatio anatomica de motu cordis et sanguinis in animalibus. An English translation with annotations by C. D. Leake. Charles C Thomas, Springfield, Illinois. His, W. 1874 Unsere Kérperform und das physi- ologische Problem ihrer Entstehung. Briefe an einen befreundeten Naturforscher. F. C. W. Vogel, Leipzig. Huxley, T.H. 1849 On the anatomy and the af- finities of the family of the Medusae. Philos. Trans. Roy. Soc. London, 739:413-434. 1853a On the morphology of the Cepha- lous Mollusca, as illustrated by the anatomy of certain Heteropoda and Pteropoda collected dur- ing the voyage of H. M. S. “Rattlesnake” in 1846-50. Philos. Trans. Roy. Soc. London, 743: 29-65. 1853b The cell-theory (Review). British & Foreign Medico-Chirurgical Review, 72:285- 314, Jung,C.G. 1927 Psychology of the Unconscious. A Study of the Transformations and Symbolisms of the Libido. A Contribution to the History of the Evolution of Thought. Authorized transla- tion, with introduction, by B. M. Hinkle. Dodd, Mead and Company, New York. Kleinenberg, N. 1886 Die Entstehung des Anne- lids aus der Larve von Lopadorhyncus. Nebst Bemerkungen iiber die Entwicklung anderer Polychaeten. Wilhelm Engelmann, Leipzig. Kowalewski, A. 1867 Die Entwickelungs- geschichte des Amphioxus lanceolatus. Mém. de l’Acad. de St. Pétersbourg, 7th series, 77, No. 4, pp. 1-17 (page numbers recorded from offprint; original journal not available). Maitre-Jan, A. 1722 Observations sur la forma- tion du poulet, ot! les divers changemens qui ar- rivent a l’oeuf 4 mesure qu’il est couvé, sont ex- actement expliqués & representés en figures. d’Houry, Paris. Malpighi, M. 1685 Dissertatio epistolica de form- atione pulli in ovo Regiae Societati Londoni ad scientiam naturalem promovendam institutae, dictata; in Bibliotheca anatomica sive recens in anatomia inventorum thesaurus completissimus ... edited by D. LeClerc and I. I. Magnetus, pp. 575-594. Chovet, Geneva. Meckel, J. F. 1821-1833 System der vergleichen- den Anatomie. 6 pts. in 7 vols. Renger, Halle. Meyer, A. W. 1935 Some historical aspects of the recapitulation idea. Quart. Rev. Biol., 10:379- 396. 1939 The Rise of Embryology. Stanford University Press, Stanford, California. Needham, J. 1934 A History of Embryology. Cambridge University Press, Cambridge, Eng- land. Oppenheimer, J. M. 1936 Historical introduction to the study of teleostean development. Osiris, 2: 124-148. 1940 The non-specificity of the germ- layers. Quart. Rev. Biol., 75:1-27. Parpart, A. K. (editor) 1949 The Chemistry and Physiology of Growth. Princeton University Press, Princeton, New Jersey. Plato [1944] The Timaeus and the Critias or Atlanticus. The Thomas Taylor translation. Pantheon books [New York]. Radl, E. 1930 The History of Biological The- ories. Translated and adapted from the German by E. J. Hatfield. Oxford University Press, Ox- ford, England. Roux, W. 1895a Der ziichtende Kampf der Theile oder die ‘“Theilauslese” im Organismus. Zugleich eine Theorie der “functionellen An- passung.” Ein Beitrag zur Vervollstandigung der Lehre von der mechanischen Entstehung des sogenannten “Zweckmassigen”; in Gesammelte Abhandlungen iiber Entwickelungsmechanik der Organismen, Vol. I, pp. 135-422. Wilhelm Engel- mann, Leipzig. 1895b Die Entwicklungsmechanik der Organismen, eine anatomische Wissenschaft der Zukunft; in Gesammelte Abhandlungen iiber Entwickelungsmechanik der Organismen, Vol. II, pp. 24-54. Wilhelm Engelmann, Leipzig. 1895c_ Ejinleitung. Roux’ Arch. Entw.- mech., 7:1-42. 1895d Ueber den “Cytotropismus” der Furchungszellen des Grasfrosches (Rana fusca). Roux’ Arch. Entw.-mech., 7:43-68, 161-202. Schrecker, P. 1938 Malebranche et le préform- isme biologique. Revue internat. de philos., 7:77- 97. Shumway, W. 1932 The recapitulation theory. Quart. Rev. Biol., 7:93-99. Singer, C. 1922 Greek Biology & Greek Med- icine, Clarendon Press, Oxford, England. [1941] A Short History of Science to the Nineteenth Century. Clarendon Press, Oxford, England. 1944 A word on the philosophical back- ground of Vesalius; in Studies and Essays in the History of Science and Learning Offered in Homage to George Sarton on the Occasion of His Sixtieth Birthday 31 August 1944, edited by M. F. Ashley Montagu, pp. 75-84. Henry Schuman, New York. Spemann, H. 1903 Entwickelungsphysiologische Studien am Triton-Ei. III. Roux’ Arch. Entw.- mech., 16:551-631. [1943] Forschung und Leben. Edited by F, W. Spemann. J. Engelhorns Nachf. Adolf Spe- mann, Stuttgart. Temkin, O. 1950 German concepts of ontogeny and history around 1800. Bull. Hist. Med., 24: 297-246. Thompson, D’A. W. 1940 Science and the Clas- sics. Oxford University Press, Oxford, England. Wheeler, W. M. 1898 Caspar Friedrich Wolff and the Theoria generationis. Biol. Lect. Marine Biol. Lab. Woods Holl, 1898, pp. 265-284. 24 PROBLEMS, CONCEPTS AND THEIR HISTORY Whitehead, A. N. 1925 Science and the Modern Wolff, C.F. 1774 Theoria generationis. New ed., World. The Macmillan Company, New York. enl. & corr. Hendel, Halle. Whitman, C.O. 1894a Evolution and epigenesis. Woodger, J.H. 1948 Observations on the present Biol. Lect. Marine Biol. Lab. Woods Holl, 1894, state of embryology; in Growth in Relation to pp. 205-224. Differentiation and Morphogenesis, Symposia of 1894b Bonnet’s theory of evolution. Biol. the Society for Experimental Biology, No. 2, pp. Lect. Marine Biol. Lab. Woods Holl, 1894, pp. 351-365. Academic Press, New York. 225-240. Section II METHODS AND TECHNIQUES JANE M. OPPENHEIMER INTRODUCTION: SOME GENERAL CONSIDERATIONS GREAT ADVANCES, in scientific history, have almost always depended more on intellectual than on technological innovations. No new technique has alone either answered any problem of primary importance or has itself set one. In biological history, specifically, so far as technique is concerned, the work of Harvey and of Darwin and of Mendel could have been performed far earlier than it was: their advances were on the intellectual side, rather than technical in any sense. Harvey’s greatest contribution was perhaps the appli- cation of the principle of measurement to biological material, and his ability to per- form his experiment was an inevitable out- come of his quantitative considerations. There was a certain greater ease for him, who was bred in the halls of Padua a generation after Galileo, than for his predecessors to think quantitatively. But there is no a priori rea- son why such thinking might not have emanated from some Greek mind near Archi- medes who in considering specific gravity in physical terms was thinking as quantitatively as Harvey who had only to collect and count a few cups of blood. Darwin was led to the formulation of his doctrine by the considera- tion of Malthus’ economic principles and Lyell’s geological ones: but might not a clue to the genetic relationship of man and the ape have come from a comparison of their faces? Indeed had it not already done so in Buffon’s concept of the ape as a “degraded” man? And what did Mendel do but sepa- rate his generations, keep careful records, count accurately and think clearly? The ma- terial of these men was the organism pure and simple; the instruments with which they attacked it were primarily their ideas. To quote Woodger, who emphasizes the same point: Neither Dalton nor Mendel were afraid to put forward their hypotheses because of the absence of 23 physical apparatus like that provided by X-ray pho- tography. Their hypotheses were devised to explain the generalizations of their day—chemical general- izations about combining proportions in Dalton’s case, and generalizations about ratios of kinds of off- spring of known parentage in Mendel’s case. The apparatus which subsequently provided confirma- tion of these hypotheses might never have been in- vented (at least im the case of Dalton) if the hy- potheses themselves had not first been invented (48, p. 360). In embryology, too, as we have seen, the concept developed before the technique to verify or refute it. It was changes in think- ing, not the mechanical tool, that permitted Wolff and von Baer to see more than Mal- pighi, in seeing less. Thinking in terms of concrete units such as Dalton’s atoms, Men- del’s unit characters in heredity and Pasteur’s germs preceded the discovery and observation of discrete particles in the form of chromo- somes and cytoplasmic inclusions by the use of improved achromatic lenses in the mas- terly cytological studies at the turn into the present century that were to become so im- portant for embryology. Boveri (’07) estab- lished the fact of qualitative difference of the chromosomes by an intellectual tour de force in his analysis of dispermic echinoderm eggs before the technical methods of genetics were available. Spemann reached his primary premises out of thoughtful consideration of methodologically simple constriction experi- ments; he was later only to test and confirm, by the application of the technically more involved transplantation methods of Born and Harrison, what he had already suspected. Indeed, one of the more curious phenom- ena of embryological history is the great lag in the application of more general bio- logical techniques to the particular problems of embryology. Vesalius’ Fabrica was pub- lished over 75 years before the treatise of Fabricius on the developing chick. Strong magnifying lenses were used fruitfully for a half century on other biological material be- fore Malpighi used the compound micro- 26 scope to examine the blastoderm of the chick. Alchemy and pharmacology had puzzled over the uses of specific salts, at least since the ninth century, for the adult organism; the modern statement of the fundamental chemical problems of embryology awaited the nineteenth century. If the embryo waited centuries for even such simple quantitative approach as had been devoted to the adult by Sanctorius Sanctorius, it had to wait millen- nia, from the time surgeons first used their scalpels on biological material to cut off an offending member, for the hot needle of Roux and the discerning eye of Chun (1880) who observed comparable effects of the stormy Mediterranean seas. John Hunter had at- tempted modern methods of grafting in the adult organism over a century before trans- plantation techniques were applied to em- bryological material. There is surely no simple reason to account for these long de- lays in embryological evolution. But cer- tainly they may be related to the fact that no method or technique developed for other sciences, even within the biological realm, has been adequate to enable the embryologist to come to grips with his fundamental and most inescapable problem, the nature of em- bryonic organization. The important progress, then, in the his- tory of embryology, has been in the gradual changes in the cast of thought and clarifica- tion, as it seems from our point of view, in the setting of the question to be answered by the embryo. How the question is ex- pressed, at any one moment in history, is of course conditioned by the technical proce- dures available at the time, as well as by the influence of more general currents of thought; and the problem of the embryolo- gist becomes the problem of asking a ques- tion, with whatever means are at his dis- posal, that the embryo can answer in a manner intelligible to the investigator. What are the means of investigation avail- able to the embryologist today? How have they developed? To what degree do they per- mit adequate reply to the problems they purport to attack? What are their limitations and how far can these be overcome? In what measure do they inhibit, as did the late nineteenth century concentration on genea- logical research, or in what way do they en- courage, as did the happy exploitation of the transplantation method by Spemann and Harrison, the posing of new and searching problems? What does the experience of the past and the present inform us at all usefully as to how the future might best be explored METHODS AND TECHNIQUES in terms of new techniques and of new prob- lems? These are questions which the em- bryologist must answer if he is to review his work in proper perspective with relation to larger fields, and only by so doing can he hope to facilitate his approach to the prob- lems next facing him. OBSERVATION VS. INTERFERENCE AS AN APPROACH TO EMBRYOLOGICAL PROBLEMS Modern embryology, since Roux, has tended strongly both in its pedagogical and investigational aspects to contrast the de- scriptive, or morphological, or observational, approach, with the so-called experimental, an only apparent distinction whose illusion of dichotomy leads to an important paradox to be taken up below. But since the observa- tional method, at least in a crude form, has always been available to the investigator, a few of the difficulties inherent in the inter- pretation of what seem to be the simplest ob- servations may be pointed out at the begin- ning of this discussion. Since even the results of experiments must be observed in some fashion, and this is only part of the paradox, these difficulties of interpretation are of sig- nificance in a much wider sense and there- fore will be discussed also in relation to the broader issues. In the first place, observation of the em- bryo can rarely, if ever, remain observation pure and simple. This seems a truism; yet there are certain inferences to be drawn from it which are not so completely obvious as it might seem. Perhaps the greatest interference with con- structive and advancing observation derives from the preconceptions already present in the mind of the observer, and there is no need further to labor the point that modern investigators, like the great minds of the past, like Wolff, like Roux, tend to see in terms of what they are looking for. It is the what-he-is-looking-for that is so strongly conditioned by the mechanical tools at the disposal of the embryologist, which both expand and limit what is visible to him. In the early days of the microscope the em- bryologist saw organs, tissues, layers, per- haps cells. With the improvement of the techniques of microtomy and staining and with the perfection of achromatic lenses he could examine parts of the nucleus and what now seem the grosser cytoplasmic inclusions. The technique of modern optics and micros- copy enable him to push his frontiers far beyond the old limits, and new instruments METHODS AND TECHNIQUES will continue to help him in this respect. Many other techniques, biophysical, bio- chemical, immunological, and so forth, allow the identification and description, if not the actual visualization, of components of the embryonic cell at the molecular and even the submolecular level. Roux was frankly skeptical of the possi- bility of reducing embryological problems to the molecular level: Auch wenn wir von den letzten Ursachen ganz absehen, so ist es doch fraglich, ob wir das von Carl Ernst v. Baer gesteckte Ziel: ““Die bildenden Krafte des thierischen K6rpers auf die allgemeinen Krafte oder Lebensrichtungen des Weltganzen zuriickzu- fiihren,” je erreichen werden, vorausgesetzt, dass die zu Grunde liegende Auffassung iiberhaupt vollkommen richtig ist (1889, cited from Ges. Abh. 2:28-29). He preferred to approach them by investi- gating, by means of interference with the embryo, what he called the Causalnexus of events. Causality is more suspect to the mod- ern scientist, and becomes mere statistical probability, but following Roux, neverthe- less, the tendency is fortunately to describe not simple structure but events, and specific- ally a sequence of events in time. If Roux believed this sequence capable of being sub- ject to causal analysis, he realized also that the results of any such analysis could become significant only in a frame of reference de- fined by the normal developing embryo, and must have taken for granted that the “con- trol” for an experimentally treated embryo must be an undisturbed one. If a primary obligation of the embryolo- gist, before he can evaluate an experimental result, is the knowing of the normal condi- tion of the embryo, the fundamental paradox arises that he cannot perhaps adequately know the normal without the benefit of the experiment, whose whole raison d’étre can surely be only that it elucidates the normal; yet he cannot interpret his experimental re- sult without comparison with the normal control. Leaving aside, however, for the mo- ment this dilemma, the task of the embryolo- gist concerned with “normal” development becomes the task of describing the sequence of events in time as accurately as possible in terms of what he can see with the tools avail- able to him, and to describe the components which are acting and their manner of action in as precise physical and chemical terms as his instruments and techniques will allow, as objectively as is possible in the light of the general biological and broader philosophical tenets of his times. 27 A first difficulty arises in that it is virtu- ally impossible to observe an embryo under external conditions that do not interfere with it. Fixation, staining and other chem- ical treatment disturb the “normal” condi- tion of the embryonic cell and alter it. Suffice it to say about the living embryo, without entering into the obvious detail, that the laboratory is not its normal environ- ment. Indeed, is it always possible to define, let alone reproduce, a “normal” environ- ment? What is the “normal” environment of a developing Bonellia or Crepidula, whose sex- ual differentiation may be modified by it, in terms of distance from another Bonellia or Crepidula (Herbst, ’36, ’37; Baltzer, ’37; Coe, °48)? What is the “normal” salinity for the development of Artemia whose form de- pends on the salt concentration of its en- vironment (Abonyi, °15)? To come to the vertebrate, which we are in the habit of con- sidering more conventional and stable em- bryological material, Fundulus heteroclitus embryos raised at constant temperatures reach specific morphological stages after dif- ferent periods of time when raised at varying salinities (Merriman, unpublished). The adults live and breed both in brackish water and sea water in nature, and the tempera- tures of the seas and the estuaries are incon- stant. What then is the “normal” environ- ment of the developing Fundulus? What are the criteria for determining objectively which set of conditions is “optimum” for the embryo developing in the laboratory, and what biological significance these have for the organism developing in nature? Comparable problems arise for the chem- ical embryologist who tries to describe the constituent systems of the embryonic cell in biochemical terms. Unpublished data of Dumm have suggested that the cholines- terase level of Fundulus embryos at particu- lar morphological stages varies according to the temperature at which the embryos have been raised; and Boell (unpublished) also has data suggesting that the course of enzyme development in the amphibian can be altered by varying the temperatures at which the embryos are reared. Weiss (’49) has intro- duced in another connection a useful con- cept of “molecular ecology” which may well serve to remind the embryologist of what he already knows but sometimes forgets, namely, that the embryo has an internal as well as an external environment that may well bear more strict definition than it has always hitherto received. 28 For the embryologist who considers him- self primarily concerned with descriptive and morphological questions, the problems arising from these examples, and the count- less others that could have been enumerated, may seem of little moment. He can evade them by ignoring them, or more profitably, he can meet them by specifying as strictly as possible the conditions under which he is making his observations. For the investiga- tor, however, who is looking for “normal” controls for experimental material, the situ- ation becomes more critical. It becomes especially so for the inter- pretation of certain isolation experiments which will be discussed in more detail be- low, and it presents very particular problems in the case of some transplantation experi- ments. The rate of growth of an eye or a limb grafted heteroplastically onto an am- phibian host varies according to whether or not the host is maximally or less than maxi- mally fed (Twitty and Schwind, 731); how often does an Amblystoma in nature enjoy a condition of maximum repletion? The growth of an Amblystoma embryo varies ac- cording to the organisms constituting the diet it is fed; who can define the ‘“nor- mal” diet of Amblystoma in nature? Mem- bers of the same species of Amblystoma reared under similar laboratory conditions develop at different rates of growth when collected in Pennsylvania and New Jersey on the one hand, in Illinois on the other (DuShane and Hutchinson, ’44); which is the more “normal” larva, that collected in the East or that which is spawned in the Midwest? This is a more than philosophical problem when experimentalists in Princeton, for instance, are comparing their results with those of investigators working at the Uni- versity of Chicago in an attempt to work out basic mechanisms of development. And while it is particularly accentuated by the results of the heteroplastic experiments, it touches the heart of each investigation in- volving the growth of the whole or parts of the developing amphibian. In the case of more complicated experi- ments it raises even more complex issues. Haploid tissues developing from the hybrid androgenetic merogons of Hadorn (734, ’37) and Baltzer (°40), for instance, die in the embryos which they constitute but survive in tissue culture or after transplantation to normal diploid hosts of one of the experi- mental species. Is a favorable tissue culture medium or the tissues of a diploid host to be defined as more “normal” or “abnormal” for METHODS AND TECHNIQUES these cells than the haploid parent embryo from which they were derived? Under which conditions are the operations of the gene in development more “normal”? In other cases, what seems “optimum” according to the sub- jective judgment of an investigator con- cerned with a particular experiment may lack biological significance to the organism. Of what biological significance is it to Lyt- echinus that its egg can be so treated in the laboratory that it is more readily activated by sperm of a foreign species than by that of its own (Tennent, ’25)? Tennent (710) has found that in reciprocal crosses of Hip- ponoé XxX Toxopneustes the larvae were of the Hipponoé type when raised in sea water of higher pH, of the Toxopneustes type when reared in sea water of lower pH, and has suggested that seasonal variations in hybrid echinoderm larvae obtained in other labora- tories might be accounted for by seasonal variations in the alkalinity of the seas; who is to say whether autumn or spring is more “normal” for echinoderm hybrids? Under which conditions, and this is the crux of the matter, does the action of the gene, which according to its end-effect varies in the dif- ferent situations, more closely simulate its norm? This is not the appropriate place to take up the developmental action of genes, which will be considered in a separate section be- low; but it is necessary to emphasize here the advantages to the embryologist of work- ing with genetically known and genetically specified material when he is able to do so. Embryos vary. Many of the factors which induce them to vary are difficult to control, as we have seen. The genetic factors are per- haps uncontrollable too when specimens are collected in nature. But when the genetic factors are known, they can be controlled, and thereby great strides into unknown ter- ritory can be accomplished (cf. Gluecksohn- Schoenheimer, ’49), and, as important, the results can be specified in the best biological sense of the word. Consideration, in fact, of all the limitations to interpretation discussed above, and of many others like them, leads to the same in- escapable necessity for specification of the conditions under which an investigation pro- ceeds. It may often accrue to the advantage of the investigation, rather than otherwise, that observations and experiments carried out in different situations lead to different outcomes as well as different interpreta- tions. Only when investigators specify as closely as possible the conditions against METHODS AND TECHNIQUES which particular outcomes eventuate will there become available new data and new ideas through which correlations and signifi- cances which currently still elude us may eventually be discerned. Granted that the conditions, intrinsic and extrinsic to the embryo, under which obser- vations are made are specified as accurately as possible, the question arises as to their ideal manner of description once they have been obtained. For embryology, the prob- lem of semantics which faces all scientists arises in a particularly desperate form, per- haps at least partly because embryology, never having formulated its own problems nor having developed its own techniques, has adopted descriptive words from the lingo of other sciences. Spemann, it may be remem- bered, concluded his great monograph with a confession that he borrowed words to de- scribe embryonic phenomena which point not to physical but to psychical analogies, to emphasize his conviction that these processes of development, like all vital proces- ses, are comparable, in the way they are connected, to nothing we know in such a degree as to those vital processes of which we have the most intimate knowledge, viz., the psychical ones. It was to ex- press my opinion that, even laying aside all philo- sophical conclusions, merely for the interest of exact research, we ought not to miss the chance given to us by our position between the two worlds (’38, p. ByiPD\e Investigators more neutral with respect to this issue use the words induction, determina- tion, regulation, organization, and so forth, borrowed, as Spemann would say, from the psychic sphere, partly out of wonderment at the unexplained powers of regulation of the embryo, but largely for lack of more clearcut or appropriate ones. The problem, however, is not so much of word as of concept; and only when the embryologist can more com- pletely emancipate himself from the domina- tion of other sciences and their techniques, and formulate his problems in his own terms, will he be motivated to create and define such terms with requisite precision. Roux seems to have been the first, and the last, to worry over this problem, sufficiently, at least, to be driven to a specific attempt to solve it. He drew up his Terminologie der Entwicklungsmechanik, a discursive text with only the remotest resemblance to our own unfortunate glossaries, with the express aim ‘‘das causal-analytische Denken [zu] for- dern und auch das vollkommene Verstandnis der Autoren untereinander [zu] erleichtern” (712, p. ix). His recognition of the difficulties 29 inherent in adapting for Entwicklungsme- chanik terms borrowed from other sciences is perhaps nowhere made clearer than by the fact that he found it necessary to include in the J'erminologie two separate definitions in sequence for J'ropismus, one for zoological material composed by himself, the other for botanical, contributed by Kiister, one of the botanical collaborators who assisted in the preparation of the book. No modern attempt to emulate the Terminologie has ever been made; no one since Roux has had either the courage or the conceit to try, and modern embryology is still confronted with the old problem of using borrowed terms. Perhaps, however, it is an advantage to the embryolo- gist to be forced to utilize, as a temporizing device, the terminology of physics and chem- istry, since in doing so he must also use their methods and resources. It is indisputably by the application of these resources that the greatest advances are being made at the present time. The re- sults of current investigations of structure and ultrastructure by phase-contrast micros- copy and cinemicrography and by electron microscopy; of molecular arrays by polar- ization optics; of chemical constitution and activity by histochemical and immunological techniques, by microspectrography and mi- crospectrophotometry; of the localization, constitution and kinetics of enzymes and en- zyme systems and of other metabolic sys- tems, by microrespirometry, by “biochemical dissection” by antimetabolites and other spe- cific poisons, by modern nutrition studies and by the use of both radioactive and stable iso- topes as tracers; of genetic effects of ioniz- ing radiations—all these will be discussed in ensuing chapters of this book. Continuation and elaboration of such phys- ical and chemical descriptions of the em- bryo, of its cells and of their components, are a conditio sine qua non for further embryo- logical progress. Such description, however, dynamic though it may seem, is essentially structural rather than functional, analytical rather than synthetic, and this new mor- phology, like the old, is not able to pene- trate to the core of the problem of organiza- tion. A physical-chemical model of the embryo may ultimately be adequate to rep- resent some phase of what embryonic or- ganization has produced, but as yet there is no assurance that it can reproduce the process by which organization has functioned. Struc- ture, in embryonic material, is not yet ade- quate to “explain” process. This may be bound up with the fact that 30 the embryo has the disadvantage, from the point of view of the investigator, of develop- ing in time. The methods appropriated by embryology from other disciplines, which have a different concern for the time factor, have been invented to describe material at a given moment, not to analyze transition from one moment to the next. Insight into the problems of organization demands new approaches to the physiology of development as such, while as yet we confine our efforts to the descriptive physiology of embryos, which is something different. However, it is possible that the clue to new methods for analyzing organization may well come from a deeper comprehension of structure than we now enjoy; and in any case, any knowledge of process we ever may hope to obtain is certain to become more meaningful in the light of as intimate an understanding as pos- sible of ultimate structure. One evident superiority of the results of the new morphology over those of the old is that they allow quantitative expression. There is no question but that embryological descriptions must be as quantitatively exact as is appropriate for the material and tech- niques in question, and there is no greater need to justify the benefits of this for em- bryology than for any other science. There are, however, certain sources of error in in- terpretation which may be pointed out which are inherent in some attempts to analyze the significance of certain aspects of growth and differentiation by presenting descrip- tions of them in quantitative form. It is well to remember that mathematical abstraction is a particular kind of abstrac- tion which is in itself highly specialized. Any progressive science deals in abstraction as well as in measurement, but it requires always to question the appropriateness of whatever abstraction it utilizes to the partic- ular material with which it deals. There is no more fascinating collection of biological facts than those by which Sir D’Arcy Thomp- son referred growth and form to certain mathematical relationships, but surely Sir D’Arcy’s analogies are provocative rather than explanatory, and he himself hardly claimed more for them. A curve, for instance, which describes the “growth” of a colony of bacteria is useful in that it designates periods of change at particular moments, and these may be periods with which the investigator may wish to concern himself; but the “growth” of a colony of micro-organisms is something dif- ferent from that of a multicellular organ- METHODS AND ‘TECHNIQUES ism, and the growth of one organism may be controlled by different factors than that of another. Weiss’ warning of a few years ago is still relevant and will continue to remain sO: A purely formal treatment of growth, as is often attempted through the interpretation of growth curves, is only a valuable guide to and supplement of, but never a substitute for, a precise analysis of the different forms in which growth manifests it- self. There can be no research on growth as such. We can only study growing objects. And different grow- ing objects follow different methods. .. . To know growth we must therefore break down each one of its manifestations into its constituent elementary processes and then study these and describe them in objective terms. This is a long way to go, but there is no short cut (’49, p. 182). Weiss’ admonition holds equally true for other aspects of development: differentiation, determination, or whatever. Exact quantita- tive description of embryological data is an- other conditio sine qua non for future em- bryological progress, but only if the embryol- ogist keeps in mind which of his problems quantification cannot solve, as well as those which it can elucidate. The success of quantitative methods in creating the new morphology has tended to encourage attempts to adapt quantitative methods to the results of the older; and this condition, together with the fact that the journals currently encourage the publication of data in graphic and tabular form, leads to a growing tendency to make material appear quantitative which may not necessarily be so in its own right. An example is the current procedure of using morphological stage num- bers from stage series to represent ordinates or abscissae of graphs. This may be useful provided the author works with a footnote, in his thinking if not on his page, calling for caution in interpretation, but it is a ques- tion to what extent this reservation is kept in mind. The presence, in the curve of such a graph, of maxima or minima, and whether a line rises or falls is surely significant; not in the same way the slope of the line nor the character of the curve in other respects. A straight line in such a graph is not what it purports to be; “morphological age,’ as Need- ham (742) calls it, is not equivalent to time, which can be quantified. Stage 5 of an em- bryo is not something that equals the sum of stage 2 plus stage 3, which, unless spe- cifically qualified, is what the conventional graphic representation implies. There are innumerable cases, too, where METHODS AND TECHNIQUES presentation of quantitative data is inade- quate and perhaps irrelevant to answer the basic question presented by the material to be analyzed. What is the meaning, for in- stance, of the “quantitative” results described in percentages of positive differentiation in erafts? Luther (’36), for instance, has con- cluded that there is a gradient of physiolog- ical activity (Aktivitdtszustand) around the rim of the trout blastoderm on the basis of the fact that differentiation occurs in a de- creasing percentage of grafts as the material for grafting is removed from progressively ereater distances from the midline of the embryonic shield. If a particular factor, or group of factors, or a certain quantity of such factors necessary for differentiation, char- acterizes the cells near the embryonic shield should not every graft from that area dif- ferentiate if the experiment is adequately performed? What is the meaning, in terms of the functions of the grafted cells, of the fact that only 84% of the grafts removed from a particular region have differentiated? May not the significance of these results be that the grafts have been removed in different ways from zones of transition, or that they have been implanted under differing experi- mental conditions? In other words, may not the quantitative variations in such results indicate variation in the technique of experi- mental procedure as well as variation in the activity of the tissue? Too many factors, which need no enumeration here, are varied in even such a simple experiment as the im- plantation of a graft on the yolk sac, which though in some ways is simple in others is drastic and crude; and the experimental procedure, which is manual and _ therefore difficult to subject to critical control, is prob- ably differently performed each time. It can be of the greatest advantage to the investigator to acknowledge that quantita- tive variation in his results reflects his own uncertainties as well as the accomplishment of his embryonic material, if he wishes to improve his experimental approach both from the technical and the intellectual as- pect. The value of statistical treatment and its advantages in connection with the en- deavor to attain the maximum precision in analysis are particularly great in the case of embryological material where so bewil- deringly many variations are inherent in the material and where so many sources of error confuse the methods of analysis. But statis- tical results must not be interpreted as final to such a degree that they mask the weak- nesses of the technical procedure where these Su actually affect the interpretation of results. Embryology has not yet sufficiently matured towards the perfection of its methods that quantification can be its only desideratum, and it may well be that the necessity to im- prove upon these methods represents the most urgent challenge immediately confront- ing us. TECHNIQUES OF INTERFERENCE WITH THE EMBRYO It was Wilhelm Roux who first had the insight to appreciate the inadequacy of the descriptive method, no matter how precise the terms in which its results are couched, to demonstrate what he called the Causal- nexus of events, and to formulate a program designed to analyze that causal relationship within sequences of events which had al- ready been so clearly expressed on an in- ferential basis by His (see quotation on p. 17, Section I). It was the simplicity of Roux’ first statement of his problem that enabled him to try to answer his question with an apparently simple experiment: Fast alle aber fiihrten im Weiterfolgen zu einer und derselben grossen Vorfrage, zu einer Alterna- tive, von welcher aus die causale Auffassung fast aller Bildungsvorgénge in zwei wesentlich ver- schiedene Bahnen gelenkt wird. Dies ist die Frage: Ist die Entwicklung des ganzen befruchteten Eies resp. einzelner Theile desselben “Selbstdifferenzie- rung” dieser Gebilde resp. Theile oder das Produkt von “Wechselwirkungen mit ihrer Umgebung?” Eventuell, welches ist der Antheil jeder dieser beiden Differenzirungsarten in jeder Entwick- lungsphase des ganzen Eies und seiner einzelnen Theile? In der Beantwortung dieser Frage liegt meiner Ein sicht nach der Schliissel zur causalen Erkenntnis der embryonalen Entwicklung (1885; cited from Ges. Abh. 2:14). The question with which he was con- cerned happened to involve the relationship of the part to the whole, and happened to revive in a new form the old controversy between preformationists and epigeneticists; but this was not its main significance. His great contribution from the methodological point of view was that he saw his problem in terms of a single alternative and in terms of clearcut relationships; relationships so expressed that he could alter them in what was to him a simple experiment. An em- bryo or an embryonic part depends for its capacity to differentiate on a mutual inter- action with its surroundings, or it does not; remove it from its surroundings, and its reply should be unequivocal. a2 The significance, however, of the behavior of a part removed from its surroundings is lost except in comparison with its behavior in those surroundings, as Roux already knew. In most cases, except where natural pig- ments are present, the observation of the egg as a whole sheds all too little light on the separate activities of its individual parts, and Roux himself attempted to circumvent this difficulty by inventing a crude marking experiment and by pricking his eggs to pro- duce extra-ovates which might serve as mark- ers. This experiment is open to the obvious criticism that it may alter the status of the part whose normal behavior it purports to elucidate, and it was to obviate this that the technique of local vital staining was devel- oped. It has reached its highest perfection as developed for application to the amphibian egg by Vogt (’25), where its success depends on the fact that inclusions of the egg adsorb the stain from the carrier more rapidly than it diffuses into the solution, and a mark of the utmost sharpness of outline is therefore achieved and maintained. The data obtained by the local vital stain- ing method are indispensable for the inter- pretation of experiments in which the ac- tivities of particular parts are to be studied by other means, and the dyes currently used (Nile blue sulfate, neutral red and Bismarck brown) are sufficiently nontoxic that the data derived from their use are thoroughly valid. There are conditions, however, under which the method has strong limitations. In some cases the dyes may be transformed to leuko- bases within the cells. In the case of embryos whose cells lack inclusions with special affin- ity for the dyes, for instance young stages of chick and teleost, the stains are far more dif- fuse and ephemeral than in the amphibian, and the results of their use may prove unreli- able when checked against results obtained by other methods. The newer method of follow- ing morphogenetic movements by the appli- cation to the cells of carbon particles (Spratt, 46) has, for instance, produced results for the chick which are incompatible with those previously derived for the same form by vital staining (Pasteels, ’37). The use of car- bon particles holds great promise for the future, but the introduction of macroscopic particles within the cells raises certain dan- gers for the interpretation of what are sup- posed to be unhampered movements; and when the particles are applied to the outer surface of the cell there may always remain some doubt as to whether they may have shifted in position. METHODS AND TECHNIQUES So far as the localization and retention of a marker is concerned, the least reproach- able method of distinguishing one group of cells from another remains the observation of forms in which natural pigments occur. It may be remembered in this connection that the method of heteroplastic grafting was at its earliest inception used to trace the migra- tion of elements distinguished by natural pigment (Harrison, ’03; see Harrison, ’35, for later uses of heteroplastic grafting). Grafting, however, as an operative method, introduces new sources of error not inherent in the methods of marking cells in an unoperated embryo. Unfortunately there is still no equiv- alent of a Geiger counter to report on the migrations of cells in the embryo. The data, however, which have been accumulated by the present techniques as applied by cautious investigators are adequate to serve as a frame of reference against which studies of the parts may be judged. Since killing a cell is probably the easiest thing that an embryologist can do, it was per- haps inevitable that the study of the behavior of parts of an embryo should have first been examined by defect methods, and it is ap- propriate next to mention some of the ways in which defects have been produced. Experi- ments involving the removal or the supposed inactivation of cells or their parts have been carried out by mechanical, chemical, thermal and electrical methods and by combinations of them, and by the use of various sorts of radiations. A primary and insidious source of error common to all these methods is that in ap- plying them the investigator may alter more factors than he knows. The classical experi- ment of Roux (1888) was designed by killing a blastomere to eliminate its influence on its neighbor: the fact that its corpse exerted mechanical influence of moment could be ap- preciated only after McClendon (710) com- pletely removed a blastomere to demonstrate a different accomplishment by the remaining cell than had been achieved in Roux’ experi- ment. Comparable sources of error may lie hidden in many if not all of the defect ex- periments subsequently performed. In dealing with the deletion or inactiva- tion of components of cells, the perils of in- terpretation may be as great. Boveri (18) long ago recognized as an inevitable source of error in experiments designed to exclude nuclear influence that nuclear residues might remain undetected in the cytoplasm, which in any case has necessarily been produced under the influence of the nucleus. The cen- METHODS AND TECHNIQUES trifuge is a tool which can translocate sub- stances from a particular part whose behavior is to be studied in their absence, but who knows what effects it may have had on the invisible components of the zone? Nor are the mechanical methods the only ones open to suspicion. It is unthinkable that in anything kinetically as complex as the simplest protoplasm, alteration of its equili- bria by chemical or thermal changes could be limited in its effects to a single system alone. Temperature changes probably affect all its constituent systems in some measure, and indeed the laws of thermodynamics are hardly such as to permit such effects to re- main local. Methods involving the use of radiations and electricity, in spite of the ad- vantage that the amounts of energy applied may be measured, cannot be construed in most cases as affecting single localizable or identifiable systems within the cell. Excep- tion, however, may be made in the cases where radiations or other agents affect the known gene; the most reliable defect method, perhaps the ideal, is that which excludes identified genes (Poulson, ’40, 45). Manipu- lation of the gene, by man or nature, is probably the most satisfactorily controlled experimental method available to the em- bryologist to date, and it will become the more useful the more exhaustively the inter- mediate steps between the primary action of the gene and the end-effects of its activity, as expressed in differentiation, become known. Despite the reservations enumerated, how- ever, excellent contributions to embryology have been made by the use of all of the de- fect methods. These methods were the first to demonstrate the high degree of regulabil- ity of which the embryo is capable, and thus have led directly to the primary embryolog- ical problem of ultimate organization, and the data which they have provided will be indispensible for the final solution of it. A fact to be kept in mind in this connection is the impossibility of consideration of the defect experiment as separate from the iso- lation or explantation experiment, which is its corollary as well as its complement. In the ideal situation the embryologist wishes to consider both experiments. In one, the in- vestigator studies what remains after some- thing has been taken away; in the other, he studies the behavior of what he has re- moved. In this sense, the study of a single blastomere isolated from a two-celled egg may be regarded either as a defect experi- ment or as an isolation experiment. Roux, as a preformationist, considered it as an isola- 33 tion experiment. He knew that in the egg one blastomere forms half an embryo, and in- terpreted the blastomere in his experiment as duplicating its normal action. He con- cluded that its processes of development were identical in normal and experimental mate- rial, and the experiment seemed to him essentially a way to confirm what he had postulated the behavior of the normal part to be. The questions that are put to the isolated part are now framed by more open minds, but they still deal in the main with the de- gree to which a cell is dependent for its differentiation on factors impinging upon it from its surroundings. To what degree does a part begin or continue differentiation when isolated from its usual cellular surroundings? Does it differentiate the same structures it was destined to form in the normal embryo? If not, what is the direction of its differenti- ation and what factors determine this direc- tion? Attempts to isolate these factors involve not only the negative phase of the experi- ment and determining what factors usually present are lacking when the part is isolated, but also the more positive one of demonstrat- ing the new or different ones to which it is subjected. This was a side of the problem on which Roux did not concentrate, though it is clear that he recognized its importance; and it is this aspect of it which raises some of the most immediate issues facing the pres- ent interpreters of isolation or explantation experiments. Ideally the investigator may express a wish to culture his isolate in a neutral or indifferent medium (cf. Needham, °42, p. 175) which will permit it to continue its own development in its own way. This is essentially what Harrison (’07) did when he isolated the neuroblast in clotted lymph, a medium which did not inhibit the produc- tion of the axon yet which excluded the presence of the cells which had been thought by some to manufacture it. If the behavior of a cell is to be studied in the absence of influence from surrounding cells, it is es- sential that the medium to which it is re- moved cannot itself alter the chain of reac- tions to be studied. To what degree can this ideal be achieved? First and foremost, the emancipation of a cell from influences emanating from its neighbors may now be recognized as more difficult to achieve than formerly was antici- pated, in view of Holtfreter’s (44) recent demonstration that some of the cells constitut- ing the very cultures being studied may at- 34 tain a sublethal state of cytolysis, as a result of reaction to the medium, which may exert hitherto unsuspected effects on other cells nearby in the cultures. Even in cases, however, where such effects may be discounted in the interpretation of results, many other difficulties arise in at- tempts to prepare a suitable medium. Cells can never be independent of mechanical factors in their environment, and indeed are notoriously susceptible to their influence, as was so clearly demonstrated by Harrison (14). In a liquid environment they take quite different shape than when they have access to a solid substrate, and the physical framework of the matrix in which they de- velop is of paramount importance in deter- mining their form (cf. Weiss, 49). Some- times hidden mechanical influences quite prejudice the interpretation of investigations designed to analyze the effects of quite dif- ferent factors; for an example, the reader is referred to Weiss’ (750) critique of Marsh and Beams’ (’46) experiments where develop- ing nerve cells were subjected in vitro to apparent modification by the passage of elec- trical currents. Indeed the demonstration of the degree to which cells are susceptible to influences of external mechanical factors has been one of the most fruitful contributions of the isolation method. From the point of view of nutritive and chemical effects of the milieu on isolates, the analysis is more highly complex. If a part is isolated in inorganic media of easily re- producible composition, many components of the normal environment are lacking which may be essential to foster the processes of normal differentiation; and even such simple factors as a change in pH (though in view of the widely divergent systems in the cell which this might affect its simplicity is only apparent) can alter the accomplishment of cells in such media (Holtfreter, ’45). If the isolate is explanted to parts of another or- ganism, as in implants to the eye cavity, the anterior chamber of the eye, the chorio- allantois, or to the various sites employed for the window techniques, or even to culture fluids containing embryonic extract, plasma or other body fluids, it is impossible in our present state of knowledge to ascertain what components are present. While to some the ultimate aim may seem the perfection of synthetic media—and the embryologists proper lag far behind the tissue culture ex- perts and the microbiologists in their progress towards this goal—it is a little soon to divine what all the ingredients of such media might METHODS AND TECHNIQUES be, since embryologists have hardly yet ex- hausted the knowledge of all the biochemical requirements of their material. A fundamental problem arises as to the criteria by which a neutral or indifferent medium could be recognized as such, granted the validity of the assumption that it exists and granted that it would be capable of preparation. Just as there are various con- ditions under which cells removed from an embryo behave differently than in the nor- mal embryo, so there are various sets of conditions under which cells removed from the embryo might carry out the same per- formance as in the embryo; the most striking manifestations of embryonic organization are those regulatory phenomena whereby processes resembling the normal are carried out under a great variety of abnormal con- ditions. Devillers (’50), for instance, has found the trout blastoderm incapable of differentiation in triple-strength Holtfreter’s solution but able to differentiate in modified White’s solu- tion. It is unthinkable that this is the single solution capable of supporting differentiation in this form. Devillers’ result demonstrates the fact that triple-strength Holtfreter’s solu- tion is unsatisfactory for his particular ex- periment, but provides no essential informa- tion about the blastoderm; the fact that modified White’s solution is more favorable furnishes little information about differen- tiation as such, but signifies primarily that the medium used permits certain embryonic processes to occur. The absence of differentia- tion of cultures in solutions of particular com- position does not necessarily demonstrate that the cells are characterized by the presence or absence of particular potencies, but rather may indicate that the media lack certain factors required as stimuli for the realization of normal potencies, or even that they include agents which may actively inhibit such real- ization. Whether or not the cells will differentiate, furthermore, is not the only test of the suit- ability of the medium; the direction of dif- ferentiation and what the factors are which determine it are as important considerations. Using the prospective nervous system of the young urodele gastrula as an example, when isolated in salt solution it will under some conditions form only simple epidermis, under others nervous tissue (Holtfreter, ’45). Im- planted in vivo, where it is subjected to a wider variety of influences, it is capable of differentiating widely divergent structures (Holtfreter, ’29; Bautzmann, ’29; Kusche, ’29) METHODS AND TECHNIQUES quite other than those it would have formed in the normal embryo, hence Bautzmann’s term bedeutungsfremde Selbstdifferenzierung. The occurrence of these and other examples of bedeutungsfremde Selbstdifferenzierung signifies that the cells have sufficient plas- ticity to differentiate in other than their normal direction as a result of change in the conditions with reference to which they are differentiating, but is not adequate to define the changes which have produced an alteration in the direction of differentiation. It has been postulated that one way, theo- retically, to come closer to a definition of conditions essential to differentiation might be the testing of differentiation capacity in the widest possible variety of media. But if the reactions and directions of differentiation should be studied under as many experimen- tal conditions as possible, how is the infor- mation to be referred to the cells in action in the normal embryo? Observations would be available as to media satisfactory to elicit all ranges of differentiation; but would not the ascertained data remain knowledge of re- actions in an external medium rather than of processes internal to the embryo itself? The comparison of different results in various solutions fails to permit reference of the ex- perimental results to the processes which might have been carried out by the cells in situ. What is the way to prove in the isolation experiment whether the cells which have differentiated particular structures are the same which would have done so in the embryo or whether they have used the same method to reach their end? If, however, the processes studied experi- mentally cannot be referred without caution back to the normal embryo, it is clear that the processes examined experimentally re- quire rigid definition as to the conditions under which they occur. No medium can be neutral or indifferent with respect to early differentiation; if it were actually either of these, isolates could continue no de- velopment at all; a medium can be neutral or indifferent only to a cell that is dead. The failure to come to grips with the prob- lem of relationship between embryo and medium grows partly out of the tendency to emphasize the degree to which cells “self- differentiate.” The concept of self-differenti- ation implies a contradiction in terms; no cell can “‘self”-differentiate, bedeutungsfremd or bedeutungsgemass, insofar as no cell can be separated from its environment. Roux expressed the problem more cogently when he opposed dependent and independent 39 differentiation, a classification which de- mands enumeration of the factors with ref- erence to which differentiation might be de- fined as either of these. One of the strong needs of the moment is more specific defini- tion of the conditions with reference to which both normal and abnormal differen- tiation occur; only this will permit evalua- tion of the meaning of changes in the differ- entiating systems proper, since only in this way can embryologists be certain when they are dealing with these systems themselves and not something extraneous to them. While attempts to invent the medium ideal to support specific types of differentia- tion may suffer in that the results of the studies lack referability to the normal em- bryo, the attempts must still be continued, though as a means to an end. Only when knowledge is available concerning the re- actions of the cells to the media in which they develop can other experiments be inter- preted which are conducted in other ways to answer the more searching questions which remain at the backs of our minds; but the fact must not be lost sight of that these more fundamental questions remain to be asked by other experimental methods. The remaining one of these methods to be discussed is the transplantation method, which allows the framing of different ques- tions than those methods already discussed, or which perhaps rather allows comparable questions to be asked in a slightly different way. Strictly speaking, the implantation of cells to such cellular environments as the chorio- allantois, the eye chamber and the other sites used for studies in vivo might be considered either as transplantation or as explantation experiments; and perhaps some of the best uses to which the transplantation method has been put have been those in which it has been employed as an isolation method, as in the case of Harrison’s (’03) early ex- periments on the lateral line. When cells are isolated in culture in some synthetic medium, the direction of their dif- ferentiation presumably may be influenced by the reaction of cells to factors in the medium; if they are transplanted to a new cellular environment, it will be conditioned in relationship to factors emanating from neighboring cells, and therefore by mutual interactions between cells of graft and cells of host. When cells are transplanted to a cellular environment they are transferred, as in many explantation experiments, to a milieu 36 in which many factors are still unknown, and to this extent the transplantation method shares the limitations of many of the isola- tion methods. In addition, furthermore, it has many of its own incident to the complex- ity of variables introduced by using a living organism as host; the results in grafting ex- periments vary according to the species used, according to the size, age and growth rate of the graft and host, according to the site of implantation and so forth. However, in spite of these, indeed perhaps because of them, the method has advantages peculiar to itself, in that it permits, and in fact leads to, the demonstration of interactions at cellu- lar rather than subcellular levels. It thus en- courages some analysis, at least, of that Wechselwirkung between cell and environ- ment postulated by Roux (cf. quotation on p. 31, this section), and the exploitation of the method by Spemann and Harrison has demonstrated the reality of the progressive quality of differentiation which Roux and His before them had postulated. The significance, after all, of what a cell can do in isolation can reach its full value only in the light of what the cell does in combination with other cells, and recombina- tion therefore by means of grafting is oblig- atory to clarify interpretation of the results of the isolation and deletion experiments. Without physical and chemical description of the cells and their components, and with- out the knowledge of the separate activity of the cells as ascertained by vital staining and defect and isolation experiments, the results of the transplantation experiments themselves might have little meaning. But it is the re- sults of the transplantation experiments which impute final validity to these others, by presenting as a frame of reference not some chance combination of inert substances but the organized living embryo itself. It is when the embryologist attempts to re- fer the phenomena which the transplantation experiments demonstrate as occurring at the cellular level to phenomena with which he is familiar at the subcellular level that he meets his greatest difficulty. But though this problem may seem to present itself more acutely at a time when biochemistry is forg- ing the most rapid advances, it is not in any Way a new one, nor was it new when Roux found himself confronted with his passage between the Scylla of the overphysical and the Charybdis of the overmetaphysical inter- pretation of his results (cf. quotation on p. 21, Section I). Nor is it any new solution to claim that between these two levels a METHODS AND TECHNIQUES biological plane exists, and to recognize that here, where problems of organization are concerned, all biology works to its least satis- faction. Embryology, as a matter of fact, oc- cupies a more favorable position in this re- spect than many other fields of biology, be- cause it is so fortunate in having had a Spemann and a Harrison whose special gen- ius lay in their ability to probe more deeply than investigators in other areas into the forbidden territories. Their method, as a syn- thetic one rather than an analytic, as a method dealing with mutual interactions in terms of cell and cell, rather than simpler re- action of cell to some less organized entity, has the unique merit to come as close to the biological plane of investigation as has yet been approached. For the knowledge to proceed still further into the investigation of these intercellular phenomena and finally into those obscurer supracellular ones which express themselves as organization, embryology must bide its time, but while awaiting the new insight it is clear what its investigators may do. They may continue to elaborate their physical and chemical descriptions as precisely as pos- sible, though recognizing the limitations of these with respect to the fundamental prob- lem of organization. They may specify, as strictly as possible, the conditions under which work is carried out, in the hope of arriving at possible correlations that may eventually provide new clues. And last, but not least, they will do well to remain as closely preoccupied as possible with the living embryo itself. Spemann, it may be re- membered, had the habit of considering the embryo as a Gesprachspartner who must be allowed to answer in his own language; as a subject, in this sense, rather than a mere object of investigation (cf. Goerttler, *50). The attitude may seem excessively anthropo- morphic, but serves to keep freshly in mind that the embryo, if given the initiative, may have some wise instruction to offer. How in- telligibly the embryo can answer the ques- tions directed towards it depends on the questions asked; these must of course be reduced to simple terms, but they must be terms which the embryo can comprehend. Roux accomplished this, in setting up his first simple alternative; Harrison did so when he isolated the neuroblast, and indeed in many of his subsequent experiments. The most for- midable task of the embryologist is the intel- lectual one of restating the problems, not the technical one of physical manipulation. The embryo makes its replies at a supracellular METHODS AND TECHNIQUES level, and inspiration as how best to address it at this level can come only from the em- bryo alive, from Beobachtung und Reflexion freely expended upon it, in the future as in all the great advances in the past. REFERENCES Abonyi, A. 1915 Experimentelle Daten zum Erkennen der Artemia-Gattung. Zeit. wiss. Zool., 114:95-168. Baltzer, F. 1937 Analyse des Goldschmidtschen Zeitgesetzes der Intersexualitat auf Grund eines Vergleiches der Entwicklung der Bonellia- und Lymantria-Intersexe. Zeitlich gestaffelte Wir- kung der Geschlechtsfaktoren (Zeitgesetz) oder Faktorengleichzeitigkeit | (Gen-Gleichgewicht). Roux’ Arch. Entw.-mech., 136:1-43. 1940 Ueber erbliche letale Entwicklung und Austauschbarkeit artverschiedener Kerne bei Bastarden. Naturwiss., 28:177-187. Bautzmann, H. 1929 Ueber bedeutungsfremde Selbstdifferenzierung aus Teilstiicken des Am- phibienkeimes. Naturwiss., 77:818-827. Boveri, T. 1907 Zellen-Studien. VI. Die Ent- wicklung dispermer Seeigel-Eier. Ein Beitrag zur Befruchtungslehre und zur Theorie des Kerns. Jena. Zeit. Wiss., 43:1-292. 1918 Zwei Fehlerquellen bei Merogo- nieversuchen und die Entwicklungsfahigkeit mer- ogonischer und partiell-merogonischer Seeigel- bastarde. Roux’ Arch. Entw.-mech., 44:417-471. Chun, C. 1880 Die Ctenophoren des Golfes von Neapel; in Fauna und Flora des Golfes von Nea- pel. Monographie I. W. Engelmann, Leipzig. Coe, W. R. 1948 Variations in the expression of sexuality in the normally protandric gastropod Crepidula plana Say. J. Exp. Zool., 108:155-169. Devillers, C. 1949 Explantations en milieu syn- thétique de blastodermes de Truite (Salmo irid- eus). Journ. Cyto-embryol. belgo-néerland., 1949: 67-73. DuShane, G. P., and Hutchinson, C. 1944 Dif- ferences in size and developmental rate between eastern and midwestern embryos of Ambystoma maculatum. Ecol., 25:414423. Gluecksohn-Schoenheimer, S. 1949 Causal anal- ysis of mouse development by the study of muta- tional effects. Growth Suppl., 72:163-176. Goerttler, K. 1950 Entwicklungsgeschichte des Menschen. Ein Grundriss. Springer-Verlag, Ber- lin. Hadorn, E. 1934 Ueber die Entwicklungsleistun- gen bastardmerogonischer Gewebe von Triton palmatus (9) X Triton cristatus ~ im Ganz- keim und als Explantat in vitro. Roux’ Arch. Entw.-mech., 731:238-284. 1937 Die Entwicklungsphysiologische Auswirkung der disharmonischen Kern-Plas- makombination beim Bastardmerogon Triton palmatus (9 ) X Triton cristatus 4 . Roux’ Arch. Entw.-mech., 736:400-489. Harrison, R. G. 1903 Experimentelle Untersu- chungen iiber die Entwicklung der Sinnesorgane Si der Seitenlinie bei den Amphibien. Arch. mikr. Anat., 63:35-149. 1907 Observations on the living develop- ing nerve fiber. Anat. Rec., 7:116-118. 1914 The reaction of embryonic cells to solid structures. J. Exp. Zool., 77:521-544. 1935 Heteroplastic grafting in embryol- ogy. Harvey Lecture for 1933-1934, pp. 116-157. Herbst, C. 1936 Untersuchungen zur Bestim- mung des Geschlechtes. VI. Mitteilung. Neue Gedanken zur Geschlechtsbestimmung bei Tieren. Roux’ Arch. Entw.-mech., 735:178-201. 1937. Untersuchungen zur Bestimmung des Geschlechts. VII. Mitteilung. Ueber die Be- deutung des SO4-ions fiir die Weiterentwicklung und geschlechtliche Differenzierung der Bonellia- Larven und iiber den Einfluss des erhéhten Ca- Gehaltes im SO4-armen Medium auf diese Pro- zesse. Roux’ Arch. Entw.-mech., 736:147-168. Holtfreter, J. 1929 Ueber die Aufzucht isolierter Teile des Amphibienkeimes. I. Methode einer Gewebeziichtung in vivo. Roux’ Arch. Entw.- mech., 117:421-510. 1944 Neural differentiation of ectoderm through exposure to saline solution. J. Exp. Zool., 95:307-343. 1945 Neuralization and epidermization of gastrula ectoderm. J. Exp. Zool., 98:161-209. Hutchinson, C., and Hewitt, D. 1935 A study of larval growth in Amblystoma punctatum and Amblystoma tigrinum. J. Exp. Zool., 71:465-481. Kusche, W. 1929 Interplantation umschriebener Zellbezirke aus der Blastula und der Gastrula von Amphibien. I. Versuche an Urodelen. Roux’ Arch. Entw.-mech., 720:192-271. Luther, W. 1936 Potenzpriifungen an isolierten Teilstiicken der Forellenkeimscheibe. Roux’ Arch. Entw.-mech., 735:359-383. Marsh, G., and Beams, H. W. 1946 In vitro con- trol of growing chick nerve fibers by applied elec- tric currents. J. Cell. Comp. Physiol., 27:139-157. McClendon, J. F. 1910 The development of iso- lated blastomeres of the frog’s egg. Am. J. Anat., 10:425-430. Needham, J. 1942 Biochemistry and Morpho- genesis. Cambridge University Press, Cambridge, England. Pasteels, J. 1937 Etudes sur la gastrulation des vertébrés méroblastiques. III. Oiseaux: TV. Con- clusions générales. Arch. de Biol., 48:381-488. Poulson, D. F. 1940 The effects of certain X- chromosome deficiencies on the embryonic devel- opment of Drosophila melanogaster. J. Exp. Zool., 83:271-325. 1945 Chromosomal control of embryo- genesis in Drosophila. Am. Nat., 79:340-363. Roux, W. 1885 “Einleitung” zu den “Beitragen zur Entwickelungsmechanik des Embryo.” Ges. Abh., 2:1-23. 1888 Beitrage zur Entwickelungsmech- anik des Embryo. V. Ueber die kiinstliche Her- vorbringung “halber” Embryonen durch Zer- storung einer der beiden ersten Furchungszellen, sowie iiber die Nachentwicklung (Postgeneration) der fehlenden Kérperhalfte. Ges. Abh., 2:419-521. 1889 Die Entwicklungsmechanik der Or- 38 ganismen, eine anatomische Wissenschaft der Zukunft. Ges. Abh., 2:24-54. 1912 Terminologie der Entwicklungs- mechanik der Tiere und Pflanzen. Wilhelm En- gelmann, Leipzig. Spemann, H. 1938 Embryonic Development and Induction. Yale University Press, New Haven, Connecticut. Spratt, N. T. 1946 Formation of the primitive streak in the explanted chick blastoderm marked with carbon particles. J. Exp. Zool., 703:259-304. Tennent, D.H. 1910 The dominance of maternal or of paternal characters in echinoderm hybrids. Roux’ Arch. Entw.-mech., 29:1-14. 1925 Investigations on specificity of fer- tilization. Carnegie Inst. Wash. Yrbk., 24:240- 242. Twitty, V. C., and Schwind, J. L. 1931 The growth of eyes and limbs transplanted hetero- plastically between two species of Amblystoma. J. Exp. Zool., 59:61-86. METHODS AND TECHNIQUES Vogt, W. 1925 Gestaltungsanalyse am Amphi- bienkeim mit Grtlicher Vitalfarbung. Vorwort uber Wege und Ziele. I. Methodik und Wirkungs- weise der Ortlichen Vitalfarbung mit Agar als Farbtrager. Roux’ Arch. Entw.-mech., 106:542- 610. Weiss, P. 1949 Differential growth; in The Chemistry and Physiology of Growth, edited by A. K. Parpart, pp. 135-186. Princeton University Press, Princeton, New Jersey. 1950 The deplantation of fragments of nervous system in amphibians. I. Central reor- ganization and the formation of nerves. J. Exp. Zool., 113:397-461. Woodger, J.H. 1948 Observations on the present state of embryology; in Growth in Relation to Differentiation and Morphogenesis, Symposia of the Society for Experimental Biology, No. 2, pp. 351-365. Academic Press, New York. Section III CELLULAK, STRUCT URESANDEAC INV TTY. CHAPTER Cell Constitution FRANCIS O. SCHMITT CLassIcaL cytology was concerned primarily with the elucidation of cell structure by examination, with the light microscope, of living cells and of fixed and stained prepara- tions. In the study of chromosomes, cytology helped provide the foundation of modern genetics by revealing a correlation between chromosome structure and genetic function. Histology has been of immense value in physiology and pathology but in a more limited sense. The limitations are primarily (1) that the fundamental physiological ap- paratus of the cell has dimensions far below the resolving power of the light microscope, and (2) that little is known concerning the chemical composition and biochemical prop- erties of the cell entities which can be op- tically resolved. The development of new optical techniques such as electron microscopy and various cytochemical techniques, particularly the iso- lation and characterization of cell particu- lates, has enormously broadened the cytolog- ical horizon. One has only to examine current texts on experimental cytology, such as the excellent compilation of Bourne (’51), to see how important the biophysical and bio- chemical aspects have become. Indeed, an impressive portion of the work in analytical cytology is now being done by general physi- ologists, biochemists, enzymologists and bio- physicists. This rapid expansion of experimental cy- tology makes it impossible, within the space of this chapter, to deal with the details of cell constitution. Rather, we shall attempt to outline the more salient features of cell structure as they relate to function, particu- larly in growth and development. Nuclear 39 structures and nucleus-cytoplasm relations will be considered in another chapter and will therefore not be included in this presen- tation. Where possible, key references will be cited from which more detailed informa- tion may be obtained. TECHNIQUES OF ANALYTICAL CYTOLOGY Most of the techniques of analytical cy- tology are physical or chemical in nature. The investigator in almost any field of modern experimental biology should have a general notion not only of the existence and potentialities of these techniques but of their limitations as well. Those wishing to employ any of the various techniques effectively as a fundamental portion of their research pro- gram must understand that is it not enough to learn the mere manipulations necessary to “make the gadget work”; it is essential to gain a good grasp of the physical and chem- ical principles underlying the techniques. For the convenience of beginners in this field, the principal techniques are listed, to- gether with references from which an in- troduction may be obtained. STRUCTURE ANALYSIS (DIRECT METHODS) Ultraviolet and Infrared Microscopy and Microabsorption Spectroscopy. Permit resolu- tion higher than the light microscope. Give information about chemical composition of regions as small as 1 sq. » in cells. References: Caspersson (’50), Loofbourow (’50), Barer etaal® C50): Electron Microscopy (EM). Provides resolu- 40 tion to or near the molecular range (15 to 200 A), depending on the resolving power of the instrument and the nature of the specimen. Ultrathin sectioning makes the method applicable to a study of cells and tissues. The necessity that the specimen be dry requires previous fixation or freeze-dry- ing. References: Schmitt (49), Drummond (s0)e"Gosslett Col), Hall) (53): X-ray Absorption Techniques. Permit de- termination of the content and location of particular elements in cells and of the mass of small objects within cells. References: Engstrom (50), Brattgard and Hyden (’52). Interference Microscopy. May be used to follow changes in dry weight of undamaged, living cells. Is more sensitive than x-ray ab- sorption technique. References: Barer (752), Davies, Engstrém and Lindstrém (53). Autoradiography. Reveals location in cells of elements introduced as radioactive iso- topes. Resolution relatively low and only roughly quantitative results so far achieved. References: Glick (49), Doniach (753). Phase Contrast Microscopy. By converting small differences of refractivity into changes of intensity this method permits visual- ization of objects having refractive index similar to that of the surrounding medium. Continuous (rather than discrete) variation of phase offers promise of detection of very small objects within cells. References: Ben- nett, Jupnik, Osterberg and Richards (51). Fluorescence Microscopy. Reveals the pres- ence of fluorescent substances normally pres- ent or introduced into cells. References: Sjéstrand (44), Hamley and Sheard ('47). Reflected Light Techniques. Permit study of opaque objects and very thin objects, such as cell membranes, one dimension of which may be submicroscopic. References: Waugh and Schmitt (’40), Pfeiffer (49). STRUCTURE ANALYSIS (INDIRECT METHODS) Polarization Microscopy. Reveals orienta- tion and state of subdivision of molecules and submicroscopic constituents. Provides some information about chemical composition of cellular constituents. Being applicable to living cells it avoids some of the indeter- minacies of fixation and is a valuable adjunct to electron microscopy. References: Schmidt (37), Frey-Wyssling (48), Schmitt (50a), Bennett (750), Seeds (53). The study of natu- ral and artificial dichroism gives information about the orientation of molecular species which absorb in the ultraviolet (especially CELLULAR STRUCTURE AND ACTIVITY valuable in studying structures, like chrom- osomes, which contain nucleic acid). Dichro- ism studies in the infrared have been em- ployed to deduce the orientation of particular bonds, such as peptide and hydrogen bonds, in proteins. Reference: Perutz, Jope and Barer (50). X-ray Diffraction. Permits deduction of in- tra- and intermolecular structure of biolog- ical materials, the degree of regularity of structure, orientation and particle size. X-ray diffraction can, under certain conditions, be applied to undried materials, e.g., fibrous tissues, crystalline proteins and the like. Though x-ray studies have been concerned chiefly with small, interatomic separations, small-angle techniques now permit investi- gating separations as large as 700 to 1000 A (e.g., large axial repeating patterns of fibrous proteins). References: Perutz (49), Hodgkin (50). Electron Diffraction. Electron microscope techniques make it possible to obtain electron diffraction data from particular, small re- gions located and observed in the electron microscope. Reference: Drummond (750). CYTOCHEMISTRY (HISTOCHEMISTRY ) As originally employed, the terms cyto- chemistry or histochemistry referred to the recognition, localization and quantitation of chemical entities, such as certain enzymes, steroids, polysaccharides, proteins and min- erals in cells and tissues. Localization was the chief purpose and this was accomplished by reactions, usually in sections of fixed mate- rial, which characterize the pure substance in vitro. Subsequently, many other types of procedures, including isolation of particulates by differential centrifugation of fragmented tissues, were included in the field of histo- chemistry by some authors. The writer will continue to use the original more restricted definition, reserving the term “analytical cytology” for the more inclusive field. Perhaps the chief pitfalls are the uncritical assumption that chemical entities react in the complex colloidal systems of protoplasm in the same way they do in simple in vitro systems and the assumption that, during the over-all procedure, the substance in question remains localized exactly as it was in life. The latter point is particularly to be con- sidered in the inevitable application of the electron microscope to the problem of local- ization. References: Glick (749), Gomori (50), Glick, Engstr6m, and Malmstrom 51): es .,§ PSH CELL CONSTITUTION FRACTIONATION OF CELL PARTICULATES The development of the technique origi- nally employed by that pioneer of analytical cytology, Robert R. Bensley, has led to great advance in our knowledge of the con- stitution and function of subcellular systems. Cells are fragmented in a medium usually containing an indifferent non-electrolyte. By differential centrifugation, particles of vary- ing size, density and composition may be isolated. Particulates so obtained exist in an environment very different from that obtain- ing in the cell. Biochemical tests show that certain of the properties of the particulates may be retained after isolation. However, much further progress may be expected from further investigation of the effect of the fragmentation medium on the system. The work of Kopac (’50a,b) strongly indicates that mere cytolysis, to say nothing of com- plete cell fragmentation, produces surface denaturation of cytoplasmic proteins, as in- dicated by the spontaneous Devaux effect. A truly indifferent “Ringer’s solution” for the suspension of cellular constituents has yet to be devised. The term “particulate” has been used to connote subcellular aggregates of dimensions near or below the resolution of the light microscope. Fractionation and characteriza- tion of much smaller particles, such as pro- tein macromolecules and complexes, is a field which is destined to play a very sig- nificant role in the next few decades. This will be facilitated by adaptation of ultra- centrifuge, electrophoresis and other physical chemical methods to deal with very small amounts of sample. References: Glick (749), Schneider and Hogeboom (’51). THE PROBLEM OF FIXATION The purpose of fixation is to treat cells in a manner such that their structure may be examined in the greatest detail with minimal alteration of the normal state, and also to gain information concerning the chemical properties of cell constituents by interpreta- tion of fixation reactions. The advent of electron microscopy has required a reinvesti- gation of the mechanism of fixation (see Palade, 52a). The present problem is essen- tially similar to that which faced cytologists of old. Aside from the fundamental indeter- minacy arising from the fact that any treat- ment must alter the cell, one is faced with the difficulty that, until one knows the struc- ture of the normal living cell, one can only 41 guess whether a given method of preparation preserves that structure unaltered. Except for gross alterations (shrinkage, swelling, etc.), judgment as to the value of a fixative depends upon what the investigator chooses to regard as normal structure. When dealing with a system such as the cytoplasmic “sround substance” where the components are of colloidal dimensions, the electron mi- croscopist’s problem is at least as great as that of the histologist working at light micro- scope resolution. No doubt we may expect in the next few decades description of new “fundamental” protoplasmic structures ob- served with the electron microscope (EM) in tissues fixed in various ways. The extent to which sound interpretation of such fixation “artifacts” can be made will probably depend upon concurrent advances in the physical and analytical chemistry of protoplasmic systems. The problem of fixation is amenable to systematic investigation and some prelimi- nary investigations have already been made. It has long been realized that the fixative should be isotonic with the tissue (especially true in the case of solutions of osmium tetrox- ide, which is un-ionized). The importance of appropriate carbon dioxide tension has not yet been sufficiently realized. The effect of pH, ionic strength, redox potential and the like have been studied (Zeiger, ’49), as well as the chemical properties of particular fixa- tives, such as osmium tetroxide (Hirschler, 43; Schmidt, ’47) and Formalin (French and Edsall, ’°45; Crawford and Barer, ’51). Since many structures can be seen in the living cells by phase contrast microscopy, useful comparisons have been made before and after fixation (Buchsbaum, ’48; Zollinger, ’48a,b). Most cellular objects exist as organized sys- tems of components. It is therefore necessary to preserve not only the individual entities but the organization of the various systems with respect to each other as well. It seems improbable that any procedure will be found which will fully meet these require- ments. The feeling is growing that it may be better not to try to “fix” the various com- ponents chemically but rather merely to remove water by freeze-drying (see Sylvén, 51). This has proven useful to many who work with the EM. THE COLLOIDAL ORGANIZATION OF PROTOPLASM The cell consists of a highly aqueous col- loidal system, enclosed in a complex limiting 42 envelope, and containing particulates of vari- ous sizes and shapes, including a tenuous fibrous system capable of sol-gel reversal. In the microcosm of the living cell the various particulates and subcellular systems have in- timate chemical and structural relationship with each other. The properties of the in- dividual components depend on this inter- relationship and the normal physiological function of the cell requires an appropriate organization of the system as a whole. It is improbable that we can fully learn the chem- ical and structural properties of the individ- ual components after their removal from the cell. However, information about such partial systems must be obtained before the types of interaction which provide the emer- gent organizational properties can be de- duced. Much of the discussion in this chapter will necessarily deal with partial systems of various sorts. For convenience of presentation these sys- tems will be considered according to the geometric form of the components rather than according to chemical composition or supposed function. It is convenient to con- sider three geometric forms: the one-, two- and three-dimensional arrays. These corre- spond, respectively, to the fibrous, membran- ous and globular particles, which will be considered separately, though not in that order. Equally important, of course, is the aqueous milieu with dissolved substances and this will be considered first. WATER AND DISSOLVED SUBSTANCES The water content of protoplasm ranges from 80 to about 95 per cent or more. In terms of molecular species, water represents over 98 per cent of the molecules present in the cell. If we were to look at a cell with “molecular spectacles” we should see little besides water molecules. Water is unique among biological liquids in regard to its physical and chemical prop- erties, owing primarily to its peculiar dipolar structure and the bonding of molecules to each other by means of hydrogen bonds (see Gortner, ’49). Similarly, water is bonded to polar groups of organic molecules in cells and this solvation determines many of the colloidal properties of cell structures. Physio- logical functions, such as contractility, de- pend importantly upon the reversible altera- tions of the reactivity of chemical groupings which, in turn, are conditioned by the shell of water molecules coordinated about the groups. CELLULAR STRUCTURE AND ACTIVITY Important for the chemical coordination of cell processes is the high solubility of organic and inorganic substances in water. Dissolved in the cell water are the mineral salts, lyoenzymes, intermediary metabolites and certain hormones. Energy-rich com- pounds, such as adenosine triphosphate (ATP), presumably diffuse freely in the aqueous phases of protoplasm from their site of synthesis, primarily in the mitochondria, to the molecular effectors where the energy is liberated. In general the water content of cells is greatest in cells in which chemical metabol- ism is high. Embryonic cells have a higher water content and a higher metabolism than do the corresponding cells in the adult or- ganism. Osmotic influx of water resulting, for example, from the hydrolysis of large mole- cules or macromolecules to form a much larger number of smaller molecules may play a role in morphogenesis through change in cell shape (particularly if the water perme- ability of one surface of a cell layer is dif- ferent than that of the other surface). PARTICULATE SYSTEMS The view that many of the essential re- actions of protoplasm occur within, or at the surfaces of, subcellular, macromolecular sys- tems having a high degree of internal organ- ization is not original with modern analytical biologists. It has been inherent in the think- ing of great physiologists, morphologists, geneticists and embryologists for over a cen- tury.* However, the isolation and analysis of several types of particulate systems has now given valuable clues as to the struc- tures responsible for certain of the major functions of the cell and there is every reason to expect that further important advances will be forthcoming by the use of fractiona- tion techniques. For purposes of convenience we shall con- sider in one category the organized particu- lates, whether they occur as microscopically visible structures in the cell or as submicro- scopic aggregates. This is probably justified since in most cases, where such systems have * To realize how fundamental this concept has been, one has only to recall the numerous terms which were invented over the years for the giant colloidal particles in which “life” was thought to re- side: biogens (Buffon, Verworn), physiological units (Spencer), bioplasts (Altmann), gemmules (Darwin), pangens (DeVries), plastidules (Haeck- el), biophores (Weismann), idiosomes (Naegeli) and many others. The gene, in its original form, is a descendent of such concepts. CELL CONSTITUTION been observed in the EM or have been iso- lated by fractionation, they have been dem- onstrated to be aggregates of quite small particles. It has been found that when fragmented cell material is centrifuged, three easily sep- arable fractions can be obtained: the “large granule” fraction, the “small granule” frac- tion and the supernate. We shall consider them in that order. LARGE GRANULE FRACTION Mitochondria. Observed in intact cells, mitochondria appear as granules, elongated threads or as intermediate forms depending on the type of cell and the conditions obtain- ing at the time of observation. In size they vary from 0.5 » to about 1 », when granular, to several micra when filamentous. They may manifest active movement, changing from elongated threads to globules, as observed in phase contrast (Hughes and Preston, 49). Filaments may curl into loops, rotating the while. They may also bifurcate and again fuse longitudinally. This incessant movement suggests both that their substance is semi- fluid in nature and that contact between cyto- plasmic substrate and enzymes contained in the mitochondria may be facilitated thereby. According to Claude (’50), mitochondria are probably capable of self-duplication. They grow by elongation but apparently di- vide transversely. Granules about 0.1 in diameter were observed within the mito- chondria. However, it is not certain that these pre-exist in vivo. EM observations in- dicate that at least certain mitochondria are enclosed in a formed membrane (Mihle- thaler et al., °50; Palade, 52b). This was presumed to be the case from the way in which isolated mitochondria swell or shrink as a function of the osmotic pressure of the suspending medium. In the sarcosomes of insect muscle it is stated that, after osmotic lysis, a residual limiting membrane may be observed analogous to the “ghosts” which re- main after lysis of red cells (Williams and Watanabe, °51, 52). These sarcosomes mani- fest most of the chemical, enzymatic and staining properties of mitochondria. Their presence in abundance in insect muscle is correlated with the need for immediate en- ergy supply to this active tissue. Birefringence has been observed in mito- chondria in living cells by Giroud, Monné and others. The analysis is only fragmentary but may be considered consistent with the view that oriented protein and lipid mole- 43 cules are present. A more detailed study of mitochondria with improved methods of po- larization microscopy may prove rewarding. The chemical composition and enzymatic properties have been studied in mitochon- dria isolated by differential centrifugation. Among the constituents are protein (ca. 50 per cent), lipids (25 to 30 per cent, con- sisting chiefly of phospholipids), nucleotides, flavins, and relatively small amounts of ri- bose nucleic acid (RNA) (Claude, 49). Mitochondria may account for 15 to 20 per cent of the cell mass and, according to Schneider and Hogeboom (751), they may represent as much as 26 per cent of the total nitrogen and 33 per cent of the total protein of the mouse liver. The enzymes demonstrated to be present in the mitochondrial fraction include cyto- chrome oxidase, succinoxidase, cytochrome c, members of the Krebs cycle, p-amino acid oxidase, diphosphopyridine nucleotide (DPN)-cytochrome c reductase, triphospho- pyridine nucleotide (TPN)-cytochrome c reductase, acid phosphatase and ATP-ase. This fairly complete complement of the oxi- dizing mechanism clearly indicates the res- piratory function of mitochondria, and their ability to carry out aerobic phosphorylation and possibly to supply energy for other syn- thetic processes including peptide and pro- tein synthesis. The mitochondria appear to constitute the “power plant” of the cell. This discovery represents one of the great mile- stones in modern biochemistry and analytical cytology. Mitochondria have been studied in some detail with the EM (see Palade, ’52b). An internal layered structure, whose perfection depends importantly on the method of prepa- ration, has been observed. Possibly these lipid-protein layers are the locus of the organized enzymes known to be present in mitochondria. Secretion Granules and Other Members of the “Large Granule” Group. Characteristic particulates are found in the large granule fraction of secreting cells, such as pancreas and liver; they are thought to derive from zymogen and other secretion granules. In the pancreas they may be about as abund- ant as the mitochondrial components. Claude (50) suggests that secretion granules may arise by a progressive transformation of mitochondria, a view expressed long ago by classical cytologists on morphological grounds. Other members of the large granule frac- tion have been assigned special names on at the basis of real or supposed function and composition (see Zollinger, *48a). Among these are: storage granules, starvation gran- ules, degeneration granules, resorption gran- ules, glycogen granules, melanin and _ per- haps the so-called “biosomes.” Centrioles and Kinetosomes. These highly important granules have not as yet been iso- lated from fragmented cells. What little we know about them rests mostly on circum- stantial evidence based on cytological in- vestigations. Being relatively large macro- molecular aggregates they are doubtless very complex chemically. Nothing approaching a fractionation and chemical characteriza- tion of these constituents has yet been at- tempted. The centrioles and kinetosomes are pre- sumably autonomous entities frequently found in association with fibrous structures (cilia, flagella, sperm tails, mitotic mech- anism, trichocysts, etc.), and are thought to be causally concerned with the production and maintenance of the fibrous arrays. In the ciliates much study has been given to these granules, particularly by Lwoff (50) and Weisz (51), from which it is concluded that they are pluripotent morphogenetically and essentially equipotential. Caspari (’50) considers them to be “visible plasmagenes”’! Though their relation to the genome and cytogenes is far from clear, there seems little doubt that these granules play a significant role in processes of differentiation, particu- larly of cortical structures of protozoa and possibly of metazoan cells as well. Aside from the problem of their role in differenti- ation, the biochemical problem is at least as interesting. If they do play a role in the formation of organized protein fibers from cytoplasmic precursors, one may expect that enzymes may be involved. It may not be too much to hope that sufficiently delicate mi- cromethods may be developed by which these structures may be isolated and studied in vitro. SMALL GRANULE FRACTION (MICROSOMES) Claude’s (43) redefinition of the term ‘“microsome,” which is an old one in cyto- logical literature, would include any partic- ulates that are separable from fragmented cells at relatively high centrifugal force (after removal of the heavier large granule fraction). There is no doubt that this frac- tion includes many types of particulates, the composition of which may vary with growth, CELLULAR STRUCTURE AND ACTIVITY differentiation and function. Therefore, anal- yses of the entire fraction cannot character- ize any particular species in the particulate population. However, such analyses indicate something of the nature and possibly the function of these components of protoplasm. According to Claude (749, ’50), micro- somes have diameters in the range of 0.06 to 0.25. They constitute some 15 to 20 per cent of the dry weight of the cell and have been shown to contain protein, lipid, ribose nucleic acid (as much as 60 per cent of that of the whole cell), some nucleotide and a number of enzymes. It is thought that the basophilic properties of cytoplasm may be due primarily to the microsomes and to the endoplasmic reticulum (which will be dis- cussed below). At least five different sub- fractions of the microsome fraction have been obtained by differential centrifugation (Chantrenne, 51, ’52). The relative amounts of RNA and other constituents vary with the size of the particle, and Brachet (50) suggests that during embryogenesis the par- ticles start out small but become more com- plex and richer in RNA as development pro- ceeds, this increasing complexity being an important factor in differentiation. Several enzymes, including phosphatase ATP-ase, dipeptidase and cathepsin have been found to be associated with microsomes. Notions as to the functions of microsomes rest chiefly on circumstantial evidence. Brachet (49, °50) believes they may be con- cerned with protein synthesis and discusses their relation to plasmagenes and viruses. Claude inclines to the view that they are concerned with anaerobic respiration. Char- gaff (45, ’49) isolated particulates 0.08 to 0.124 in diameter from lung and other tis- sue which have high thromboplastic activity and he suggests that cell structuration may be influenced by such intracellular enzyme- containing particles (perhaps along the gen- eral lines of the clotting process in blood). Basophilic Components. The basophilic eranules and filaments were early grouped together under the term “chromidia” by Hertwig. This term has been revived re- cently by Monné (’48) and Lehmann and Biss (’49) to denote the “self-duplicating” RNA-containing submicroscopic particles (biosomes) to which have been ascribed most of the vital properties of protoplasm, including respiration, metabolism, irritabil- ity and morphogenesis! It is possible that a portion of the chromidial apparatus is ac- tually composed partly of adlineated micro- somal particulates, as suggested by Bessis and CELL CONSTITUTION Bricka (49) and partly of the so-called “endoplasmic reticulum” which Porter (753) and Porter and Kallman (’52) regard as a system of RNA-protein-containing structures which may appear canalicular or reticular and which may be concerned with synthetic processes in the cell. A somewhat different description of the same component has been given by Sjéstrand and Rhodin (53). This class of cytoplasmic organelles is probably of considerable functional importance and worthy of further detailed study. SUPERNATE The supernate which remains after cen- trifugal removal of the small granule frac- tion contains chiefly globular proteins (Sorof et al., 51a, b, c; Gjessing et al., 51), organic and inorganic compounds of relatively low molecular weight, “soluble” enzymes, some RNA, nucleotides and the like. LIPOCHONDRIA (GOLGI SYSTEM) Despite the enormous literature (over 2000 papers) written during the last half century on the Golgi apparatus, “we know very little and our hypotheses are pure guesses,” to quote Bensley (751). After reviewing the literature, Palade and Claude (49a, b) con- clude that the Golgi system as seen in cyto- logical preparations is an artifact produced by the action of the alcohol in the fixative upon pre-existing highly refringent droplets which stain with Sudan black and have an affinity for neutral red; the more slowly dif- fusing osmium tetroxide then fixes these arti- facts in the form characteristic of the Golgi apparatus. Structures closely resembling Golgi apparatus were produced by the action of fixatives on lipid-containing fractions of cells. Palade and Claude ignore the evidence that the Golgi material can be seen in the living cell. According to Baker (49), “the Golgi bodies are the most evident objects in the cell,” where they appear as spheroids when viewed with phase contrast. Bensley (51), Gatenby and Moussa (’51) and others favor the view that the “apparatus” pre- exists in the cell as vesicles and canals simi- lar to the structures seen in fixed prepara- tions. The evidence seems to the present writer to indicate that the Golgi bodies represent a pleomorphic system which pre-exists in the cytoplasm and whose colloidal organization is sensitive to alterations of the environment, 45 particularly to acids and organic solvents. Although little is really known about their chemical composition (Sosa, ’48) there seems little doubt that lipids are an important com- ponent and may be primarily responsible for the pleomorphism. It may be desirable, therefore, to adopt Baker’s (50) suggestion to drop the name Golgi apparatus (since this may indeed be an artifact) and refer to the system as “lipochondria.” Adopting this relatively non- commital name may restore respectability to a cellular system which is almost certainly an important protoplasmic constituent and which, because of colloidal sensitivity, has suffered much uncritical abuse. The evidence in the cytological literature that lipochondria are intimately associated with the process of secretion, though circum- stantial, is perhaps as good as that which, in the same literature, associated mitochondria with respiration (which later biochemical evidence proved to be correct). To what ex- tent the lipochondrial system will prove to be multicomponent, with “internal and ex- ternal” components, remains to be seen. The author leans to the view that a portion of the instability and of the difficulty of obtain- ing the material as particulates from frag- mented cells (see Worley, 51) may be due to the fact that certain of the components exist in the smectic paracrystalline state (a view not inconsistent with the meager polar- ized light data available). If so, the pleo- morphism, ease of solvation and sensitivity to alterations in the chemical environment would be expected. Fairly substantial evidence points to the possibility that ascorbic acid and phospha- tase occur in lipochondria. Gersh (’49), ap- plying the Hotchkiss reaction to frozen-dried material, concluded that the material which reacts positively is a glycoprotein which may provide ‘fa suitable frame-work for the or- derly arrangement of enzymes and other activities.” This must be regarded as pure speculation at the present time. The determination of the composition and function of lipochondria has lagged behind that for mitochondria for the reason that the subject is very much more complex and diffi- cult. However, especially if the material is involved in the vital process of secretion, the problem is all the more challenging and worthy of a renewed vigorous attack in- volving the chemical and enzymatic as well as the morphological approach. Quite pos- sibly progress would be accelerated by the exploitation of the most favorable source of 46 material, whether that may prove to be long known or yet to be discovered. FIBROUS SYSTEMS OF CYTOPLASM The linear, fibrous array is geometrically, physically and chemically well adapted to subserve a number of fundamental biological processes. It permits the specific adlineation and segregation of chemical and biological entities, as in chromosomes. It provides me- chanical strength and rigidity: the tensile strength of collagen fibers may be as high as 100 kg./cm.? (as high as some metals). Con- tractility is almost uniquely a property of fibrous material. Most fibers of biological origin, particularly those composed of pro- teins and their complexes, are capable of contraction when subjected to appropriate chemical environment. To grasp the proc- esses underlying the basic phenomena of srowth and differentiation, insofar as they concern the molecular machinery of the cell, the investigator must delve fairly deeply into the physical and chemical mechanisms of contractility. This involves the submicro- scopic fibrous lattices responsible for sol-gel transformation as well as the more obvious fibrous systems described in classical cy- tology. Most of our detailed knowledge about fibers has been derived from a study of pro- teins, polysaccharides, nucleic acids and their conjugates which can be purified and subjected to physical chemical analysis. This knowledge has been of great value in study- ing the intracellular fibrous systems which have not yet been isolated and characterized. Accordingly, it will be useful to discuss a few of the salient features of purified pro- teins as well as to describe the far less per- fectly understood intracellular systems. POLARIZATION OPTICAL ANALYSIS Protein fibers usually manifest positive in- trinsic and form birefringence. This permits detection, in living cells, of fibrous arrays and their direction of orientation even though the structures are too thin to be resolved by the light microscope. This is made possible by high sensitivity of the polarization optical method (Swann and Mitchison, 50; Inoué, Hell) Application of the Wiener theory of form birefringence led to the deduction that fibrils which show positive form birefringence are composed of still finer fibrous structures whose thickness is small with respect to the CELLULAR STRUCTURE AND ACTIVITY wave length of light. This deduction has been verified with the EM in each case so far investigated. Indeed, electron microscopy has shown that the fibrous elements respon- sible for the form birefringence may occur as filaments of characteristic width, fre- quently about 100 to 300 A. When lipid or nucleic acid is associated with fibrous proteins their presence can be deduced from polarized light examination, since the sign of their birefringence is oppo- site to that of fibrous proteins. X-RAY DIFFRACTION ANALYSIS The eventual aim of x-ray diffraction crystallography is to determine not only the major structural features of the molecule but also the exact position of each atom within the molecule. This has not yet been possible with the crystalline, globular proteins which produce hundreds or thousands of diffraction spots in the x-ray pattern. The task is enor- mously more difficult in the case of the fibrous proteins whose patterns may be rela- tively poor in diffractions. The polypeptide chain is the structural unit of fibrous proteins. The diffraction analysis centers around the determination of the configuration of the polypeptide chains, the interchain relationships and the larger structural features which relate to molecular domains or supermolecular pat- terns. Configuration of Polypeptide Chains. In cer- tain types of fibrous proteins, such as silk, feather keratin and highly stretched hair and muscle protein, the polypeptide chains are thought to be nearly fully extended (Ast- bury’s beta type). Extended chains may form fabrics which Pauling and Corey (’51a) call “pleated sheets” because the two chain bonds of the a carbon atom form a plane which is perpendicular to the plane of the sheet, form- ing “pleats.” However, in most fibrous, and probably globular, proteins the chains have a folded configuration (Astbury’s alpha type) whose detailed configuration may differ among the various types of proteins. The precise configuration of the folded chains has not yet been unequivocally dem- onstrated in any protein. The best evidence at present favors the view that the chains may be helically coiled, the pitch and num- ber of amino acid residues per turn varying with different proteins (Pauling and Corey, *Hilibje; Bear, “52: "Bulle °52) Hora discus sion of recent developments in this field see Kendrew (754) and Edsall (’54). CELL CONSTITUTION Large Periods in Protein Fibers. Small- angle x-ray data have shown that some fibrous proteins are characterized by large axial and transverse periods (Table 1). Al- though it has been suggested in several cases that the axial period represents the length of the molecule, this has not yet been proven. Indeed, in some cases (collagen, paramyosin) it has been possible, by solution and appro- priate reconstitution, to produce fibrils which manifest an axial repeat period several times that of the native fibers. It is of course pos- sible that, in the native state, the chains are coiled and that, in passing from the normal length of the period to the “long-spacing”’ form, the chain molecules become fully ex- tended and then manifest the long-spacing period. From the x-ray data, it has been possible also to deduce something about the intra- period fine structure. In the case of collagen, Bear (752) suggests that the neighboring chains form parallel, ordered regions alter- nating with more poorly ordered regions. These regions have precise position within the main repeat period, giving rise to intra- period bands as seen in the EM. The rela- tive order or disorder is thought to depend upon the length of the side chains and the perfection of packing of these side chains between neighboring polypeptide chains. This would imply that there is a fairly pre- cise sequence of amino acid types, if not of amino acids themselves, along the chain molecules, a view for which there is analyt- ical evidence in the case of globular proteins (Sanger, *52). It seems possible that characteristic, long axial repeating periods may be discovered in many other fibrous proteins thus far not ade- quately investigated. This is of considerable importance because such periods provide molecular “fingerprints” by which proteins may be identified in tissues and in various kinds of preparations and give information about molecular structure. Classification of Fibrous Proteins. Astbury early pointed out that the wide-angle, alpha, patterns of certain proteins are closely simi- lar. Keratin, myosin, epidermin and fibrin are grouped together as the “KMEF class.” A different large-angle pattern characterizes a group of proteins called by Astbury the “collagen class.” The latter group includes the collagens of vertebrates and invertebrates and a number of proteins from forms as low phylogenetically as the Porifera. Astbury suggested that the polypeptide chains of members of the collagen class are not folded 47 upon themselves as are those of the KMEF class but are essentially fully extended. Re- cent analyses (Bear, °52) indicates that chains of the members of the collagen class probably have a helical configuration as do members of the KMEF class although the geometry of the helix may differ in detail. It is interesting to note that, despite strik- ing differences in chemical composition, various members of the collagen class which Tasie 1. Large Repeating Periods in Fibrous Proteins REPEATING PERIOD PROTEIN FIBER AXIS LATERAL PERIOD, A PERIOD, A 8 Keratin 95 34 a Keratin 198 82 Collagen 640 a2 ‘Long Spacing” (~2100) = Paramyosin 720 261 “Long Spacing” (~1400) Muscle (Actomyosin?) 405 86 Fibrin 230 == Trichocyst Protein 550 = (2200) have been carefully studied manifest not only the same large-angle pattern but the same axial long period (about 640 A) as well. Each member of the KMEF class has a different, characteristic axial long period (see Table 1). The meaning of these facts will probably not be fully understood until much more is known about the chemical composi- tion and molecular configuration of these proteins. ELECTRON MICROSCOPE ANALYSIS The hierarchies of fiber sizes are well demonstrated in EM studies. It has long been known that macroscopic fibers, such as muscle or tendon, having widths from 10 to 100, are composed of thinner fibrils which may be at or below the limit of resolution of the light microscope (Heidenhain). The elec- tron microscope has shown that these fibrils (whose presence was deduced from their positive form birefringence) have widths ranging from several hundred to several 48 CELLULAR STRUCTURE AND ACTIVITY thousand Angstr6m units. The fibrils, in chain. Figure 1, taken from Bear (’52), illus- turn, are composed of still finer arrays which trates this fibrous hierarchy. Long axial re- may be called filaments and which may be peat periods demonstrated by small-angle of the order of 100 A in width, or smaller. x-ray studies have been observed in the EM The ultimate fibrous array is the polypeptide in the form of cross-striations consisting of Microscopic Level ground substance fibroblast collagen primitive fiber elastic fiber | L Ul tramicroscopic Level tii Electron Optical Level HUAN ate /] ill, WE MN it 6 AN b >, axial period ql ll | UH filament UU Uae X-Ray Diffraction Level interband SSS SS ee x x og a, SN fa CELL CONSTITUTION bands of varying density, thickness and af- finity for stains like phosphotungstic acid. This is true for vertebrate muscle, paramyo- sin and collagen. Detailed intraperiod struc- ture, in the form of fine bands having char- acteristic position and density, has been ob- served in several fibrous proteins. The com- bination of x-ray and EM data may provide clues as to the physical and chemical mean- ing of such accurately repeating axial struc- ture. In some cases it is suspected that a globular component may be associated with the fibrous component in forming the pe- riodic structure and in determining the properties of the fiber. Cilia, Flagella and Sperm Tails. It has long been suspected, on the basis of their positive form birefringence (Schmidt, ’37), that these microscopically visible fibrous structures are composed of still finer submicroscopic fibrous components. EM examination has demon- strated that they are, in fact, bundles of fibrils, each fibril being 300 to 600 A in width and running the full length of the cilium, flagellum or sperm tail. A very interesting point is that the num- ber of these fibrils is relatively constant. In mammalian sperm tails there are usually eleven fibrils, of which two may be thinner than the remainder. According to Fawcett and Porter (752), molluscan cilia also con- tain eleven fibrils, two in the center sur- rounded by nine peripheral filaments. The detailed mechanism of the formation of these fibrils, presumably in association with the centriolar apparatus, constitutes a challeng- ing subject for future EM study. There is some indication of an axial re- peating pattern in such fibrils (Grigg and Hodge, °49; Fawcett and Porter, ’52) but thus far no clear-cut proof of such a period has been demonstrated. Around the sperm tail may be distin- suished a sheath composed of one or more helically coiled fibrils. For a good descrip- tion of the fine structure of plant cilia the reader is referred to Manton (52). The structural and chemical basis of the contraction of cilia, flagella and sperm tails has not yet been demonstrated. Unlike mus- cle, the contraction is not a longitudinal shortening and thickening of the fibers but rather a screw or helical type of beating. It is not clear whether the filaments are them- selves contractile or whether they provide chiefly mechanical rigidity as was supposed in older theories. The flagella of motile bacteria appear to be individual fibrils about 120 A wide pro- 49 truding from the bacteria rather than bun- dles of fibrils aggregated to form individual flagella. The flagella may be isolated in rela- tively pure form and their composition and structure studied. According to Astbury and Weibull (49) they give the alpha wide- angle diffraction pattern characteristic of the KMEFF class. An axial period of about 400 A, similar in magnitude at least to that of ver- tebrate muscle, has also been found by Ast- bury (personal communication), who sug- gests that such flagella may be considered “monomolecular hairs or muscles” (Astbury, A) Trichocysts. The Paramecium trichocyst is an elongate, thin-walled tube terminating in a dense, pointed tip. The wall as well as the tip shows birefringence positive with respect to the axis of the trichocyst (Schmidt, ’39). In the EM the wall of the tube shows cross- striations with an axial period of 550 A (pos- sibly the main period is four times this) and intraperiod fine structure (Jakus, °47). It would appear that the trichocyst wall is composed of a fibrous protein or conjugated protein of unknown nature, the fiber axis being parallel to that of the trichocyst shaft. It will be interesting from the comparative biostructure view to discover to which class of proteins, in the x-ray diffractionist’s terms, this material belongs. CYTOPLASMIC SUBMICROSCOPIC FIBROUS STRUCTURES The view that cytoplasm contains, besides the fibers visible in the microscope, a sub- microscopic fibrous network or lattice rests not only on the tendency for fixatives to pro- duce fibers but also on a number of physical properties manifested by unfixed protoplasm. When protoplasm is made to flow through a capillary tube the flow is non-Newtonian in nature, e.g., the rate of flow is not propor: tional to the force applied (Pfeiffer, °40). This is a property of solutions containing elongate, threadlike, rather than spherical, particles. When frog eggs were centrifuged in a centrifuge microscope equipped for ob- servation in polarized light, it was found that birefringence was developed which is positive with respect to the direction of the centrifugal field (Pfeiffer, ’42). This indi- cates the existence in the cytoplasm of elon- gate, fibrous particles which were aligned by the centrifugal force. The viscosity of cyto- plasm can be estimated by a variety of methods (Heilbrunn, 52). Although it is possible to do this only semiquantitatively, 50 the results support the view that cytoplasm contains submicroscopic fibrous particles. The thixotropy sometimes displayed by cy- toplasm favors the same conclusion. When cells at rest or in division are subjected to high hydrostatic pressure the phenomena ob- served (decrease in viscosity, reduction of amoeboid movement and cell division) re- semble those observed in certain fibrous sys- tems (Marsland, 51) and are interpreted as due to the breaking of interparticle bonds faster than the rate of their reformation (Eyring, Johnson, and Gensler, ’46). When mineral oil is injected into the axoplasm of a squid giant fiber the oil takes on an elon- gated, rather than a globular, shape (Cham- bers and Kao, *52), indicating the presence in the axoplasm (a specialized form of cyto- plasm) of an oriented submicroscopic fibrous lattice. From studies involving magnetic con- trol of artifically introduced intracellular metallic particles, Crick and Hughes (749) suggested the presence in cytoplasm of sub- microscopic particulates in fibrous arrays. It is supposed by some that this fibrous submicroscopic lattice—sometimes called the “cytoskeleton” —is an internal molecular framework characteristic of all cells. Polarized Light Evidence. Tissue cells as such have received relatively little careful study in polarized light as compared with muscle and nerve fibers and other specialized structures which lend themselves well to such study. Fresh nerve cells show bire- fringence of the type expected if they pos- sessed submicroscopic fibrous arrays with the same orientation as the neurofibrils of fixed and stained preparations (Chinn, 738). Hil- larp and Olivecrona (’46) investigated many types of epithelial cells and observed form birefringence positive with respect to the base-apex axis, together with intrinsic bire- fringence negative with respect to this axis. This was interpreted to indicate the presence of protein (or possibly nucleoprotein) sub- microscopic filaments oriented parallel to the base-apex axis with lipid molecules oriented with their paraffin chains perpendicular to the protein filaments. In the fresh marine egg it is possible to demonstrate regions in the cytoplasm which show positive birefrin- gence (Inoué and Dan, *51; Swann, ’51a, b; Hughes, 52). In certain instances it has been possible to show that such regions do, in fact, possess a fibrous structure as seen in fixed cells with the electron microscope (Mc- Culloch, 52). In the case of nerve axoplasm a semiquantitative analysis has been made of the form birefringence, which is positive CELLULAR STRUCTURE AND ACTIVITY with respect to the long axis of the fiber (Bear, Schmitt and Young, ’37). The con- clusion that an oriented fibrous submicro- scopic array exists has been confirmed by subsequent EM study (Schmitt, ’50b; Fernan- dez-Moran, ’52). Electron Microscope Evidence. In most of the cell types thus far studied, including fibroblasts, blood, tumor, nerve and_ liver cells, examination of thin sections has re- vealed a reticular fibrous system in the cyto- plasm. The appearance of the fibrils depends importantly on the fixative used. Acid fixa- tives usually produce coarse, somewhat ir- regularly contoured, fibrils probably result- ing from the aggregation of finer filaments. Lehmann and Biss (49) identify the rather coarse fibrous structures which they observed in Tubifex eggs with the chromidial strands of Monné, though they may also be inter- preted as cytoplasmic filaments upon which microsomes are attached. In well fixed prepa- rations the fibrous structures may appear as thin filaments or as the system which has been called by Porter “endoplasmic reticu- lum.” The latter may appear tubular or as a linear series of elongate, bladder-like struc- tures. While this structure appears to be rather characteristic of cells generally, the structure in unfixed cytoplasm which corre- sponds to this picture in fixed cells is diffi- cult to assess. In a careful EM study of the unfertilized sea urchin egg, McCulloch (52) observed bundles of nodose fibrils, about 1100 A in thickness, oriented in directions consistent with the positive birefringence which he ob- served in the living egg cells. McCulloch’s failure to observe fibrous structures generally distributed in the cytoplasm may suggest that such fibrous molecules are either too thin to be resolved by the EM under the conditions or that the fixation caused reac- tions which obscured them. While the EM evidence to date is not in- consistent with the view that the ground substance of cytoplasm contains a fibrous component, it must be realized that the sys- tem is highly complex and that much must be done before the high resolution of the EM will reveal the true molecular basis of the fibrous system in cytoplasm. Chemical Identification of the Fibrous Con- stituents. To gain an understanding of the role of the fibrous cytoplasmic constituents in cell functioning it is necessary eventually to isolate these constituents by fractionation procedures and to determine their composi- tion and physical and chemical properties. CELL CONSTITUTION The history of muscle physiology reveals that it was only when the fibrous proteins of muscle were isolated (at least with a sem- blance of purity) that rapid advances were made in our concepts of the mechanism of muscle contraction. We are still far from a full understanding of this process despite the enormous work which has been devoted to its investigation by the most competent biochemists over the past few decades and despite the fact that muscle proteins may be obtained in any desired quantity. The prob- lem of the cytoplasmic fibrous proteins is enormously more difficult because their con- centration in cytoplasm is low and there are many other substances present which com- plicate isolation without significant chemical alteration. Constituents other than the pro- teins may also occur as fibrous particles in cells (nucleic acids and polysaccharides and their complexes with proteins, lipids and other materials) and this further complicates the problem. Protein fractionation procedures which proved so effective in isolating the proteins of blood plasma are currently being applied to the fractionation of tissue proteins in vari- ous laboratories. However, even granting that this had been successfully achieved, the problem of the localization of components in particular cell types would remain. In fortu- nate instances the localization problem may be overcome by the use of especially favor- able material. Thus the use of the giant nerve fiber of the squid permits the extrusion of axoplasm uncontaminated by nonaxonal ma- terial. Already one axonal protein in mono- disperse, relatively pure form has been char- acterized and the existence of several other proteins demonstrated (Maxfield, 751). Studies in the author’s laboratories are di- rected at the isolation and characterization of the protein of the axon filaments (the fibrous constituents of neurofibrils) and some progress has already been made. Perhaps when success has been achieved we shall have a better notion of the role of these pro- teins in nerve function. A beginning has also been made in the isolation of proteins from single cells. Mirsky (36) had suggested that a “myosin-like” protein is present in sea urchin eggs and that the state of aggregation of this macromolecu- lar component varies with physiological state. Monroy (750) has undertaken to fractionate the macromolecular constituents of the sea urchin egg and to characterize them by elec- trophoretic and _ ultracentrifugal methods. Though he has not yet succeeded in isolating at individual components in pure form he has obtained evidence for the presence of at least five components and has demonstrated changes which occur in several of the com- ponents on fertilization. The adaptation of physical chemical meth- ods (ultracentrifuge, electrophoresis, stream- ing birefringence, viscosity, etc.) to very small samples (e.g., from single cells) would greatly facilitate the analysis of cytoplasmic constituents. Intracellular Fibrogenesis. Fibroblasts are thought to produce a protein which is the precursor of the collagen fibers of connective tissue. Thus far, there is little evidence for the formation of fibrous collagen (identifiable with the EM) inside the fibroblast cells. Porter and Vanamee’s (749) EM studies of collagen formation by fibroblasts in tissue culture suggest that the collagen fibrils are formed at or near the external surface of the cells. It is possible that further detailed stud- ies of this sort may lead to information about the process by which the elongate collagen molecules are formed in the cell. If, as the writer suspects, the intracellular precursor occurs as discrete, probably elongate mole- cules, the task will require a chemical as well as a morphological approach. For present purposes we are primarily con- cerned with the mechanism of formation of the fibrous materials characteristic of the cell itself rather than with the “secretion” of fibrous substances. Fibrous arrays are the chief building stones of cell structure. Yet almost nothing is known about the process by which the constituent molecules are as- sembled into fibers. Definitive evidence must await chemical isolation and characteriza- tion as discussed in the preceding section. Meanwhile background information obtained by indirect methods and by a study of fibro- genesis in vitro will prove useful in guiding our thoughts and in devising direct experi- ments in cells themselves. Intracellular fibrogenesis may be thought to occur along two (not mutually exclusive) lines: (a) that which may occur spontane- ously without the known action of enzymes, and (b) that known to be controlled enzy- matically. The precursor molecules, which become linked by covalent or electrostatic bonds, may themselves be highly elongate or more nearly spherical in shape. Fibrilization without Enzyme Action. Tn this type the adlineation of protein molecules, chiefly through electrostatic or hydrogen bonds at opposite ends of molecules, is chiefly involved. Stability of the fiber depends im- a2 portantly on ionic strength and on linkage through particular types of groups, such as SH groups. A good example of this type is the muscle protein, actin. In solutions of low ionic strength the viscosity is relatively low and the actin exists as globular particles. In- crease in ionic strength increases the viscosity markedly due to the formation of fibers by the adlineation or polymerization of globular actin molecules. Reduction of ionic strength, as by dialysis, reconverts the fibers to globu- lar molecules. This type of reversible process may play an important role in muscle con- traction. According to Straub and Feuer (50) the globular actin contains ATP as a functional group; removal of the terminal phosphate causes the linking of actin mole- cules, in which process SH and Ca may be involved. However, the process requires no enzyme such as ATP-ase. Tropomyosin, investigated extensively by Bailey (’48a,b), and thought to be the pre- cursor of myosin, is also capable of reversible elobule-fiber transformation depending upon the ionic strength. Insulin, whose chemical composition and structure have been thoroughly investigated, normally exists in the form of globular mole- cules but can be converted to the fibrous form by heating in acid solution. The fibrous form, which has no biological activity, can be re- converted to the globular form with full restoration of activity (Waugh, °48). If a minute amount of the fibrous form is added to a solution containing globular insulin, all the insulin comes down in fibrous form. The transformation is quantitative and occurs even in the presence of other proteins and foreign substances (the process has been used by Waugh for in vitro assay of insulin from crude preparations). This suggests the possibility that fibrous proteins may be formed autocatalytically in cells once the fibrous form is produced. Whether a similar tvpe of process plays a role in the differentia- tion of cellular proteins remains to be de- termined. Enzymatically Induced Fibrilization. The classic example of this type is the conversion of elongate fibrinogen molecules into fibrin under the influences of thrombin. Neglecting the complex system of activators and inac- tivators which control the formation of thrombin from prothrombin, the fibrogenesis may be thus described: Fibrinogen molecules, having dimensions of about 35 * 600 A, un- der the influence of thrombin, are converted into fibrils which appear cross-striated in the CELLULAR STRUCTURE AND ACTIVITY EM. The axial period is about 230 A (Hawn and Porter, ’47). Hall (49) has demonstrated intraperiod fine structure and suggests that the striations are due to lateral alignment by colloidal forces of components within the fibrinogen molecules. Ferry (’52) has offered additional suggestions concerning the forces and groups involved in the lateral and Jon- gitudinal aggregation of fibrinogen mole- cules to form fibrin. It is possible that fibro- genesis of other proteins may involve a complex system analogous to that of blood clotting. Underlying the complex, balanced sys- tem of activators and inactivators in blood clotting is the requirement that the clotting system be under the strictest biological con- trol; breakdown of this control may lead to death. The mitotic mechanism, with its elaborate spindle and astral fibrilization, may have similar general properties. Many investiga- tors have suggested a basic similarity between the mitotic mechanism and blood clotting and have suggested that it may be no less complex. Heilbrunn (752) and his associates have emphasized this view and find that cer- tain anticoagulants (heparin, Dicumarol) prevent formation of the mitotic figure in the marine egg. They believe release of Ca** from the cell cortex is an essential factor in cytoplasmic clotting. However, un- til the structural proteins and other compo- nents are isolated and their mechanism of action elucidated, such analogies must be considered speculative. Chargaff (45, °49) finds that particles from the large-granule fraction of lung cells have high thrombo- plastic action. He suggests that such granules may be involved also in intracellular struc- ture formation. Careful polarization optical studies of di- viding cells have been made by Swann and Mitchison (’50), Inoué and Dan (51) and Swann (’51a,b). The results, well summa- rized by Hughes (°52), indicate that the spindle and asters contain oriented fibrous protein particles which must be very thin (possibly tens or hundreds of Angstrém units) and about as highly hydrated as the proto- plasm surrounding them. Swann concluded that the chromosomes liberate a “structural agent” which affects the organization of the spindle and astral fibers, decreasing their birefringence. No evidence is yet available as to the nature of such a substance. The experiments demonstrate the great sensitivity of the polarization optical method to detect alterations of protoplasmic ultrastructure. CELL CONSTITUTION THE PARACRYSTALLINE (MESOMORPHIC) STATE; TACTOIDS, COACERVATES AND LONG-RANGE FORCES The crystal is possessed of perfect, non- statistical order. However, symmetry may occur in certain types of materials in one or two dimensions as well as in three dimen- sions. There are, in fact, various transitions of ordered arrangement between that in a crystal and the lack of order characteristic of a liquid. Hermann showed that there are 18 possible transitional states, which have been called paracrystalline or mesomorphic. Two of these, the nematic and smectic states, are of great importance in the microstructure of protoplasm and will be briefly character- ized. The nematic state concerns the fibrous systems discussed in the preceding and fol- lowing sections. The smectic state concerns more importantly a subsequent section on lamellar, membranous systems. However, for the sake of clarity, both states are considered jointly. THE NEMATIC STATE This state is characterized by arrays of thin elongate particles which are constrained to remain oriented parallel to a preferred axial direction. The particles are free to rotate about their axes or be translated later- ally or axially. Such systems are birefring- ent and show evidences of the parallel orien- tation of the particles in x-ray patterns. The particles may be regularly spaced laterally, the interparticle distance depending on the concentration, the charge distribution on the particles, the ionic strength and the pH of the aqueous medium. The conditions which determine the state of a nematic system were investigated in some detail in the case of solutions of tobacco mosaic virus (TMV) by Bernal and Fankuchen (’41) and by Oster (50). This will be discussed below in connection with tactoid theory. EM studies have greatly en- hanced our knowledge of such systems. THE SMECTIC STATE In smectic systems there is one degree of freedom less than in nematic systems. The elongate molecules, though oriented in a common direction, are constrained to lie in planes perpendicular to the direction of molecular orientation. The system has thus a layered, planar structure. The molecules may have freedom of rotation and of lateral translation but are constrained to remain a3 within their own planes (like people stand- ing on floors of a building but unable to walk between floors). The lipids compose the chief type of smectic system encountered biologically. In such systems the lipid molecules usually oc- cur as bimolecular layers, the polar ends of the molecules being located at the aqueous surfaces. The distance between double layers depends upon the water content, ionic strength, specific ions and pH. In some cases, e.g., cephalin, lipid double layers may be separated by water layers thicker than the double layers themselves (Palmer and Schmitt, ’41). Because such layered lipid and lipid-protein systems are of great importance biologically, as in the nerve myelin and in cell membranes, it is important to gain an understanding of the forces and conditions which determine their stability. Most cellular lipid double layers are com- posed of mixed lipids: phospholipids, cerebro- sides and steroids. The ability to incorporate water extensively between double layers of mixed lipids is determined primarily by certain of the lipids, notably the cephalins, which present negative charges at the aque- ous surfaces. When positively charged ions, particularly multivalent cations such as Ca** or histones and protamines, are added the water is expelled from between the double layers, causing the system to precipitate and lose its characteristic colloidal texture. This illustrates the great sensitivity of such sys- tems to changes in the ionic environment. Both nematic and smectic systems show birefringence which is positive with respect to the optic axis which parallels the direc- tion of orientation of the molecules. The sign of the birefringence depends upon the nature of the molecules and on the distance between layers (extensive solvate layers may produce lamellar form birefringence which is uni-axially negative). TACTOIDS AND LONG-RANGE FORCES Mesomorphic systems, both of the nematic and the smectic types, frequently form tac- toids and coacervates (Kruyt, 49) in which phases form spontaneously depending on the concentration of the material and the na- ture of the environment. The phases may separate microscopically or macroscopically in vitro. These are illustrated in the case of TMV suspensions. The threadlike macromole- cules aggregate laterally to form birefrin- gent lens- or spindle-shaped droplets. These are called positive tactoids (Bernal and 54 Fankuchen, ’41). Negative tactoids may also be formed, in which case the dispersed phase is low in protein content. Tactoids may also be formed by smectic systems such as dis- persions of certain types of lipids, soaps and certain inorganic materials such as iron oxide and vanadium pentoxide. The forces invoived in the formation and behavior of tactoids have received consider- able attention theoretically and experimen- tally (for recent reviews see Verwey and Overbeek, °48; Oster, ’50). The behavior of the system is determined importantly by the shape, size and concentration of the particles (rodlets or platelets), their charge density, the concentration and valence of the ions in solution and, of course, the pH. De- terminative is the type and extent of the ion atmosphere about the particles, since the interaction between the particles is primarily electrostatic in nature. Except at the isoelectric point the particles will bear a predominantly negative or positive charge. Since the reaction of most protoplasmic sys- tems is near to neutrality and since most of the proteins have isoelectric points in the acid range, the charge will, in general, be negative (basic proteins such as histones and protamines are of course positive). Assembled about the particles are the ions of opposite sign (counter ions). The thickness of the diffuse double layers, hence the interaction between the particles, is inversely propor- tional to the square root of the salt con- centration. It has been shown that colloidal particles of this sort will, in the presence of their counter ions, repel each other, the equilibrium distance of separation varying with the concentration of ions, the repulsive force between the particles and the potential of the particles. Verwey and Overbeek sug- gested that van der Waal’s attraction between particles of this sort will, in the presence of their counter ions, repel each other, the atmospheres. Van der Waal’s forces are very short-range (inversely proportional to the sixth power of the distance in the case of atoms) but, being additive, become important in the case of large particles such as virus particles, fibrous proteins or smectic lipid systems. The intensity of the van der Waal’s attraction and of electrostatic repulsion de- pends upon the conditions obtaining in each particular system and it would be pointless for our purposes to discuss the theoretical aspects further. Suffice it to say that the evidence that long-range attractive forces need be invoked to explain the behavior of CELLULAR STRUCTURE AND ACTIVITY most colloidal systems which have thus far been carefully studied is highly questionable. Long-range forces have been supposed by some to be important in determining the structure of systems composed of long chain molecules containing alternate single and double bonds. London (42) has shown that such conjugated double bond chains may at- tract each other over distances comparable to the lengths of the chains owing to the fact that they behave like oscillators. Such forces may be quite specific. Rothen (747, 50, °52) has invoked them to explain his experiments in which he finds that films of antigen and antibody may interact with each other even when separated by inert films several hundred Angstrém units thick. Enzymes were also thought to attack sub- strate molecules separated by similar dis- tances. Although this finding has given rise to much speculation on the role of long-range forces in protoplasmic systems [even includ- ing the attractive forces during the somatic pairing of dipteran chromosomes (Cooper, 48) |, Rothen’s experiments have been criti- cized on technical grounds (Iball, 49; Karush and Siegel, ’48; Singer, °50) and on theoreti- cal grounds (Pauling, "48; Winter, °52). At the present writing it seems that Rothen’s re- sults are susceptible of explanation without reference to long-range forces. However, as pointed out above, each system must be con- sidered as a special case in order effectively to analyze the possible role of long-range forces. Bernal (49) is convinced that “they must play a very large part in the inner organization of the cell” and that the prop- erties of the mitotic spindle are explicable in terms of a tactoidal organization. The latter view has been strongly criticized (Schrader, 44; Hughes, 52). SOL-GEL TRANSFORMATIONS, CONTRACTILITY AND THE CELL CORTEX In a sol the asymmetric colloidal or macro- molecular particles have a relatively large average interparticle separation, depending on the concentration of particles, charge on the particles, pH, ionic strength and type of ionic environment, as discussed in the pre- ceding section. When these environmental factors are altered the particle interaction may be greatly increased, causing the sol to be transformed into a gel. Thus a sol containing elongate macro- molecules of nucleic acid dissolved in salt solution may be transformed into a gel by CELL CONSTITUTION removal of salt by dialysis. An acetic acid solution of collagen containing only 0.1 per cent protein may be converted into a gel by dialysis against water. In this case the collagen filaments are very thin (< 50 A) and very long (~ micra); there are many cross-bonds between particles and a gel is formed although the concentration of protein is very low. When the average par- ticle length is much less, as in gelatin, the concentration necessary for gelling is much higher. Sol-gel transformations of such sub- stances are freely reversible. Gels may be broken down also by the action of depolymerases which convert the elongate macromolecules into lower poly- mers or monomers. Ribose and desoxyribose nucleic acid depolymerases and hyaluroni- dase are examples of such action. About the enzymatic polymerization of such nucleic acid and polysaccharide molecules much less is known. Much has been learned by a study of sol- gel transformations produced in systems con- taining purified components. Indeed, only in such systems can we possibly hope to evalu- ate the physical and chemical factors in- volved. However, biological systems mani- festing sol-gel transformations may be very much more complex, frequently involving a system of enzymes, kinases and antikinases. Although it is useful as a first approximation, and for economy of thought, to visualize protoplasmic sol-gel transformations as obey- ing a common set of rules, this is by no means necessarily true (see Kopac, 751). Care- ful distin-tion must be made between specu- lations based on the assumption that par- ticular cellular processes behave as do known partial systems and the demonstration that such is, in fact, the case. To really understand a system such as the cortical gel-sol trans- formation or the formation of the mitotic mechanism there is only one way which, in the end, will suffice: the isolation of the individual components and the analysis of the physical chemical factors involved. This may be a discouraging point of view for the enthusiast who would seek a simpler ap- proach by a study of the system in cells them- selves—a field in which much valuable work still remains to be done. However, the prob- lem should not be regarded as insuperable: it will certainly yield when attacked with persistence and patience by modern micro- methods of fractionation and physical chem- ical analysis. Contractility is a property of most gelled 5) systems of biological materials. The contrac- tion may be isodiametric, as in the syneresis of many gels in which the constituent par- ticles have little or no preferred orientation. Anisodiametric contraction presumes prefer- ential orientation. The precise mechanism of such contractility has not yet been clearly demonstrated even in the much studied case of muscle, in which the proteins may be obtained in kilogram amounts. In recent years emphasis has been placed on the view that the protein polypeptide chains them- selves contract to form a configuration more highly folded or helically coiled than is char- acteristic of the chains in the uncontracted state. However, the alternative view, pro- posed half a century ago, that the process involves a change in orientation of the par- ticles without change in their internal or- ganization, has received strong support re- cently (Huxley and Hanson, 754). Space permits mention of only a few illus- trative cases of intracellular sol-gel trans- formations which are of unquestioned physio- logical and embryological significance. Dur- ing mitosis the internal organization of the cytoplasm undergoes striking alterations leading to the formation of the spindle and asters. When tested with a micromanipulator the spindle is found to be a fairly stiff gel capable of being moved about as a semi- rigid structure (Chambers, 751). The isola- tion of the gelled mitotic apparatus from fragmented cells recently reported by Mazia and Dan should prove very valuable in work on this subject. Although the fibrous par- ticles composing the spindle are oriented, as shown by their positive birefringence, they must be extremely thin (probably less than 100 A). The robust fibers in electron micro- graphs of sections shown in certain published work are almost certainly aggregates due to the action of the fixative. The region of cytoplasm lying immediately below the plasma membrane and having varying thickness (one to several micra de- pending on the cell type), is usually in a gelled state and is known as the cortex or cortical gel. From birefringence and EM data one may suppose that this region con- tains very fine threadlike particles, having strong interaction with each other and hav- ing orientation predominantly parallel with the surface. Lipid molecules in the cortex are oriented with paraffin chains normal to the surface plane. Although little is known about the com- position of the cortical gel, important prop- 56 erties have been attributed to it. One of these *is concerned with the mechanism of ameboid movement and perhaps of cell move- ment generally. As DeBruyn (’47) points out, current theo- ries of ameboid movement again stress the contractility of the plasmagel as fundamental to the movement. It is also frequently sup- posed that protein passes from cortical gel to endoplasm at the “tail” of the ameba and from endoplasm to cortical gel in the ad- vancing pseudopod. It has further been sug- gested that the cortical gel contains fibrous proteins analogous to the actomyosin system of muscle and that contraction may depend upon the interaction of such proteins with ATP. According to Goldacre and Lorch (’50), injection of ATP into ameba causes contrac- tion and liquefaction of the cortical gel; this liquefied gel is squeezed forward to form more gel on the surface of the advancing pseudo- pod. Kriszat (50) found that ATP causes the ameba Chaos chaos to contract, presumably because of an increase in the rigidity of the “round cytoplasm.” Lettré (752) suggested that the stage of contraction of the cortical gel depends on the ATP level and thus on cell metabolism. It is interesting that Loewy (52) has demonstrated the presence of an actomyosin-like substance in a myxomycete plasmodium. Goldacre (’52) speculates that the fibrous protein chains of the cortical gel upon con- traction fold and remain in this condition in the plasmasol; when these particles reach the front of the advancing pseudopod the chains again unfold to form cortical gel. For such speculations there is as yet little direct evidence. From the ultrastructural, and possibly from the chemical, viewpoint the process causing ameboid movement may be fundamentally similar to muscle contraction, e.g., localized changes in affinity of fibrous proteins under the influence of ATP, together with changes in configuration or relative positions or orien- tations of protein particles. In muscle the fibrous proteins are highly oriented, mak- ing possible rapid reversible and anisodia- metric contraction. In the ameba the fibrous particles are apparently oriented with long axes predominantly in planes parallel with the surface but the structural organization is of low order, making contraction relatively slow and uncoordinated. Lewis (’47) has invoked the “contractile tension” of the cortical gel to explain changes in cell configuration and movements occur- ring in embryogenesis. Equatorial constric- CELLULAR STRUCTURE AND ACTIVITY tion resulting in cell division is also attrib- uted to contraction of the cortical gel (Marsland, 51). Many other cases might be cited in which investigators have considered the cortex to be contractile. Several observations may be pertinent on this matter. Although really very little de- tailed knowledge exists about the ultrastruc- ture and composition of the cortical gel, the inference that it contains a lattice of very thin protein filaments seems justified. Such a system may well exhibit contractility. How- ever, the contractile phenomena attributed to this gel would seem to require specific orientation of the fibrous components. For this there is little evidence except the polar- ization optical indication that the anisodi- ametric protein components may lie pre- dominantly in planes parallel to the cell surface. Finally, direct physical measure- ments of force generated in the cortex at times when contraction is supposed to occur are either lacking or unconvincing. © This is said not to reflect skepticism about the contractility of the cortex but rather to emphasize the desirability of a direct physi- cal or physical chemical investigation of the properties of this important region of the cell. Other instances of intracellular contrac- tility for which no mechanism has yet been demonstrated might be mentioned. One thinks, for example, of the movement of pig- ment granules in melanophores. Under cer- tain physiological conditions the granules move out into the cell processes while under other conditions they move into the center of the cell. The movement in or out may be produced by drugs which cause the contrac- tion or relaxation, respectively, of smooth muscle. A preliminary unpublished investi- gation in this laboratory by J. B. Finean failed to reveal a fibrous structure resolvable under the conditions with the EM. Similarly the mechanism of cyclosis in plant cells and the rhythmic contraction of slime molds (Seifriz, ’43, Loewy, ’49) remain challenging subjects for investigation with modern micro- methods. THE CELL MEMBRANE The limiting envelope or cell membrane, though representing but a very small frac- tion of the cell volume, is a highly critical structure because so many aspects of cell function depend upon it. Being only rela- tively few molecules in thickness, contigu- ous with the cortical cytoplasm on the inner CELL CONSTITUTION side and the environment on the outer side (which may include connective tissue con- stituents, “cement” substances and _ other poorly defined materials), it has been diffi- cult to obtain reliable evidence of a direct nature as to its composition, structure and function. Many of our concepts originated from indirect evidence, chiefly from the lore of permeability studies, from a consideration of the limiting envelope of the mammalian erythrocyte (which may lack important as- pects of the membrane of tissue cells) and from a consideration of the properties of thin surface and interfacial films as studied by physical chemists. For detailed discussions of the literature on the cell membrane see Ponder ('48), Waugh (’50), and Davson (’51). It will be our purpose in this section to draw attention to a few aspects of the subject which may be particularly significant for the student of erowth and development and to pose a few problems worthy of further investigation. First, to what extent is the limiting en- velope or cell membrane an entity having characteristic ultrastructure, composition and function and to what extent does it repre- sent merely a non-specific interfacial struc- ture—a repository of all the various mate- rials merely adsorbed upon it from the underlying cytoplasm and external environ- ment? Unfortunately, direct evidence is meagre; one can only present a point of view with the hope that it may stimulate further investigation. The polarization optical evidence suggests that the ultrastructure of the cell membrane, the nuclear membrane and the membrane surrounding cytoplasmic vacuoles have the common property of being composed of pro- tein layers thin with respect to the wave- length of light. The membranes show nega- tive uni-axial form birefringence with optic axes normal to the plane of the membrane. Estimates of the thickness of the red cell en- velope range from 50 A (Hoffman and Hillier, 52) to several hundred Angstrom units (Waugh and Schmitt, ’40) depending on the method used. Since this includes the lipid components it is obvious that the pro- tein layers must be very thin. The polariza- tion optical evidence gives no clue as to whether the protein or the lipid component is external, whether they are interleafed or arranged in a mosaic. It is consistent with the view that the protein may be a meshwork of very thin filaments lying in the plane of the membrane. Rather ill-defined filaments have been observed with the EM in fixed Sy preparations of red cell envelopes by some authors, although Latta (52) found the sur- face free of discontinuities within the reso- lution of his preparations (60 A). No signi- ficant fine structure has thus far been ob- served with the EM in the plane of the cell membrane of tissue cells.* In the red cell envelope and in the cell membrane of several other types of cells which have been studied, the lipid molecules are oriented with paraffin chains normal to the plane of the membrane. The intrinsic birefringence is uniaxially positive with optic axes perpendicular to the plane of the membrane. It seems probable that the lipids occur as bimolecular leaflets of mixed lipids (chiefly phospholipids, galactolipids and steroids). However, there is little direct evi- dence for or against the view that such lipid bimolecular layers are continuous over the surface of the cell—a matter of considerable importance in permeability theory. The proteins of the red cell envelope, though still poorly understood, seem charac- teristic of this type of cell. Chief among these is stromatin, which has not yet been isolated in pure form but which appears to have similar amino acid composition in a variety of mammalian species (see Ponder, *48). Evidence has been presented by Moskowitz et al. (’50, 52) for another protein, “elinin,” which is said to be an elongate macromole- cule and to contain the Rh antigens. No doubt other proteins, present in relatively small quantities in the red cell envelope, remain to be discovered. However it seems probable that the bulk of the protein moiety is composed of complex molecules which are characteristic of this envelope and which re- semble each other in various animal forms (as do the muscle proteins). Almost nothing is known about the pro- teins of the membranes of embryonic cells and of various tissue cells. One good reason for this is that, unlike the mammalian eryth- rocyte, it is very difficult to separate the limiting envelope of tissue cells from the re- mainder of the cell material. It is therefore impossible to say whether there is a class of proteins or macromolecular complexes which is characteristic of the cell membrane gen- erally. It is also impossible to say whether * On the other hand, considerable structure has been observed in the nuclear membrane. In the case of the amphibian egg, Callan and Tomlin (50) found that the nuclear membrane consists of an outer porous layer and continuous inner layer. The pores in the outer layer have a diameter of 400 A and are regularly arranged. 58 the protein complex of the membrane re- sponds dynamically to changes in chemical environment, for example to the presence of ATP. This possibility may repay inquiry, for a dynamic system of this sort might well be capable of altering permeability (through change in state of aggregation of the protein fabric of the membrane) in response to physiological changes in the cytoplasm. In such considerations, it is of course difficult to rule out the possible role of the cortical cytoplasm immediately underlying the mem- brane, for it too contains macromolecular protein complexes which are known to be re- sponsive to changes in the environment. That the membrane is indeed highly re- active is shown by the fact that the per- meability for sodium ions, which must de- pend in some way on the molecular lattice of the membrane, may be increased several hundred fold in a ten-thousandth of a second in response to the passage of current. This has been clearly demonstrated in the mem- brane of nerve and muscle fibers; it might also be demonstrable, with different time constants, in other types of cells if adequate techniques were available. The surface membrane is intimately con- cerned with the establishment of the char- acteristic electrolyte pattern and with bio- electric phenomena which occur at this dis- continuity. The facts in this area of physiol- ogy bear importantly on the concept of the surface membrane as a structure in dynamic relationship with metabolic reactions in the underlying cytoplasm. For example, certain ions, such as Na’*, are present in much lower concentration inside the cell than outside. This concentration gradient is maintained by an ion “pump” capable of moving Na’, which enters the cell at a low rate, back across the membrane into intracellular space (see Steinbach and Moog in this volume, Section III, Chapter 2). The energy necessary for this osmotic work derives from metabolic reactions but the coupling of the reactions with membrane structure is not known. One possibility is that the “pump” may be located in the mem- brane itself, the ability of a carrier molecule to combine with or release Na* being deter- mined by its reaction with molecules in- volved in intermediary metabolism. A redox carrier mechanism proposed by Conway (751) encounters certain theoretical difficulties but serves to illustrate the idea (see also Ussing, 49; Rosenberg and Wilbrandt, °52). Bio- electric phenomena depend upon such ion regulatory mechanisms, as do ion secretory CELLULAR STRUCTURE AND ACTIVITY processes such as the secretion of hydro- chloric acid, Cl, etc. The phenomena which occur at or near the surface of the egg cell when activation (natural or artificial fertilization) occurs are very complex (see Runnstrom, 749). In a not too literal sense they resemble the ac- tivation of the irritable membrane in nerve and muscle though of course the structural and chemical impedimenta of this cytoplas- mic system differ markedly from those of nerve and muscle. Enzymes form an important part of the chemical complement of the surface mem- brane. Cholinesterase, catalase and several other enzymes occur in the red cell envelope. Alkaline phosphatase has been demonstrated in the surface of the cells of the intestinal epithelium and of the proximal convoluted tubules of the kidney. A variety of hydro- lytic enzymes concerned with phosphate and sugar metabolism (phosphatase, invertase, lactase, sucrase, trehalase and ATP-ase) are thought to be located at the surface of yeast cells; enzymic phosphorylation of sugar is generally believed to be necessary for en- trance of the sugar into the cell (cf. Roth- stein, Meier, and Hurwitz, *51; Brown, ’52; Rosenberg and Wilbrandt, *52). Living cells are usually resistant to the action of tryptic enzymes. This may be due in part to the presence in the cell surface of certain polysaccharides which are powerful tryptic inhibitors (Runnstrém, 49). Some difference of opinion exists as to whether the cell membrane is in fact an or- ganized, complex molecular lattice as indi- cated in the above discussion or whether it is merely surface film which forms by ad- sorption of solutes from the underlying cyto- plasm. It has long been known that when the cell wall is ruptured, a film forms very quickly over the naked protoplasm and this film may have some of the semipermeable characteristics of the normal membrane (Naegeli). Chambers points out that Ca** is necessary for this reaction. However, pro- ponents of this view must demonstrate that such a spontaneously formed interfacial film has more of the properties of the normal cell membrane than mere water immiscibility and impermeability to certain colloidal dyes. On the other hand, too much preoccupa- tion with the “fixed” lattice of the mem- brane at the expense of study of the dynamic aspects is also undesirable. We have stressed certain of these dynamic aspects above. An- other aspect of the molecular ecology (to borrow a phrase from Paul Weiss) of the CELL CONSTITUTION cell membrane is the incorporation into the film of molecules which have diffused from the cytoplasm to the region of the surface film. Experiments with monofilms show that molecules from the subsolution may readily penetrate the film, increasing the film pres- sure. Since the tension at the surface of the cell is very low, sensibly zero, penetration from the subsolution may increase the sur- face area and possibly the shape of the cell. This was shown in model systems by Lang- muir and Waugh (738). Conversely, mole- cules in the surface film may be ejected either by increase in surface pressure or by a decrease in their affinity for film molecules. The degree to which forces in the surface film may determine the shape of cells, es- pecially free cells, cannot yet be accurately assessed. The matter will be considered more in a subsequent section. An important factor not thus far men- tioned is the surface charge of the cell mem- brane. Electrophoretic measurements show that the net charge is negative on the sur- face of most cells. The red cell membrane is negative over the entire range in which the cell is stable (down to pH 4). This nega- tivity may be due in part to ionization of phosphoric acid groups in_ phospholipids (chiefly cephalin), although proteins and possibly also acid polysaccharides may also be involved. Dan (’47) has made electrophoretic stud- ies of the sea urchin egg treated in various ways. The surface charge is negative even at pH’s as low as 2. Fertilization is said to re- duce the negativity. Dan also studied the ef- fect of Ca**, Ce*** and other ions on the elec- trokinetic potential and the role of this poten- tial in surface adhesiveness and agglutination phenomena. Such quantitative studies, rela- tively rare in the literature, are to be encour- aged. They give information of the net charge, but not of the particular types of ionized groups in the exterior surface of the cell. While the electrokinetic potential is of importance in determining cell to cell inter- action; the entire constellation of charges and ion atmosphere must also be considered. SOME PHYSICAL CHEMICAL CONSIDERATIONS OF MORPHOGENETIC PROCESSES STRUCTURAL PATTERNS AT VARIOUS LEVELS OF ORGANIZATION Chemical analytical and crystallographic data, as well as biological properties, support the view that there is very precise regularity 59 of structure in protein molecules, reaching to the atoms themselves. To understand this regularity of pattern, it is mecessary to sup- pose that each native protein molecule is composed of a specific number of amino acid residues arranged in a specific sequence of residues or residue types and that the specific configuration of the polypeptide chains is that which has maximum stability under any particular conditions. We are still far from an understanding of the mechanisms by which protein molecules are formed. Much work is currently being done on the biosynthesis of peptides utilizing energy-coupling reactions revealed in recent years. However, the process by which the specific sequences of amino acid residues are joined and the chains characteristically folded remains a matter of speculation. Precise patterns of organization exist also at the level of the giant macromolecular com- plexes. Illustrative are the fibrous proteins (Table 1) which manifest axial periodic structure so regularly repeating as to give dozens of orders of x-ray diffraction. This regularity depends in turn upon a precise sequence of amino acid types along the chains, giving rise to alternating regions of relative order and disorder in the adjacent chains which form the fibrils. These regions are thought to correspond to the bands, or cross-striations, seen in the EM. Important light on the processes by which such fibrous patterns are formed is thrown by experiments made some years ago by Nageotte and Fauré-Fremiet, in which it was demonstrated that collagen fibers may be dissolved in dilute acid and reconstituted by neutralization or addition of salt. The re- constituted fibrils have the same _ period (ca. 650 A) and fine structure as the native fibrils, as seen in the EM (Schmitt, Hall and Jakus, ’42). Appropriate adjustment of ionic strength and pH causes the dispersed, solvated chains to aggregate again in perfect register with respect to the axial discontinuities. When serum acid glycoprotein or certain other substances are added to the acid solu- tion of collagen, dialysis vields fibrils with a new axial period (Highberger, Gross and Schmitt, ’51) several times greater (2000 to 3000 A) than that of native collagen. Ap- parently the added substance combines with the collagen chains to produce a new pattern of structure. Bizarre two-dimensional pat- terns have been observed in which grids were formed by the intersection of long- spacing fibrils radially directed from several centers, 60 An equally striking reconstitution of a new fibrous repeating pattern has recently been accomplished by Hodge (’52). The fibrous protein, paramyosin, of the adductor muscles of the clam has an axial repeat pattern of 145 A with a main period five times this value, or 725 A (Hall, Jakus and Schmitt, °45). The paramyosin fibrils dissolve in di- lute acetic acid. Reconstitution of the native- type structure has not yet been accomplished. However, upon increase in ionic strength, a new fibrous form was obtained by Hodge which has a repeating period of about 1400 A. This pattern, never observed in nature, has intraperiod band structure strikingly similar to that of skeletal striated muscle; the coun- terparts of the A, J, Z, M, H and N bands are all represented. However, the axial repeating period is only one-twentieth to one-thirtieth that of striated muscle. Whether this similar- ity is merely coincidental remains to be determined. It is possible that systems which form such highly ordered structural patterns in vitro are, in fact, multicomponent systems in which small amounts of non-fibrous material are required to integrate the protein chains in particular structural arrays. At any rate, the systems are chemically complex and little is as yet known about the conditions re- quired to produce the patterns. Two new patterns have already been obtained (colla- gen and paramyosin) and it may be expected that more will be found when reconstitu- tion studies are made of other fibrous proteins. The results described above suggest a point of view regarding the mechanism by which fibrous and possibly other types of structural patterns are formed in cells. This may be illustrated in the case of muscle. In the premorphological stages of differentia- tion the proteins and other components are synthesized by the cell but have not yet been fashioned into the form characteristic of the differentiated cells. The presence of myosin and actin in predifferentiation stages of muscle has been demonstrated by Her- mann and Nicholas (’48). When the physical chemical environment of the cell is favorable, the components of the fibrous pattern may “crystallize” out spontaneously in a manner similar qualitatively to that observed in in vitro experiments. The pattern may form in several stages. Thus in muscle the first stage of morphological differentiation is the for- mation of unstriated fibers which show posi- tive birefringence. It is not known whether, at this stage, the fibers have an axial perio- CELLULAR STRUCTURE AND ACTIVITY dicity of 400 A, as do fully differentiated fibers, or whether the characteristic ratios and geometric relations of myosin and actin have already been achieved. Formation of the banded structure (Z, A, J, M and other bands) follows in a sequence and manner which is not yet fully understood. The finally differentiated structure shows a high degree of regularity of axial repeating pattern. As many as four orders of diffraction, repre- senting the sarcomere length included be- tween Z bands, have been observed with visible light (Buchthal and Knappeis, ’40). How can such linear repeating patterns, hav- ing periods as large as 3 to 15y and in a few cases very much larger, be produced? It seems probable that the pattern-forming po- tentialities reside in the fibrous system itself and that the process is spontaneous as in the in vitro cases. Any other explanation would require that some kind of equivalent regular structural discontinuities which direct the process preexist in the cell; we would then be required to explain the origin of this intrinsic precursor pattern. The longest axial period thus far observed by reconstituting fully dispersed fibrous proteins is about 0.3 (3000 A)—the so- called collagen long-spacing. It seems reason- able to expect that when the system contains many more constituents (several fibrous pro- teins in muscle and an undetermined num- ber of other participating compounds), the emerging pattern may have much larger dimensions and be more complex. It is in- teresting to note in passing that the pattern of axial structure in striated muscle, as manifested by band characteristics, is the same whether the repeating sarcomere period is 2 » or 15 » [in the proventriculus muscles of certain marine annelids (cf. Schmidt, ’36), the period may be as much as 100 xz]. It seems improbable that the myosin and actin molecules in the different species vary so markedly in properties. It is more likely that the differences depend on other pattern- modulating substances and circumstances. It is difficult enough to attempt to inter- pret the phenomena which occur when fibrous proteins as well known chemically as collagen are reconstituted in vitro. Experi- ments with more complicated systems such as skeletal muscle would be very empirical indeed. However, it will be many years be- fore the chemistry of muscle constituents is fully known. Meanwhile, stimulating new discoveries and ideas may come from em- pirical attempts at reconstitution of muscle structure which may throw light not only on CELL CONSTITUTION muscle structure itself but also on the gen- eral conditions which determine the forma- tion of complex structural patterns in cells, tissues and organisms generally. Thus far only structureless or periodic fibers have been reconstituted in vitro. To reconstitute fibers with aperiodic axial structure, such as chrom- osomes, would seem much more difficult be- cause matching points along the chains do not occur periodically but are presumably unique for each point. Nevertheless, con- ditions may be found which would permit even so improbable a process to occur. It is not our intention to attempt to force all biological pattern formation into rela- tively simple concepts such as those dis- cussed above. The value of such suggestions has an inverse relation to the complexity of the system. One might, for example, suggest that cells themselves may be able to “crys- tallize” into structural patterns or tissues when the chemical environment is appropri- ate. The experiment, using certain types of free cells, or cells liberated from tissues or embryonic masses by the use of trypsin, seems quite feasible and may be a rewarding exercise, particularly in the analysis of fac- tors concerned in cell-to-cell interaction. It is conceivable that an important new con- cept may emerge from such experiments. However, they should in no way distract attention from the straightforward analyti- cal approach to the complex problems of morphogenetic fields and the genesis of patterns of organization. FACTORS INVOLVED IN CELL-TO-CELL AND CELL-TO-SUBSTRATE INTERAC- TION For purposes of simplification let us con- sider a somewhat idealized free cell, neglect- ing any surface coats or other organic matrix surrounding the cell. The surface will bear electric charges depending upon the dis- sociation of groups in the molecules com- posing the surface envelope. Depending on the ultrastructure and composition of the sur- face molecules the charges will have configu- rational arrays and will be both positive and negative. As indicated in the previous sec- tion, the negative charges will, in general, exceed the positive charges considerably so that the net charge will be negative. Surrounding the cell there will be an ion atmosphere consisting of ions of sign op- posite to those of the fixed charges on the surface molecules. These counter ions will be predominantly positive and their density 61 will grade out from the surface, forming a diffuse double layer, as exists around charged colloidal particles. The density and extent of the ion atmosphere will depend upon the density of fixed charges on the membrane and upon the ionic strength of the medium sur- rounding the cell. The ionic strength of the medium in vertebrate cells being rather high (between 0.1 and 0.2), the ion atmosphere will not extend far out into the medium. The ion atmosphere exists in an aqueous medium or water shell which forms a part of the fixed environment of the cell. At least so far as our idealized cell is con- cerned, cell-to-cell interaction will be gov- erned by the same laws which govern the interaction of colloidal particles. Like cells will have a similar ion atmosphere. Neglect- ing long-range forces (which may actually exist between such giant macromolecular systems), there will be little interaction be- tween cells until they approach within dis- tances equal to their ion atmosphere. There would then be a repulsion (because they bear the same net charge) unless the distribution of positive and negative charges is such as to permit a “matching,” in which case the cells would form stable aggregates. It should be pointed out that, particularly in processes of growth and development, the probability that an appreciable fraction of the cells would have the “ideal” properties assumed above is very small. Adhesion of cells to other cells and to sub- strates probably depends most importantly upon the formation of electrostatic bonds be- tween groups of opposite sign and also to hydrogen bonds; there is little clear evidence for covalent bonds. The force necessary to separate a cell from another cell or from a substrate (hence the stability of the cohesion) depends upon the number and types of linkages between the surfaces. Cells, such as mammalian erythro- cytes, may adhere strongly to a hydrophilic surface such as glass covered with thorium- conditioned stearate layers. This is a non- specific adhesion due to the presence of a large number of attractive groups per unit area in the substrate. When there is steric conformity between the molecular config- urations in the opposing surfaces the prob- ability of strong adhesion is greatly increased. Thus, if the surfaces of two cells contain a fabric of the same fibrous protein (presum- ably not combined at the surface with other substances, thus saturating outwardly di- rected bonds), these proteins might combine to form patches in the interface which would 62 represent a true union between the cells. In such highly localized regions, it would be impossible to tell which part of the surface belongs to which cells.* A possibly analogous situation may exist when cells behave ‘as though they knew their own kind.” An example is the recom- bination or self-sorting of coelenterates, hy- droids and sponges with cells of their own kind to the exclusion of heterologous com- binations (Brgndsted, °36). Gnmunologically specific cell aggregation or agglutination in- volves both steric matching and the resultant formation of stable bonds between surface molecules. Phenomena of this kind may be partially analyzed by studying the adhesion of cells containing an antibody upon a slide coated with a film of antigen suitably pre- pared. The possibility of intercellular ad- hesion by antigen-antibody like surface bonds was long ago pointed out by Weiss (741), who has developed the idea in a series of papers (see particularly Weiss, ’47, 50). In actual biological situations, materials in the intercellular medium play a signifi- cant, sometimes dominating role. The pres- ence of Ca** and other multivalent cations is particularly significant (see the reviews of Robertson, ’41, and Reid, ’43). Such ions may cause cell aggregation in the same manner in which they cause precipitation or co- acervation of colloids. They may actually bond surface molecules of adjacent cells by combination with negative charges in the apposing surfaces (as they do in built-up layers of fatty acids). Organic cations, such as histones and protamines, may form very stable bonds between negatively charged cell surfaces, such as those of mammalian erythro- cytes (Schmitt, 41). Rouleaux of red cells, presumably involving bonding by certain poorly identified hydrophilic substances, rep- resent a similar example. In both cases the bonding is so firm that considerable mechan- ical force is required to separate the cohering cells. Valuable analyses of the factors promoting cohesion of epithelial cells are those of Her- mann and Hickman (748), in which it was possible to estimate the force necessary to separate the epithelium from the under- lying stroma and to separate individual epi- * This possibility might be susceptible of experi- mental test if a type of free cell were found capable of being coated with a dispersed fibrous protein such as collagen. By appropriate manipulation of the lonic environment, it might then be possible to cause the cells to aggregate owing to the affinity of the coating collagen filaments for each other. CELLULAR STRUCTURE AND ACTIVITY thelial cells from each other. Cohesion is decreased by proteolytic enzymes, anionic detergents and high pH (> 9). It was impos- sible to arrive at a common primary cohesive mechanism for cells generally; there is great variability in tissues of different types. Connective tissue, the chief components of which have low if any antigenicity, may serve to separate cell types in embryological development, thus permitting strongly in- teracting cells to form tissue anlagen without interference from contact attractions from other adjacent cell types (Weiss, ’41). Surface interactions between cells may have an important influence on cell shape. Since the tension at the surface of cells is very low (near zero), surface interaction with other cells or with substrate may pro- foundly alter cell shape (assuming that cell volume does not change appreciably during the process). Strong intercellular bonding, causing cells to share surfaces to a maximal extent, may be expected to lead to the for- mation of tall, columnar epithelia while low interaction leads to flat, cuboidal epi- thelia (Schmitt, 41). Where cells are free to migrate, as in tissue cultures, the con- figuration of the substrate molecules or mac- romolecules may determine the shape of the cells (Weiss, °49; Weiss and Garber, 52). When the cell-to-substrate interaction is low the shape and movements of the cell are determined primarily by properties inherent in the protoplasm, especially in the cortex, favoring ameboid movement (Holtfreter, ’46, 47). High cell-to-substrate interaction would cause the cells to flatten and round up when the substrate has an isodiametric molecular organization. When the substrate has an an- isodiametric molecular organization the cells may form elongate processes because of strong interaction with elongate substrate particles having preferred orientation. Such processes lead also to directional migration of cells, depending upon the orientation of substrate particles. Apparent attractions be- tween cells over large distances may be ex- plained by such cell-to-oriented-substrate in- teractions (Weiss, °52). Some _ evidence indicates that cells may liberate organic materials which influence not only their own contact relations but also directions of migration along macromolecular pathways having preferred orientations (Weiss, ’45). Possibly optical and electron optical investi- gation of these phenomena would throw light on the mechanism of such effects. In the preceding section, it was suggested that the surface envelope, or plasma mem- CELL CONSTITUTION brane, of “typical” tissue cells may be the seat of very dynamic processes involving energy coupling with metabolic processes and possibly alterations in the configurations and interactions of the macromolecular constitu- ents which form the fabric of the membrane. Such processes may be expected to be par- ticularly significant in stages of embryo- logical development and growth. The syn- thesis or activation of enzyme _ systems (Spiegelman and Steinbach, 45; Boell, ’48) and of new proteins and macromolecular constituents exposes the surface membrane to possible penetration by such substances. If the penetration is not readily reversible and ephemeral, these substances will play a role in surface interactions of cells. MORPHOGENETIC FIELDS AND THEIR REGULATION It is not in the competence of this writer to evaluate, at the biological level, the vari- ous factors involved in the genesis and regu- lation of morphogenetic fields which lead to the orderly development of the embryo. How- ever, it may be useful to suggest a point of view regarding the analysis of such com- plex phenomena and to offer a few sug- gestions as to methods which might prove fruitful. First it is necessary to state some of the limiting conditions defining the problem. It seems clear that intimate contact between cells is a necessary condition for induction (see Weiss, 47; McKeehan, ’51). To obtain a normal induction for a particular locus the appropriate cells must be in apposition. This is made possible by the orderly movement of cells, bringing acting and reacting cells together at the proper time. What occurs between cells at this time is not known, but induction involves the specific structural and chemical properties of both the acting and reacting cells. The idea that a few hypo- thetical diffusible substances, perhaps occur- ring in gradients of concentration, may trig- ger off the complex field and regulatory processes seems to have been abandoned as fruitless. Variation in the chemical environ- ment, as by the addition of lithium, ammonia and assorted other inorganic and organic substances, affects the differentiation of in- ductors or the properties of reacting systems but these results seem to be purely empirical; there is no coherent body of chemical facts which explain the effects. All this leaves us with respect for the specificity of structure and composition ot 63 cells in close contact and for the complexity of the problem, but with no detailed analyt- ical facts from which to proceed. Useful in filling this factual vacuum are speculations and working hypotheses based primarily on observed and inferred surface interactions of cells (see particularly Weiss, 50). It seems probable that various types of cell-to-cell and cell-to-substrate interactions, some of which have been discussed in the present chapter, play a significant role in the aggregation of cells into specific tissue or tissue anlagen (see also Loeb, °45; Tyler, 47). However, the complexity of the situa- tion is seen by the necessity of invoking subsidiary hypotheses about the dynamic in- teraction of surface membrane constituents with the cytoplasm and with solutes in the environment. Almost no direct analytical physical or chemical data are at hand con- cerning these surfaces with properties which are, or can be made, specific with respect to developmental processes. The failure of the enthusiastic chemical attack of some years ago is voiced by Holt- freter (51): “. . . until more adequate bio- chemical methods of investigation are found, the burden of elucidating the problems of induction rests more upon the shoulders of the analytically minded morphologist than upon those of the biochemist.” But what would constitute more adequate biochemical and biophysical methods? How can one de- vise methods unless the problem to be solved is clearly focussed? It seems clear that an- swers are to be sought chiefly at the molecu- lar level of organization.* At this level the distinctions between morphology and chem- istry largely vanish. But what techniques will provide the “molecular spectacles” with which to discern the critical phenomena in- volved in embryonic fields and their regu- lation? The direct frontal attack on cell ultra- structure has been greatly implemented by modern techniques of electron microscopy. Important new facts may be expected to result from a systematic EM examination of embryological material in thin sections. How- ever, even EM techniques require consider- able development before it will be possible to * Many valuable investigations have been made of the relation of biochemical processes, such as high energy phosphate bond transfer (Barth and Barth, 651) and respiratory metabolism (Barth and Sze, 651), to embryonic development. However, these do not lead directly to clues to the mechanism of induction and the regulation of morphogenetic fields. 64 resolve the detailed fine structure of the cell membrane, to say nothing of transient molecular constituents within the membrane, which appear or disappear as development proceeds. And the indeterminacies of fixa- tion are always with us! There is of course no substitute for the slow, painstaking structural and chemical analysis of embryonic cells by ever improv- ing techniques. However, pending the ac- cumulation of such data, useful new data and concepts may result from consideration of partial systems along lines briefly out- lined below. Let us regard embryogenesis as a process of “crystallization” in which the constituent parts are themselves highly diverse, changing in composition and position as development proceeds. Induction of a particular structure corresponds to a crystallization of a pattern within the larger complex of the spatially and temporally integrated complex of pat- terns. How would such a view lead to fruitful experimentation? Recalling the results on the in vitro crys- tallization of fibrous macromolecular pat- terns described above, we discern certain similarities to induction at a very much lower level of organization. From an acid solution of collagen may be formed not only typical collagen fibrils but also fibrils hav- ing one-third the normal period, fibrils with no axial period, or long-spacing fibrils, de- pending on the conditions. The specificity and morphogenetic potentiality resides in the fibrous, dissolved collagen; various sub- stances when added to the system may “evoke” one or another of the patterns prev- iously described (Gross, Highberger and Schmitt, 52; Schmitt et al., 53). What forms depends upon the concentration of collagen and of non-collagenous organic material, upon the ionic strength, and upon pH and other factors which can be controlled. The reconstitution of paramyosin affords similar challenging opportunities for studying the way in which the protein molecules can in- teract despite the fact that almost nothing is known about the composition of paramyosin. May it not be possible similarly to study the ability of embryonic partial systems, pos- sibly of in vitro suspensions of individual cells or groups of cells, to form specific tissue- like aggregates under conditions in which the chemical environment of the cells is subject to reasonably strict control? Such experi- ments may involve development of special technigues and preparation of material on a rather heroic scale in order to make possible Bennett, H. S. CELLULAR STRUCTURE AND ACTIVITY all types of combinations and to permit a statistically significant evaluation of the re- sults. Under such controlled conditions (as- suming that the viability of the cells is not damaged too much) it may be possible to introduce materials—or even cells—suspected of having inductive or regulatory signifi- cance. To obtain even a single type of cellular system which is amenable to such manipula- tion may require considerable effort. How- ever, once obtained, it may yield information about factors involved in the formation of cell patterns obtainable in no obvious way from a study of the whole embryo or of parts transplanted into embryos, because under these conditions there are too many variables to permit drawing any simple conclusions directly. REFERENCES Astbury, W. T. 1951 91-24. , and Weibull, C. 1949 X-ray diffraction study of the structure of bacterial flagella. Nature, 163:280-281. Bailey, K. 1948a Tropomyosin: A new asym- metric protein component of the muscle fibril. Biochem. J., 43:271-279. 1948b Molecular weight of tropomyosin from rabbit muscle. Biochem. J., 43:279-281. Baker, J. R. 1949 Further remarks on the Golgi element. Quart. J. Mic. Sci., 9:293-307. 1950 Morphology and fine structure of organisms. Nature, 765:585-586. Barer, R. 1952 Interference microscopy. Nature, 170:29. , Holiday, E. R., and Jope, E. M. 1950 The technique of ultraviolet absorption spectros- copy with the Burch reflecting microscope. Bio- chim. Biophys. Acta, 6:123-134. Barth, L. G., and Barth, L. J. 1951 The relation of adenosine triphosphate to yolk utilization in the frog’s egg. J. Exp. Zool., 116:99-121. , and Sze, L. C. 1951 The organizer and respiration in Rana pipiens. Exp. Cell Res., 2:608- 614. Bear, R.S. 1952 The structure of collagen fibrils. Ady. Protein Chem., 7:69-154. , Schmitt, F. O., and Young, J. Z. 1937 The ultrastructure of nerve axoplasm. Proc. Roy. Soc., (London) B, 123:505-519. Bennett, A. H., Jupnik, H., Osterberg, H., and Richards, O. W. 1951 Phase Microscopy. John Wiley & Sons, Inc., New York. 1950 Microscopical investigations of biological materials with polarized light; in McClung’s Handbook of Microscopical Tech- nique, pp. 591-677. Paul B. Hoeber, Inc., New York. Bensley, R. R. 1951 Facts versus artifacts in cytology: The Golgi apparatus. Exp. Cell Res., 2: 1-9. Flagella. Sci. Amer., 184: CELL CONSTITUTION Bernal, J. D. 1949 The structure and interac- tions of protein molecules. Exp. Cell Res., Suppl., 1:15-23. , and Fankuchen, I. 1941 X-ray and crys- tallographic studies of plant virus preparations. J. Gen. Physiol., 25:111-165. Bessis, M., and Bricka, M. 1949 Nouvelles études sur les cellules sanguines au microscope élec- tronique avec une étude particuliére de leur ul- trastructure. Arch. Anat. mic. Morph. exp., 38: 190-215. Boell, E. J. 1948 Biochemical differentiations during amphibian development. Ann. N. Y. Acad. Sci., 49:773-800. Booth, F. 1953 Recent work on the application of the theory of the ionic double layer to colloidal systems. Progress in Biophysics and Biophysical Chemistry, 3:131-194. Bourne, G. H. 1951 Cytology and Cell Physiol- ogy. 2d ed. Oxford University Press, Oxford, Eng- land. Brachet, J. 1949 L’hypothése des plasmagenes dans le dévelopment et la différenciation. Publ. Staz. Zool. Napoli Suppl., 27:77-105. 1950 The localization and the role of ribonucleic acid in the cell. Ann. N. Y. Acad. Sci., 50:861-869. Brattgard, S. O., and Hyden, H. 1952 Mass, lipids, pentose nucleoproteins and proteins deter- mined in nerve cells by x-ray microradiography. Acta Radiologica Supp., 94:1-48. Brondsted, H. V. 1936 Entwicklungsphysiolo- gische Studien tiber Spongilla locustris. Acta Zool., 17:75-172. Brown, R. 1952 Protoplast surface enzymes and absorption of sugar. Intern. Rev. Cytology, 7:107- 118. Buchsbaum, R. U. 1948 Individual cells under phase microscopy before and after fixation. Anat. Rec., 102:19-35. Buchthal, F., and Knappeis, G. G. 1940 Diffrac- tion spectra and minute structure of the cross- striated muscle fibre. Skand. Arch. Physiol., 83: 281-307. Bull, H. B. 1952 The chemistry of amino acids and proteins. Am. Rev. Biochem., 27:197-208. Buschke, W. 1949 Studies on intercellular co- hesion in corneal epithelium. J. Cell Comp. Physiol., 33:145-175. Callan, H. G., and Tomlin, S. G. 1950 Experi- mental studies on amphibian oocyte nuclei. I. In- vestigation of the structure of the nuclear mem- brane by means of the electron microscope. Proc. Roy. Soc., B, 137:367-378. Caspari, E. 1950 Visible plasmagenes. Evolution, 4:362-363. Caspersson, T. 1950 Cell Growth and Cell Func- tion. W. W. Norton & Co., Inc., New York. Chambers, R. 1951 Micrurgical studies on the kinetic aspects of cell division. Ann. N. Y. Acad. Sci., 57:1311-1326. , and Kao, C. Y. 1952 The effect of elec- trolytes on the physical state of the nerve axon of the squid and of stentor, a protozoon. Exp. Cell Res., 3:564-573. Chantrenne, H. 1951 Recherches sur le mécan- 65 isme de la synthése des proteines. Pubblicazioni della Stazione Zoologica di Napoli, 23:70-86. — 1952 Acides ribonucleiques et biogenése des proteies. Symposium on the Biogenesis of Proteins, 2nd Int. Cong. of Biochem. (Paris), pp. 85-95. Chargaff, E. 1945 Cell structure and the prob- lem of blood coagulation. J. Biol. Chem., 760: 351-359. 1949 Recent studies on cellular lipopro- teins. Trans. Faraday Soc., No. 6, pp. 118-124. Chinn, P., 1938 Polarization optical studies of the structure of nerve cells. J. Cell. Comp. Phys- iol., 72:1-16. Claude, A. 1943 The constitution of protoplasm. Science, 97:451—-456. 1949 Discussion in symposium on lipo- proteins. Trans. Faraday Soc., No. 6, pp. 125-129. 1950 Studies on cell morphology and functions: methods and results. Ann. N. Y. Acad. Sci., 50:854-860. Conway, E. J. 1951 The biological performance of osmotic work. A redox pump. Science, 113: 270-273. Cooper, K. W. 1948 The evidence for long range specific attractive forces during the somatic pair- ing of dipteran chromosomes. J. Exp. Zool., 708: 327-335. Cosslett, V. E. 1951 Practical Electron Micros- copy. Butterworths Scientific Publications, Lon- don. Crawford, G. N. C., and Barer, R. 1951 The ac- tion of formaldehyde on living cells as studied by phase-contrast microscopy. Quart. J. Mic. Sci., 92:403-452. Crick, F. H., and Hughes, A. F. W. 1949 The physical properties of cytoplasm. Exp. Cell. Res., 1:37-80. Dan, K. 1947 Electrokinetic studies of marine ova. V. Effect of pH-changes on the surface po- tentials of sea-urchin eggs. Biol. Bull., 93:259-266. Davies, H. G., Engstrém, A., and Lindstrém, B. 1953 A comparison between the x-ray absorp- tion and optical interference methods for the mass determination of biological structures. Na- ture, 772:1041. , and Wilkins, M.H.F. 1952 Interference microscopy and mass determination. Nature, 169: 541, Davson, H. 1951 A Textbook of General Phys- iology. The Blakiston Company, Philadelphia. DeBruyn, P. P. H. 1947 Theories of amoeboid movement. Quart. Rev. Biol., 22:1-24. Doniach, A. H. 1953 Autoradiography. Progress in Biophysics and Biophysical Chemistry, 3:1—26. Drummond, D.G. 1950 The practice of electron microscopy. J. Roy. Mic. Soc., 70:1-141. Edsall, J. T. 1954 Pasadena conference on the structure of proteins. Science, 719:302-305. Engstrém, A. 1950 Use of soft x-rays in the as- say of biological material. Progress in Biophysics, 1:164-196. , and Lindstrom, B. 1950 A method for the determination of the mass of extremely small biological objects. Biochim. et Biophys. Acta, 4: 351-373. 66 Eyring, H., Johnson, F. H., and Gensler, R. L. 1946 Pressure and reactivity of proteins, with particular reference to invertase. J. Phys. Chem., 50:453-464. Fawcett, Don W., and Porter, Keith, R. 1952 A study of the fine structure of ciliated epithelial cells with the electron microscope. Anat. Rec., 113-33) Fernandez-Moran, H. 1952 The submicroscopic organization of vertebrate nerve fibres. Exp. Cell Res., 3:1-83. Ferry, John D. 1952 The mechanism of poly- merization of fibrinogen. Proc. Nat. Acad. Sci., 38:566-569. French, D., and Edsall, J. T. 1945 The reactions of formaldehyde with amino acids and proteins. Adv. Protein Chem., 2:278-336. Frey-Wyssling, A. 1948 Submicroscopic Mor- phology of Protoplasm and Its Derivatives. Else- vier Publishing Co., New York. Gatenby, J. Bronte, and Moussa, T. A. A. 1951 The nature of the Golgi apparatus. The liver cell and the Palade-Claude mash cytology. La Cellule, 54:49-64, Gersh, I. 1949 A protein component of the Golgi apparatus. Arch. Pathol., 47:99-109. Gjessing, E. C., Floyd, C. S., and Chanutin, A. 1951 Studies on the proteins of the particulate- free cytoplasm of rat liver cells. J. Biol. Chem., 188:155-165. Glick, D. 1949 Techniques of Histo- and Cyto- chemistry. Interscience Publishers, New York. , Engstrom, A., and Malmstrém, B.G. 1951 A critical evaluation of quantitative histo- and cytochemical microscopic techniques. Science, 114:253-258. Goldacre, R. J., 1952 The folding and unfolding of protein molecules as a basis of osmotic work. Int. Rev. Cytol., 7:135-164. , and Lorch, I. J. 1950 Folding and un- folding of protein molecules in relation to cyto- plasmic streaming, ameboid movement and os- motic work. Nature, 166:497-500. Gomori, G. 1950 Pitfalls in histochemistry. Ann. N. Y. Acad. Sci., 50:968-981. Gortner, R. A. 1949 Outlines of Biochemistry. John Wiley & Sons, New York. Grigg, G., and Hodge, A. 1949 Electron micro- scopic studies of spermatozoa. Australian J. Sci. Res., 2:271-286. Gross, J., Highberger, J. H., and Schmitt, F. O. 1952 Some factors involved in the fibrogenesis of collagen in vitro. Proc. Soc. Exp. Biol. Med.., 80:462-465. Hall, C. E. 1949 Electron microscopy of fibrin- ogen and fibrin. J. Biol. Chem., 179:857-864. 1953 Introduction to Electron Micros- copy. McGraw-Hill Book Co., Inc. New York. , Jakus, M. A., and Schmitt, F. O. 1945 The structure of certain muscle fibrils as revealed by the use of electron stains. J. App. Phys., 76: 459-465. Hamley, D. H., and Sheard, C. 1947 Factors in fluorescence microscopy. J. Opt. Soc. Amer., 37: 316-320. / Huxley, H., and Hanson, J. CELLULAR STRUCTURE AND ACTIVITY Hawn, C. V. Z., and Porter, K. R. 1947 The fine structure of clots formed from purified bovine fibrinogen and thrombin: A study with electron microscope. J. Exp. Med., 86:285-292. Heilbrunn, L. V. 1952 The physiology of cell division; in Modern Trends in Physiology and Biochemistry, edited by E. S. Barron, pp. 123-134. Academic Press, Inc., New York. Hermann, H., and Hickman, F.H. 1948 The ad- hesion of epithelium to stroma in the cornea. Bull. Johns Hopkins Hosp., 82:182-207. , and Nicholas, J. S. 1948 Quantitative changes in muscle protein fractions during rat development. J. Exp. Zool., 107:165-176. Highberger, J. H., Gross, J., and Schmitt, F. O. 1951 The interaction of mucoprotein with solu- ble collagen; an electron microscope study. Proc. Nat. Acad. Sci., 37:286-291. Hillarp, N., and Olivecrona, H. 1946 Structural proteins and oriented lipoids in the cytoplasm of secreting and resorbing epithelial cells. Acta Anat., 2:119-141. Hirschler, J. 1943 Grundsatzliches tiber Osmi- umfixierung und Osmiumfarbung. Z. wiss. Mik., 59:113-130. Hodge, A. 1952 (with energy packet) Oxidized substrate + reduced cofactor (with energy packet) Sugar, to feed the mechanism, may enter from the environment or from endogenous stores such as glycogen. In either instance, stages to be discussed CELLULAR METABOLISM FRUCTOSE +ATP 73 GLYCOGEN sighs GLUCOSE +ATP Fructose-6- phosphate =Glucose- 6-phosphate> Glucose- 1-phosphate ATP Yy Fructose-1,6-diphosphate Ww a Triosephosphate-+DPN-5Flavoprotein I ?5Cyt. CSCyt. ox.50 Inhibited by IAA Ai Phosphoglyceric acid Inhibited by NaF Phosphopyruvic acid 2 Nv Pyruvic acid ———————__¥ ? —_—_____», Cytochrome system>0., Vv Acetyl > [Fen Oxalosucciniec acid Inhibited . by arsenite Inhibited by malonate ? F Fumaric acid Malic VY Oxaloacetic acid Vv Isocitric acid 4+ TPN Flavoprotein II Cyt. C5Cyt. uy a-Ketoglutaric acid———y ? —_______» Cyt. C Cyt. Succinic acid ———__—____5 ? _______+ Cyt. C—Cyt. acid ——3 DPN >Flavoprotein I ?SCyt. C3Cyt. Ox. 05 ox.> 0, ox.—70,, 50 OX. > Fig. 4. Intermediate catalysts involved in the transfer of hydrogen and electrons to molecular oxygen in the stepwise oxidation of carbohydrate in muscle. Abbreviations: JAA, iodoacetic acid; OAA, oxaloacetic acid; Cyt. C, cytochrome c; Cyt. ox., cytochrome oxidase. The factor of Lockhart and Potter, and Altschul et al. is referred to as Flavoprotein I with a question mark. This factor is specific for DPN. Flavoprotein II, specific for TPN, is the cytochrome reductase of Haas, Horecker, and Hogness. Whether cytochromes a and b are involved in any of the steps is not definitely known. The important discovery has recently been made that the phosphorylation of glucose by ATP to glucose-6-phosphate, catalyzed by the enzyme hexokinase, is under the dual control of hormones from the later give the sugar a preliminary working over to result in the phosphorylated form shown. Direct oxidation without preliminary phosphorylation pos- sibly may also occur (cf. Barron, ’43), but its phys- iological role is still to be established. The usual hexose for the scheme shown would be glucose. There is evidence that other sugars may also enter the system provided the appropriate enzymes are present (cf. Lardy, 50). A preliminary splitting of the 1,6-fructose diphos- phate leads to the formation of two triose monophos- phates, which, in the presence of triose isomerase, are interconvertible. The phosphoglyceraldehyde formed, in the presence of phosphate and a phos- phate acceptor (ADP), is then oxidized by phos- phoglyceraldehyde oxidase to yield phosphoglyceric acid and ATP. At this important step, two ubiquitous coenzyme anterior pituitary and adrenal cortex on one side, and insulin on the other. (From Ochoa, ’47.) systems are introduced: the pyridine nucleotide* coenzymes (cf. LePage, in Lardy, 50) concerned with hydrogen transport, and the adenylic acid systemt+ (ATP and ADP) concerned with energy packet storage and transport. For the reaction to be * The diphosphopyridine nucleotide is best known and will be referred to as oxidized (DPN) or re- duced (DPNH2) coenzyme. This is the coenzyme I of Warburg. + Adenylic acid is a nucleotide containing one phosphate group per molecule. This form (adeno- sine monophosphate, or AMP) can add phosphate by energy-rich bonds (cf. Lipmann, ’41) to form adenosine diphosphate (ADP) or adenosine triphos- phate (ATP). In most of the discussion that will follow, only the change ADP + P == ATP will be considered, although the other forms may also be concerned. 74 . carried out successfully, all components—phos- phate, triose, oxidized DPN, ADP and the appropri- ate enzymes—amust be present. In the absence of DPN the oxidation will not take place; in the ab- sence of phosphate and phosphate acceptor, on the other hand, or in the presence of a competing sub- stance for phosphate, such as arsenate, the oxidation Glucose A. & oof Glucose-6-Phosphate + Lactate, +2H SSS PROTEIN Phosphate + CO5 +Hp5 <—+HOP03 N N yy fo) ° Sy) a | ] fo) ey ; | | ain + Acetoacetate ————=. 2 3 Ss = Gl ES a Pp a Es o @o bo Gs} on qo, fy O s < 2 a Co) 13) < eo > » 3) is) < Acetyl ey cag CELLULAR STRUCTURE AND ACTIVITY After rearrangement and addition of water, phos- phoenolpyruvic acid is formed from phosphogly- ceric acid. Phosphoenolpyruvic acid contains a high energy phosphate linkage. In the presence of ADP, transphosphorylation occurs to yield pyruvic acid and ATP. Enolase, the enzyme leading to the forma- tion of the enol form, is especially sensitive to fluor- + =2H ecO2 Malate 5 A) = § -Co a 2 Oxalacetate 6 NH} Alanine Flavin (e) N N ° = 1) a5 | \ | oN re PO ° sO i=] » od < Q =) wt a Oo o 4 + cal a a » = o ov o E Lactate ANAEROBIC Fig. 6. Aerobic and anaerobic mechanisms of hydrogen and electron transfer in the oxidation of triose to the glyceric acid level in muscle. (Although not represented in this diagram, oxidation of triosephosphate- glyceraldehyde-3-phosphate involves uptake of one molecule of inorganic phosphate and leads to 1,3-diphos- phoglyceric acid.) (From Ochoa, *47.) system having been involved but with no net change in oxidized or reduced forms. In yeast cells and in some other forms, pyruvate is decarboxylated to form acetaldehyde and carbon dioxide, and acetalde- hyde is then reduced with DPNHg to alcohol. Pyruvate may also be reversibly reduced and aminated to form alanine, or condensed with carbon dioxide to form a four-carbon acid (malic acid). Of greatest interest for the general pathway, however, is the entrance of pyruvate into the so-called Krebs— Szent-Gyoérgyi cycle (citric acid cycle, etc.). The exact details of the entrance of pyruvate into this cycle are not known, but the condensation of the three-carbon pyruvate with a four-carbon acid re- sults in the formation of a six-carbon citric acid derivative of some sort, plus carbon dioxide. A hypo- thetical seven-carbon ‘“procitric acid” may be formed (cf. Green, in Lardy, ’50), or a preliminary details of the metabolism of the various components that may enter this cycle. It is here that many fatty acids and other fuels may come into the common scheme. | The enzyme system catalyzing this truly cyclic process has been variously named. A useful single term is cyclophorase (cf. Green, in Lardy, *50). As yet, although individual reactions may be studied, the complete system remains associated with the insoluble particles which are collected by centrifugation of homogenates of cellular systems and which are apparently closely allied to the ribose-nucleoprotein components (cf. Green, in Lardy, ’50). At least parts of the system are widely distributed (cf. Barron, 43). In its complete form both decarboxylation and oxidation are involved. In large part the system is reversible, a fact of consid- erable interest in view of carbon assimilation by a * Since this paper was written, the accumulation of evidence has clearly decided in favor of the second of these alternatives. It now appears that in the pres- ence of DPN and coenzyme A, pyruvate is split to yield reduced DPN, COe2, and acetyl CoA. In the presence of a condensing enzyme, the latter sub- stance transfers its acetyl to oxalacetate, thus form- ing citrate. Acetyl CoA can also be generated from acetaldehyde, fatty acids, B-keto fatty acids, and acetate (cf. Ochoa and Stern, 52). 76 wide variety of organisms (cf. Werkman and Wood, 42). Parts of the system, such as the reversible oxidation of succinate to fumarate, may function in oxidations in other steps. The succinic fumaric sys- tem (catalyzed by succinic dehydrogenase) is notable in part because of the relative specificity by which the competitive inhibitor, malonate, may be used (cf. Krebs, ’43). Green (in Lardy, 50) notes that apparently there are single enzymes in cells that perform individual reactions of the cycle with- out, however, being necessarily a part of it. While the Krebs—Szent-Gyérgyi cycle may be essentially reversible, it is also true that the net di- rection of flow will result in decarboxylation and oxidation if a source of reduced carbon chains is continually being fed in (pyruvate, for example) and if oxygen is available. The disposition of car- bon dioxide need not be considered here, but it should be noted that a supply of oxidized coenzyme must be present in order for the cycle to keep going. Since the total absolute amount of coenzyme per unit volume of cell is small, a regenerating system to provide the oxidized coenzyme must be present. Otherwise the over-all reaction would fail even in the presence of carbon chain substrate, oxygen, and enzyme protein. By links that are not completely known at present (cf. Potter, in Lardy, *50), the major regeneration of reduced coenzyme is accom- plished in aerobic cells by the cytochrome system, together with the large number of other enzyme systems that involve the reversible oxidation of DPNHp: or TPNHoe. With the cytochrome system, an elegant method is provided for the excretion of the hydrogen of the original carbon chain food by transferring electrons to the final acceptor oxygen, which then is free to unite with the protons left by electron removal from hydrogen. The review by LaValle and Goddard (’48) gives an excellent back- ground for understanding the oxidative steps medi- ated by the cytochrome system. The diagram of Figure 6, taken from Ochoa (47), illustrates one way in which linkage of the DPN system with the cytochrome system may take place. With heart muscle systems it has been reported (Ochoa, 47) that the over-all reaction pyruvate + 2.5 Og —» 3 COs + 2 H20 yields fifteen labile phosphate bonds. While not all of the steps are known it is reasonable to assume that throughout the oxidative cycles the two coen- zyme systems, pyridine nucleotide and adenylic acid, are so linked that the energy packets released in each oxidative step are stored for further use by the phosphorylating coenzyme system or its equiv- alent. A general summary of the fate of pyruvate in cellular metabolism is given by Barron (752) and is schematized in Figure 5, which is taken from his paper. Carbohydrate metabolism, as a model sys- tem, may now be summarized even more briefly than has been done in the description of Figure 3. Sugar may possibly be oxidized directly, but the main pathway known in- volves a preliminary phosphorylation to a CELLULAR STRUCTURE AND ACTIVITY diphosphate form. The breakdown of this to pyruvate (three-carbon) involves an oxida- tive step via DPN=DPNHe and an acidic enol formation, both of which, in the pres- ence of the adenylic acid coenzyme system for phosphorylation, result in the formation of energy-rich phosphate bonds. Pyruvate, the common focal point, may be reduced with DPNHz to lactic acid or may enter the oxi- dative cycles of the six-, five-, and four- carbon acid systems and the cytochrome sys- tem, again using DPN and the adenylic acid systems. The ubiquity of the two coenzyme systems has already been remarked. Their importance as possible controlling factors needs further emphasis. Each system is typically present in more or less constant amount. That is, there is little net change during short time intervals, the systems acting in catalytic con- centrations. So long as it is possible for the DPN system to be reversibly reduced, or for the adenylic acid system to be reversibly phosphorylated, orderly reactions can pro- ceed. Should either system be fixed in one state, however, far-reaching changes might, and do, take place. For example, in the Pasteur reaction, where fermentation is in- hibited in the presence of oxygen, the DPN system has been implicated (cf. Lardy, ’50). Under aerobic conditions where the revers- ible system DPN=DPNHe might be far to the left, there would be insufficient reduced coenzyme to form lactate (or alcohol) from pyruvate, and the rate of fermentation would be correspondingly slowed down. Meyerhof (49) has shown that the stabilizing of the adenylic acid system as ATP can result in decreased fermentation of yeast extracts, a condition that can be explained by the failure to provide ADP as necessary phosphate ac- ceptor for some of the early stages of glycol- ysis. Addition of an enzyme converting ATP to ADP allows the reaction to proceed. These considerations should be kept in mind in experiments where metabolism is tampered with, by adding either poisons or excess substrates. For example, it might be tempting to think of adding the universal substrate, pyruvate, to a cellular system. Under aerobic conditions, however, pyruvate might be oxidized rapidly with the resultant formation of excess ATP and oxidized DPN, thus throwing out of gear all of the pre- pyruvate steps of the metabolic flow line. An effect possibly like this has been noted by Goldinger and Barron (’46) with sea urchin eggs, where addition of pyruvate has little effect on the Qos even though pyruvate is CELLULAR METABOLISM metabolized rapidly. Presumably, the added pyruvate enters the system and, in effect, blocks the utilization of endogenous sub- strates. Partly because relatively few coenzyme systems are essential to many different reac- tions, the metabolism of the cell represents a closely integrated system in which inter- ference with one chemical step may have profound results on some remote part of the living system. The coenzyme systems repre- sent not only important energy pools for the coupling of a variety of reactions, but also coordinating mechanisms. ENERGY-UTILIZING SYSTEMS The burning of food results, in the cell, in the formation of low-energy* waste prod- ucts from high energy precursors. In analogy with the more obvious engines of industry, the living machines must have mechanisms for converting the energy of burning to use- ful functions, unless the whole process is to result in waste heat. To use terms common in physics, cells must have transducer mech- anisms. There are no apparent structures in cells that could function as heat expan- sion engines. It is possible that electrical work may be done and keyed directly to metabolism through the existence of oriented electron transfer systems, e.g., Fe-oxidases oriented at surfaces (cf. Lundegardh, 45), but this has yet to be demonstrated. The only method by which oxidative energy is known to be made use of in quantitatively important amounts is by the transfer of chemical energy directly to other reactions or to devices for producing movements of fluids and cells. SYNTHESIS In order for the energy of one reaction, e.g., a step in the oxidative cycle, to be used for driving a synthetic reaction, the two must have a common reactant (cf. Johnson, in Lardy, *50). Presumably a common reac- tant could take a variety of forms. It is a matter of considerable interest that relatively few such common reactants are known. The best understood, which are implicated in many cellular functions, are the members * For obvious reasons the details of the energetics of metabolism cannot be given here and the term “energy” must be used rather loosely. The reader is referred to Clark (’49) or to the chapter by John- son in Lardy (50) for further definitions of the terms used. 77 of the phosphorylation coenzyme system, ADP and ATP. Lipmann (741) has discussed in a very readable fashion the formation of energy- rich phosphate bonds. In general the proc- ess may be illustrated by the oxidation of an aldehyde (such as the glyceraldehyde of Fig. 3) to an acid (glyceric). In the test tube, in the presence of water, this presum- ably takes place with the formation of an unstable hydrate and the subequent removal of hydrogen to give the terminal carboxyl group instead of the aldehyde. This process is essentially irreversible and the energy yielded is lost as heat. In the presence of orthophosphate, however, water is replaced by phosphoric acid and the oxidation then yields an acid anhydride. The link between the phosphoric acid and the organic acid is called the energy-rich phosphate bond. Lip- mann (741) may be consulted for the further specification of energy-rich linkages. The oxidation of the aldehyde in the pres- ence of phosphate results in a loss of heat per unit of hydrogen transferred less than that during oxidation following hydration, and the reaction is reversible. Of greatest im- portance to the present discussion is the fact that two groups, phosphate and organic acid, have been linked together, a synthesis has taken place, and the groups so attached can now be shifted around by a variety of en- zymes in cells. Especially, the labile phos- phate can be shifted by transphosphoryla- tion to such phosphate acceptors as creatine, arginine and the adenylic acid system with the retention of the “high-energy” character- istics, or from the adenylic acid system to sugars or other forms to give ordinary ester linkages. The formation of the bonds between phos- phate and other groups has acquired height- ened interest from the finding that the group that attaches to the phosphate becomes it- self transferable in most cases, and thus synthesis of different complex organic com- pounds becomes at least theoretically pos- sible. This sequence of events is illustrated bv outlining the synthesis of sucrose from glucose and fructose (Hassid and Doudoroff, 50). The over-all reaction is glucose + fruc- tose = sucrose. The reaction was first known as it occurred in ordinary hydrolysis where the reaction went far to the side of the split products. For many years, under the spell of the idea of reversibility of hydrolytic en- zyme action, biochemists thought that sucrose synthesis, which takes place rapidly in plants, was caused by sucrase acting in some special 78 fashion. It is now known that the synthesis is linked to other processes and must be formulated in stepwise fashion. The reaction starts with the linkage of some group to one sugar. In bacterial cells, at least, this is apparently by phosphoryla- tion, the phosphate group being furnished by ATP: Glucose + ATP = Glucose—P + ADP FATS, PROTEINS —> ETC: ENERGY POOL CELLULAR STRUCTURE AND ACTIVITY anisms to give a high energy bond; (2) transphosphorylation to an intermediate takes place; (3) transphosphorylation to a sugar occurs; (4) a “transhexose” reaction now takes place with the phosphate acting as donor rather than transferee, and the glucose is linked to another sugar, fructose, to form the end product of the synthetic reaction, sucrose. Fig. 7. Schematic flow diagram, assuming that the total activity of a cell may be regarded as a chemical manufacturing plant composed of unit processes. Preliminary stages such as glycolysis provide raw materials which are burned to COg and H20 in the citric acid and oxidase units, energy being collected in the “pool” of coenzymes and wastes eliminated. The materials of the energy pool, with other end products or raw materials, then are used for synthetic reactions or the “work of the cell.” A major purpose of the diagram is to illustrate the interrelationships of unit processes and to emphasize that control could reside either in distribution of materials (extrinsic factors of control) or in the machinery of each unit (intrinsic control). The transphosphorylation once accomplished, the ester linkage is there (roughly equivalent to a glucosidic linkage) and the phosphate can, so to speak, with the proper enzyme toss the glucose to the waiting fructose: Glucose—P + fructose = sucrose + Phosphate Since the ATP, in intact cells, is regener- ated by the oxidative uptake of inorganic phosphate, the over-all reaction in respiring cells may be written as sucrose — glucose + fructose and the catalytic components, ADP and phos- phate, have not changed in amount. The pattern of sucrose synthesis thus emerges as follows: (1) phosphate is linked to an organic molecule by oxidative mech- It is probable that this series of reactions, like those of glycolysis and oxidation, offers a general model for various types of syn- theses taking place in cells. For example, polysaccharide may be built up by trans- ferring glucose from the glucose—fructose linkage rather than from glucose—P link- age (Hassid and Doudoroff, 50). Nucleotides appear to be formed from pentose and nitrog- enous base by similar mechanisms (Kalc- kar, ’45). Once a group needed for synthetic processes is linked to any compound in proper fashion, it appears probable that the cells possess the essential enzyme systems to trans- fer the group to other compounds, the main energy expenditure being in the formation of CELLULAR METABOLISM the initial linkage. A series of “trans” en- zymes have been described (cf. Cohen, in Lardy, *50) in addition to the transphosphor- ylases already mentioned. Transaminase, transiminase, transulfurase, transmethylase have all been studied with considerable pre- cision. The dehydrogenase enzymes are essen- tially transhydrogenases and, with the DPN coenzyme system, function in a fashion per- fectly analogous to the transphosphorylases working with the adenylic acid coenzyme system. While mechanisms of protein and fat syn- thesis are not clearly understood as the car- bohydrate models, essentially similar systems have been postulated on the basis of very suggestive data (cf. Ratner, “49; Lipmann, 49; Borsook et al., ’49).* Fruton (50), in an interesting paper, suggests that cellular pep- tidase may act in building up polypeptides and proteins by transferring peptides to al- ready existing peptide chains, the process being in many respects similar to the reac- tions noted for the formation of polysac- charide by tranferring a monosaccharide from a disaccharide to an existing chain of sugar residues. THE CONTROL OF CELLULAR METABOLISM+ A living cell at work must depend upon the precise integration of a complex mosaic of active units. The examples chosen from the glycolytic and oxidative cycles illustrate both the multiplicity of the parts involved and their interdependence. In addition, it * The discussion by Borsook is especially illumi- nating with respect to the general nature of the problems considered and the careful speculations about cellular mechanisms that may be concerned with amino acid incorporation into tissues. + In this section, as in the preceding one on gly- colytic and oxidative mechanisms, we shall discuss examples rather than attempt a complete coverage of all experiments designed to elucidate control of metabolism. Much good work has been done in an attempt to analyze the relationship of physiological processes and chemical reactions, mainly by defin- ing enzyme systems in tissue breis and by deter- mining the effects of added inhibitors on functions, or the effects of addition or depletion of substrate. No single formula is yet available that will ensure that such studies are meaningful in terms of the life of the cell. The interpretation of studies which in- volve upsetting the metabolic machinery must al- ways be made cautiously against the background of the material so well set forth by Heilbrunn in his “Outline of General Physiology,” as well as the details of our knowledge of enzyme systems. ie) must be remembered that there are many special pathways associated with special cel- lular functions. The existence of multi-unit systems that function in an orderly manner implies also the existence of precise con- trolling mechanisms. Probably the most chal- lenging problem confronting the general physiologist today is that of piecing together our knowledge of fragments of the active framework of the living cell in such a fashion that not only will the energy balances hold true, but the orderliness of life will also follow. If the components of the glycolytic and oxidative systems are mixed together in a test tube, some orderly controlled reactions will take place by virtue of the sharing of reactants and reaction products of unit proc- esses of the system. For example, in the fermentation of sugar by cell-free extracts of yeast, the rate of carbon dioxide production can be controlled by the concentration of such components as sugar, phosphate, ATP and so forth. If the enzymes hydrolyzing ATP are blocked, phosphate tends to disappear as new ATP is formed; thus further phosphory- lation of sugars is inhibited and hexose mono- phosphate accumulates. Thus a “control” of sugar phosphorylation can be effected by controlling ATP dephosphorylation. This type of control, however, would have dis- tinct limitations and, in particular, it is difficult to visualize regulation of the many rather fast “off-on” shifts in metabolism in this fashion. Fortunately, there is good evi- dence that the components of the reactions are not free to mix as in the test tube, but are highly and precisely localized within the structure of the cell. Evidence for intra- cellular localization of enzymes has been accumulating over a period of years and need not be detailed here (cf. Schneider, in Lardy, 50). Ever since the work of Warburg, it has been recognized that the metal-con- taining oxidases are associated with insoluble cell fragments in a suspension of fragmented tissue. Cytochrome c appears to be reasonably soluble but the other cytochromes and the oxidase are put into solution only with great difficulty. Similarly, the cytochrome-linked dehydrogenases are separated from insoluble cell particles only with difficulty (cf. Mc- Shan, in Lardy, ’50). Green (in Lardy, ’50) discusses evidence that the enzymes of the citric acid cycle are always associated with the insoluble residue of tissue homogenates; partly on this basis he names the system cyclophorase. The enzymes primarily asso- 80 ciated with glycolysis appear, on the other hand, to be more soluble* than those asso- ciated with decarboxylation and oxidation (cf. Dixon, ’49). In addition to enzymes, many small-mole- cule substances such as ATP and acetyl- choline have been considered to be at least in part bound in the normal living cell (cf. Caspersson and Thorell, ’42). In a general fashion, it may be fruitful for the cell physiologist to regard those en- zyme systems which are fixed to insoluble units of cell homogenates as being of pri- mary importance in the regulatory activities of cells. A notable exception to this general- ization would be the apparent local ac- tivity of phosphorylating mechanisms at the cell surface (Rothstein and Meier, 48). Thus a discussion of the control of cellular metabolism should not be limited to test tube factors, but must also include factors relating to the discrete structural organization of en- zyme systems. Basically it might be said that variations in metabolic intensity are due to differences either in molar enzyme concentration or in enzyme activity. But molar enzyme concentrations are rarely known and, under the heading of enzyme ac- tivity, it will obviously be necessary to con- sider spatial orientations as well as material supply and availability of cofactors. DIFFERENCES IN ENZYME CONCENTRATION The rate at which an enzymatic reaction proceeds may be limited either by the chem- ical reactants in the usual sense, or by the catalytic system. This fact was first clearly established by Michaelis and Menten (713) in studies on the inversion of cane sugar. Re- cent work has shown that the rate of oxygen uptake may be regulated by the concentra- * The word solution is poorly defined in the bio- chemical literature. Due to the lack of precise usage of words by enzymologists, it is difficult to decide whether an “extract,” for example, means a solu- tion of the substance in question or whether it is a suspension of granules. Enzymes are here desig- nated as soluble if they have been prepared and crystallized by relatively mild methods. It should be understood that the so-called soluble enzyme may be loosely associated with structures insoluble in the cell or, on the other hand, enzymes difficult to separate from insoluble material may, under cer- tain conditions, be floating freely. Greenstein et al. (49) record an interesting example of a change of type of enzyme, more liver glutaminase appearing in insoluble fraction with the onset of cancerization while the activity of the soluble enzyme diminishes. CELLULAR STRUCTURE AND ACTIVITY tion of electron donor (cytochrome c), the concentration of electron acceptor (oxygen), or the concentration of enzyme (cytochrome oxidase). In this section we shall consider the possibility of variations in activity result- ing from variations in the molar concentra- tion of a given enzyme—necessarily assuming that substrates and cofactors are not limiting in the cell or at localized regions in the cell. Unfortunately, this assumption is not sup- ported by clear-cut evidence at the present time. Assays of enzyme content of cells are car- ried out routinely by homogenizing the tis- sue to break down the cell structure, adding appropriate amounts of substrate and _ co- factors, and then measuring the rate of the reaction in question. Under these conditions one expects that total enzyme content will be determined, any control by cell structure having presumably been abolished by the homogenization procedure. Large numbers of studies made with this technique have shown that adult tissues pos- sess distinctive enzyme patterns, differing markedly in “amounts of various enzymes per unit of tissue. In many cases the differ- ences are plainly correlated with the func- tion of the tissue—for example, high cholin- esterase activity in nerve tissue, high apy- rase activity in muscle, high phosphatase activity in kidney. Under pathological con- ditions enzyme patterns are significantly al- tered (cf. Greenstein, 47). The activity of enzymes as determined in test tube assays is obviously less meaningful than activity as deduced from metabolic studies on whole cells. Technical difficulties have prevented such comparisons from be- ing frequently made, but where they have been made (particularly in the field of oxygen uptake), wide discrepancies have been revealed. The work of Spiegelman and Stein- bach (45), showing a large increase in cyto- chrome oxidase activity on homogenization in frogs’ eggs, is an example of this. As a matter of fact, it seems reasonable to assume that most cells of adult organisms ordinarily + Activity of an enzyme is usually expressed as rate of substrate change per unit; but the appropri- ate unit basis on which to calculate activity is a difficult matter to decide. Some authors use a standard number of cells, some dry weight, some nitrogen content. Since enzymes are probably local- ized within cells, it is probable that any base unit is valid only in the most general sense, the most specifically valid being the cellular unit, if it can be clearly defined (cf. Davidson and Leslie, 50). CELLULAR METABOLISM carry an inactive reserve of enzymes to be used only in periods of unusual stress. This is self-evident in tissues like muscle, in which there is a clear-cut difference between rest- ing and activity metabolism; but it is prob- ably true for many other tissues also. Cells of embryonic tissues, however, may well operate with virtually no safety factors in enzyme concentration. Such cells presumably are never “resting” in the sense that adult muscle fibers or gland cells rest, and accord- ingly enzyme systems might be expected to function continuously at near maximum activity. A second element of uncertainty in the homogenate technique is the influence on activity of association of enzyme with struc- tural components of the cell. Although grind- ing tissue may destroy all the coarser ele- ments of cell structure, it is well known, as we have pointed out before, that different enzymes show varying affinities for particu- late matter in breis. Since enzyme activity must certainly depend in large part on pre- cise steric configurations, such as may be involved in the antigen-antibody situation (cf. Pauling, 48), it would not be surprising if a given number of enzyme molecules bound onto particles would exhibit a differ- ent activity from the same number of molecules in free solution (Mazia and Blum- enthal, *50). That such variations do occur is indicated by several studies in which different homog- enization media were used. In the case of muscle apyrase, for example, a water ho- mogenate has high activity but is insensitive to Ca**, whereas extraction with strong po- tassium chloride produces a soluble enzyme preparation of low activity which is strongly activated by Ca** (Steinbach, ’49). Similar effects have been noted with apyrase in the eranules of chick embryo breis (Steinbach and Moog, ’45). In neither case is there any assurance that only one enzyme is being dealt with, but the results are consistent with the view that association of enzymes with insoluble material alters their activity. The same interpretation can be placed on the report of Tyler (50) that respiration is in- hibited by dinitrophenol in tissue breis pre- pared as water homogenates, but is activated by dinitrophenol when the brei is prepared by homogenizing in Ringer’s solution with glucose added. The same enzyme might possibly vary in its structural relations from tissue to tissue. For this reason the differences in enzyme patterns previously cited may only partly 81 reflect differences in actual concentration of enzymes. Enzyme patterns have also been studied by the methods of chemical genetics. Here en- zyme activity is measured in terms of ac- tivities of intact cells, the assumption being made (but rarely tested) that other factors are not limiting. An excellent summary of these studies has been published by Tatum (49). DIFFERENCES IN ENZYME ACTIVITY Given a fixed number of enzyme molecules in a cell, their in vivo activity will depend SUPPLY ENZYME UNIT SUPPLY REVERSIBLE ASSOCIATION LOSS REGULATED BY OTHER INTRACELLULAR SYSTEMS OR BY EXCHANGE WITH ENVIRONMENT Fig. 8. Diagram to illustrate some possible meth- ods by which unit enzyme systems might be con- trolled. The enzyme unit (its protoplasmic analog might be a mitochondrion) is pictured as connected to the rest of the living system by (1) substrate supply, (2) waste or product removal and (3) dif- fusible co-factor supply. Control of activity might be accomplished by control of rate of flow of any of these factors. In addition, the state of the enzyme within the unit could be altered by (1) reversible association with inhibitors or (2) reversible change of activity by association or dissociation of enzymes into struc- tures. From our present knowledge of enzymatic content of particulate units it can be deduced that all these factors, and possibly more, might enter into the delicate control of cellular metabolism. on the quantity of the substrate available, on the adequacy of the cofactor supply, and on geometrical relations tending to inhibit or promote the reaction in which the enzyme participates. The term “activity,” applied to enzyme systems, bears an obvious relation- ship to the same term used in orthodox chem- istry, 1.e., a fudge-factor introduced to ac- count for the differences between observed rates of actions and those calculated on the 82 basis of known molar concentrations. In the case of enzyme activity, neither molar concentrations nor causes of activity varia- tions are usually known. In this section we shall consider enzyme activity* in terms of variations in molar concentrations and in activities. The diagram in Figure 8 may help to illustrate some of the variety of factors that might be operative in controlling the rate of an over-all reaction catalyzed by a given enzyme system. Supply of Substrate and Cofactors. A gen- eral illustration of this type of control in- volving the cofactor ATP has been developed with special acuity by Potter (’44). The way in which ATP and DPN aid in regulating phosphorylations and fermentations should be kept in mind here, for the general rea- soning used in these cases probably applies to other substances both in the entire cell and in localized areas within the cell. In- teresting modifications in patterns of activity must arise when any single substance or cofactor serving several enzyme systems is in short supply. Under such conditions com- petition for the limiting factor might even result in the complete exclusion of one re- action in favor of another. Exogenous substrate supply may be called a controlling factor of the second order, since | it appears to be itself under the control of selective mechanisms in the outer layers of the cell. Phosphate and sugar, for example, appear to enter yeast cells only during active metabolism (Spiegelman and Kamen, ’47), and even simple elements such as potassium have selective exchange systems related to the production of specific organic acids (Roth- stein and Meier, *48). The clearest examples of active regulation of movement of material in a living system are contained in the studies on sugar resorption in the kidneys (Pitts and Alexander, 44); and these find- ings and interpretations are useful in con- sidering the penetration of foodstuffs into cellular systems. The more orthodox aspects of permeability do not seem to shed much light on the problems of transfer of material in living cells (cf. material in the mono- graph by Davson and Danielli, ’43). An interesting aspect of surface control is the finding of Rothstein and Meier (’48) that numerous esterases and other enzymes local- ized on the outer layers of yeast cells will break down various tested intermediates of the glycolytic and oxidative cycles, and re- lease the hydrolytic products quantitatively into the medium. Thus they may guard the * See preceding footnote. CELLULAR STRUCTURE AND ACTIVITY orderly train of events going on within the cell by preventing the entrance of unplanned- for intermediates. We might, for exam- ple, infer that glucose, in order to be properly channeled into cell metabolism, must origi- nally enter into the metabolic reactions of the cell by being phosphorylated in the ini- tial stages; previously formed glucose phos- phate appears to be excluded. The Structural Orientation of Enzymes. The tendency of many enzymes to be associated with insoluble particulates after cellular dis- ruption has already been pointed out. At the present time a tentative understanding of many significant intracellular phenom- ena can be arrived at by assuming that such associations exist within the cell, and that they control activity both by keeping en- zyme and substrate apart or bringing them together, and also by bringing enzyme into contact with activators or inhibitors. Such geometrical arrangements are probably at the basis of very rapid changes in activity— for example, the rise in respiration in muscle and nerve that can be detected within milli- seconds after the onset of electrical activity. This rise is probably due to the “switching in” of enzymes previously immobilized, to act on substrates previously held in reserve. That the system switched in may be at least in part different from the resting system is indicated by studies like those of Stan- nard (739), who showed that azide does not affect resting muscle respiration, but abol- ishes the increase of respiration upon acti- vation. In comparing enzyme reactions in tissue under different physiological states, time is an important criterion in determining whether observed increases are due to al- terations in activity of pre-existing enzyme molecules, or to synthesis of new molecules. A short time change is of more diagnostic value because it most likely indicates en- zyme activation or inhibition. A change oc- curring over a long time, on the other hand, might indicate either enzyme activation or enzyme formation. There are numerous cases suggesting, but not proving, that enzyme localizations may serve as isolating mechanisms. All the stud- ies showing enzyme binding to cytoplasmic particles might be cited here, subject, of course, to confirmatory studies showing evi- dence of controlled change during activity. Chantrenne’s (’43) examination of the effects of moving the cytochrome oxidase-rich gran- ules of intact Amphiuma liver cells by cen- trifugation is especially interesting. In these CELLULAR METABOLISM large cells the granules can be seen packed into the centrifugal region. Concomitantly the Qo, is depressed, apparently because the enzyme has been concentrated to a point where the uniformly distributed substrate becomes limiting, even though the total amounts of both substrate and enzyme per cell are unchanged. Similar observations and conclusions have been reported for silk worm eggs (Wolsky, *50). The methods of cytochemistry have re- vealed many suggestive localizations of en- zymes in cells (cf. Bradfield, ’50). The Go- mori technique for alkaline phosphatase par- ticularly has been examined with sufficient thoroughness that we can rely on some, at least, of the intracellular localizations it has revealed; but again, correlated studies are needed to show how these localizations aid in controlling the functions of the cell. One in- stance apparently demonstrating functional change in location is the reported move- ment of ATP of striated muscle from the A bands, where it occurs during rest, to the I bands during activity (cf. Caspersson and Thorell, *42). To explain the very fast off-on effects that occur in response to stimulation, it may be necessary to seek out something more pre- cise than simple isolation between enzyme and substrate, which would apparently re- quire diffusion of substance over a finite dis- tance. Minor structural alterations, convert- ing an inhibited to an activated unit or vice versa, might be a faster mechanism. Among the few studies that appear to demonstrate a possible mechanism for such a type of control is that of Zamenhof and Chargaff (49) who found that DNase activity in cell- free yeast preparations is increased with age. The increase was shown to be due to the destruction of an inhibitor which is bound to the enzyme in fresh preparations. Prob- ably the same association of enzyme and inhibitor in the intact cell controls the rate of breakdown of the cellular DNA. In the same category we may perhaps place the report of Barron and his associates (’48) that the fixed sulfhydryl-containing enzymes in cells are poised at a given state of oxida- tion, and thus of activity, by the influence of other sulfhydryl groups freely diffused in the cytoplasm. One more possibility that should be men- tioned in this section is that the activity of an enzyme might be controlled simply by being bound or not bound to structural units. The evidence that the association of enzyme molecules into formed structures in- 83 fluences their activity considered. has already been THE REGULATION OF METABOLISM IN DEVELOPMENT That growth and differentiation must in- volve enzyme production is axiomatic. Recog- nition of this fact brings into sharp focus the problems concerned with the mecha- nisms by which the newly produced en- zymes are integrated into the orderly scheme of developmental events. In this section we shall attempt to determine to what extent the integrations of the embryonic period may be explained in the same terms as have proved useful in dealing with controlling mechanisms in adult cells. No attempt will be made to discuss the general energetics of development, since this topic is covered in Section VIII. It is necessary in this context first to dis- tinguish between metabolism as an aspect of development, and metabolism as a “cause” of development. Failure to make this distinc- tion has led to considerable confusion in the past. The melange of events that make up the life of the embryo is biochemical and functional as well as structural. That the chemogenetic events in some instances are - causally related to the morphogenetic events is reasonable to assume; but attempts to establish the causality are largely prema- ture in the present state of our knowledge. It seems to us that studies of regulating mechanisms in general embryonic biochem- istry may provide the most fruitful clues to the nature of the linkages between the chem- istry and the morphology of embryos. Three special precautions must be kept in mind in dealing with researches on en- zymes in embryos. First, the assumption that enzyme, cofactor, or substrate is not limiting where one of the three is being dealt with is less acceptable than in adult tissue, for ob- vious reasons. Second, the small quantities of embryonic tissue generally available, and its relative instability, make negative results more than usually suspect. The reported demonstration that the chick embryo does not have phosphorylative glycolysis, for example, was completely refuted by the more careful studies of Novikoff, Potter and LePage (’48). Third, it must never be forgotten that an embryo is in a continual state of differen- tiation. Any circumstances or conditions es- tablished for one stage, therefore, cannot be assumed to hold for any other stage unless specifically shown to do so. 84 ENZYME CONCENTRATION IN EMBRYONIC LIFE The most straightforward way in which embryonic metabolism is altered is through changes in relative concentrations of differ- ent enzymes. Among the few studies clearly bearing on this problem are a sufficient num- ber on one form, the chick embryo,* to enable us to state with assurance that em- bryonic tissue at every stage is enzymatically individual, rather than being merely a sim- plified or diluted version of any adult type (cf. Moog, *52). During the first few days of development, the tissue of the embryonic chick is relatively poor in phosphatases (Moog and Steinbach, 46; Moog, ’46); rela- tively rich in various peptidases (Levy and Palmer, *43); and very rich in cytochrome oxidase and succinoxidase [this fact is cal- culated in part from the data of Albaum et al. (46) who, however, reported their find- ings only in terms of total quantity per embryo]. Interestingly enough, the period of intensive histogenetic activity beginning at the end of the first week of incubation is preceded by a slackening in rate of pro- duction of all enzymes so far studied. Finding a differential increase in enzyme activity per unit of embryo by the homog- enate method may, of course, mean only that a given tissue rich in the enzyme in ques- tion is increasing faster than other tissues. Undoubtedly this is the explanation for nu- merous cases of enzyme increase per unit of whole organism. But it has by now been amply demonstrated that differential changes in enzyme concentrations also occur in iso- lated tissues; in the brain of the honey bee, in fact, cholinesterase rises while cell num- ber declines (cf. Rockstein, 50). Studies like those of Sawyer (’43a, °43b) on cholines- terase in muscle and nerve, Hermann and Nicholas (48) on apyrase in muscle, and * This section will be largely, but not wholly, confined to the chick embryo. By reason of easy availability and relative purity of cytoplasm, the chick embryo seems better adapted to the type of research we are dealing with here than any other form. The enzymology of amphibian embryos has been much studied, but in these the difficulty of separating inert yolk from active cytoplasm, or of determining how fast one is converted into the other, has meant that there has been no satisfactory unit basis to which enzyme activity could be re- ferred. Gregg and Lovtrup (50) have, however, recently proposed a method for determining non- yolk nitrogen in salamander eggs. At the present time, no form has been explored fully enough to serve as the sole basis of even a brief résumé of embryonic metabolism. CELLULAR STRUCTURE AND ACTIVITY Moog (750, 51) on phosphatase in intestine, seem to be establishing as a fact the repeated finding that in advanced stages of develop- ment an enzyme accumulates only in cor- relation with the function it subserves— either in parallel with the function, or in slight forward reference to it. If this view is correct, it will serve as a valuable guide in examining the poorly ex- plored enzymology of the early stages of development. It is in these early stages (roughly the first six days in the life of the chick, for example) that we can with ap- proximate correctness speak of ‘embryonic tissues” as a type, in contrast to the partly or fully differentiated tissues of later stages. Finding that enzymes do not behave inde- pendently of their correlated function may justify our concluding that some, at least, of the enzymes of very young embryonic tissue are actively related to the proper func- tion of such tissue, i.e., development. With embryonic material just as with adult, values for enzyme activity obtained for homogenates or extracts cannot be uncriti- cally accepted as reflecting the state of affairs in vivo. Quite possibly, however, the enzyme activities obtained with embryonic homog- enates are generally closer to the activities in intact tissue than is the case with adult material. As pointed out before, the embry- onic cell might be expected not to store a large reserve of enzymes for emergency use, since the orderly nature of embryonic life is itself a guarantee against the stresses that the adult must face. ENZYME ACTIVITY IN EMBRYONIC LIFE The Influence of Substrate and Cofactor Concentration. That the activity engaged in by a given number of enzyme molecules may be limited by availability of substrate or of cofactors has already been pointed out. It might be supposed, then, that the rate at which raw materials are supplied, by the yolk sac or placenta or other commissary agency, would control the rate at which bio- chemical reactions go forward within the embryonic body. At the present time, how- ever, no good evidence bears on this prob- lem. One might call to mind the old obser- vation reported by Needham (’31), that the tendency of the chick embryo to use its energy sources in succession—carbohydrate first, then protein, then fat—is not altered by the injection of large amounts of glu- cose at the period when protein is in prin- CELLULAR METABOLISM cipal use. But in this case it was not clear that the injected material was actually carried into the embryo at an enhanced rate. There are no unequivocal cases, either, to show that differentiated enzymes are in- hibited by lack of necessary cofactors. In a very thorough study, Novikoff et al. (48) demonstrated that even at three days, the earliest stage investigated, the chick embryo contains ATP, DPN and phosphocreatine in amounts falling within the same range as in various mature rat tissues. They also found that numerous intermediates of the phos- phorylating glycolytic system are present in substantial quantities, indicating that this system proceeds with about equal intensity in embryonic and mature tissues. The find- ing of Potter and DuBois (’42) that cyto- chrome c is present only in traces in the six- day chick embryo is, on the other hand, an example of the difficulty of accepting nega- tive evidence in this field. For we know that during the first three days the intact embryo takes up oxygen freely (Philips, ’40), that it gives a strong Nadi reaction that is inhibited by azide (Moog, °43), and that homogenates contain abundant cytochrome oxidase (AI- baum et al., °46); the latter authors make the rather equivocal statement that in young embryos “autoxidation” of added substrate in the absence of added cytochrome c repre- sents a large fraction of the observed oxygen uptake. But perhaps it is naive to assume that the embryo differentiates or accumulates some elements of its biochemical machinery in anticipation of others. The lack of evi- dence demonstrating this type of control may in itself be evidence that the various factors are produced only as they are needed, and in correlation with other necessary factors. This view is in harmony with the previously mentioned parallelism between enzyme ac- cumulation and function. Recent studies on the formation of the so- called adaptive enzymes may provide a more subtle approach to the problem of enzyme regulation by material supply. The im- portant discussion by Spiegelman (748) sug- gests a mechanism by which substrate avail- ability may direct the actual formation of enzymes, and thus control the processes of growth and morphogenesis dependent on the enzymes formed. In essence the hypothesis of long-term regulation of enzyme _ pat- terns in cells, whether embryonic or adult, by adaptive formation of enzymes, rests upon three postulates, which we will consider 85 here. Each of these postulates, it should be noted, is a demonstrated fact in itself; the extent to which these three processes actually cooperate in regulating cell metabolism, how- ever, remains to be studied in detail. 1. The amount of enzyme normally pres- ent in a cell represents a balance between synthesis and destruction. 2. Enzymes tend to be stabilized (i.e., de- struction is retarded?) when combined with suitable substrate (cf. Bayliss, ’19). Accord- ing to more recent evidence, substances other than normal substrate can promote the for- mation of adaptive enzymes (Spiegelman, 48). 3. Most, perhaps all, of the protein of a living cell is potential material for the formation of enzymes. Thus synthesis of special enzymes must involve competitive interactions. In yeast, under conditions of starvation, even an enzyme regarded as con- stitutive (glycozymase) can be depleted dur- ing the formation of a clearly adaptive en- zyme, galactozymase (Spiegelman and Dunn, "47). A plausible, though not yet demonstrated, integration of these three facts into a single regulatory mechanism is shown in Figure 9. This figure illustrates how, in the normal steady state, with raw materials supplied and breakdown products removed, the rela- tive amount of activity of enzymes A and B will depend on availability of substrate, among other factors. With low supply of substrate for B, for example, the total amount of B would diminish; but increase of sub- strate for B would raise the activity only as fast as new B could be formed. Presence of a stabilizing (inhibiting) substance (B’) would, however, save the enzyme from de- struction and allow immediate increase of activity of B if normal substrate were added in excess, the excess substrate then out- competing the inhibitor. Although enzyme patterns have been al- tered by supplying excess substrate in yeasts and bacteria, it should be noted that such treatment does not necessarily affect enzyme concentration. For example, the substrate might be present in excess to begin with, or an enzyme might be held in large quan- tities as a form of storage protein. Some seeds contain tremendous quantities of urease, which falls sharply during development (Williams, *50). Assuming that the urease- synthesizing system is a good competitor, one might interpret the observation to mean that the seed uses urease as a means of stor- ing protein out of reach of other enzyme 86 systems, which might otherwise get out of hand. Hypothetically, at least, the scheme of competitive interaction is applicable to the problem of embryonic differentiation. Since no egg is completely homogeneous in the distribution of its chemical constituents, a way is immediately open for the differential RAW MATERIALS A ~ OR CELLULAR STRUCTURE AND ACTIVITY bryos is alterable by addition of succinate (Boell, ’49), but in this case normal develop- ment is possible over only a narrow range of enzyme enhancement. Probably a more fav- orable type of material for experiments on adaptation is made available in the im- portant experiments of Spratt (49), who has shown that explanted chick blastoderms ch r ca L oS SSeS: Fig. 9. Diagram to illustrate some of the principles pertinent when two or more enzymes, subject to deg- radation, are competing for a common stock of enzyme precursor. Raw materials, presumably from endog- enous synthetic mechanisms, supply precursor. Enzymes, once formed, are stabilized by (A and B) com- bination with substrate or (B’) combination with inhibitor or structural analog of normal substrate. If pre- cursor is limiting, A or B enzyme systems might dominate depending on (1) the effectiveness of the syn- thetic system for A or B and (2) the relative stabilization by substrate. Thus starvation might have highly differential effects. production of various enzymes in different regions. As the enzymes so produced must themselves affect their surroundings, con- tinued differentiation becomes self-perpetu- ating. How amenable this scheme is to ex- perimental verification is, however, another matter. Work with great numbers of inhibi- tory agencies has taught us little except how easy it is to upset fatally the delicate bal- ance of developmental events. Will excess substrates merely add themselves to the list of inhibitors, if they have any effect at all? One preliminary report does indicate that succinic dehydrogenase in amphibian em- can use a variety of six-carbon sugars as exogenous energy sources. The Influence of Structural Orientation. Em- bryology provides us with many instances of systems held in an inhibited state. The dormant seed of a green plant, the orthop- teran egg in diapause, the winter egg of rotifers, the unfertilized egg of marine in- vertebrates, the prolonged blastocyst of the deer, the unincubated cicatrix of the hen’s egg, are examples of such restraint. In many cases the system can be released suddenly, indicating that structural factors were re- sponsible for the blockage. Experimentally CELLULAR METABOLISM the low rate of oxygen uptake of frogs’ eggs can be raised merely by homogenizing (Spiegelman and Steinbach, °45). The most thoroughly explored case of sud- den activation in normal development is that of the fertilization of sea urchin eggs. In this field we owe most of our knowledge to the laboratory of Professor Runnstrém, who has recently reviewed the work (cf. Runnstrém, °49). We have no room here to do more than point out some of the major evidences that the first few minutes following sperm en- trance involve simultaneous and no doubt correlated changes in enzymatic activity and structural characteristics. During this period the rate of oxygen uptake is raised or at least stabilized (Borei, ’48), with the prev- iously immobilized cytochrome system com- ing into action; there is a rapid breakdown of high molecular weight carbohydrates (Or- strom and Lindberg, ’40); and nonprotein ni- trogen increases rapidly (Orstrém, °41). At the same time calcium is liberated, appar- ently from a proteinate (Mazia, °37); an increasing amount of protein becomes insolu- ble in molar potassium chloride (Mirsky, 36); and cephalin changes to an ether-insol- uble form (Ohman, ’44). The egg surface is strongly affected, the thin jelly-coated vitel- line membrane giving way to the tough, birefringent fertilization membrane. Of course the practical outcome of all this post- fertilization activity is the lifting of the block to development and the initiation of cleav- age. In this whole system, it would seem, we have the most favorable material now avail- able* for showing on the one hand the rela- tionship between protein structure and en- zyme activity, and on the other the nature of the linkage between the structure-controlled activity and the events of development. Sudden alterations of activity are, however, probably a minor aspect of the life of the embryo. The proper business of an embryo is differentiation, and one would expect that the enzymatic orientations of the embryo dif- ferentiate in the same sense as does the body structure itself. In a few cases we have evi- dence that biochemical elaboration of formed elements does occur. * The much-studied diapause state of the orthop- teran egg has previously seemed to be an equally useful case of inhibition of development paralleled by immobilization of enzyme systems. The recent finding that the respiratory intensity of diapause homogenates is lower than that of homogenates from active stages (Bodine, ’50) indicates, however, that the situation is more complex than previously supposed. 87 The granular structure of embryonic cy- toplasm is the one problem in this field that has been studied with any degree of thor- oughness. The real existence of granules in intact embryonic cells has recently been shown by experiments in which P*? was in- jected at unspecified stages into hen’s eggs or gravid mice (Jeener, *49). In embryos homogenized two hours after the injection, the specific activity of radiophosphate was found to be differentially distributed among the nucleic acids of granules isolated at 13,000 g. and at 60,000 g., and of the final supernate. Since the differences are too great to be accounted for merely by the exchange rate, it appears that the granules are not merely artifacts accumulated during the homogenization process, but occur as real entities in the intact cells. In the chick embryo from seven to eleven days of incubation, centrifugable granules accumulate an increasingly large portion of the total nitrogen, and at the same time be- come relatively richer in apyrase and alka- line phosphatase activity (Steinbach and Moog, 45; Moog and Steinbach, ’46). The granules of amphibian eggs gradually bind all of the nucleoprotein, which is entirely free at the beginning of development, and they also become associated with catalase, dipeptidase, ribonuclease, and alkaline phos- phatase, although they are, surprisingly, said to be poor in respiratory enzymes (Brachet and Chantrenne, ’42); it is unfortunate that these fascinating results have never been well documented. In sea urchin eggs a num- ber of enzymes are bound to granules, but others are dispersed in the hyaloplasm (cf. Holter, ’49). It is not known, however, if the enzymatic character of the granules changes during development. Among the few other instances of biochem- ical elaboration of structure during develop- ment, the most interesting is the finding that in the muscles of the rat, just before and after birth, apyrase accumulates at a faster rate than myosin (Herrmann and Nicholas, 48). The enzyme activity, it appears, is added to the protein structure as warranted by the needs of the whole organism. A similar, though less clear, case is the shift in cal- cium activability of apyrase in homogenates of chick liver. Under the same conditions of preparation, the enzyme is slightly inhibited by calcium before hatching, but strongly stimulated after hatching (Moog, °47); as pointed out earlier, a difference in structural association could account for such an effect. A recent study correlating quantitative and 88 histochemical results has shown that the tremendous quantities of alkaline phospha- tase which accumulate in the chick intestine at hatching are largely restricted to the presumable functional localization in the striated border (Moog, 50, °51). If this is a general situation—and other cytochemical studies indicate that it is—one may be justi- fied in concluding that differentiating cells control the activity of newly synthesized en- zymes by placing them immediately in the positions in which their activity can be directed and used. Thus the middle of the twentieth century finds us with but a rudimentary understand- ing of the field of metabolic regulations dur- ing development. Plainly we do not have the answers to either of the two aspects of the problem as stated at the beginning of this section. Before we can hope to understand either the way in which the differentiating threads of the total metabolism are woven together into a unified whole, or the way in which morphogenesis emerges from the controlled metabolism, further developments along two lines are necessary. First, our knowledge of the control of embryonic en- zymology can move forward only as fast as our general knowledge of the ultrastructure of protoplasm advances. In this realm un- doubtedly lie many explanations that we can only hint at now. Second, within the field of embryology itself, more study of isolated tissues is indicated. Although what has been called the actuarial approach to embryonic enzymology has yielded some valuable in- formation on the enzymatic characteristics of differentiating tissue, the validity of such a method is obviously limited. It is with the aid of the methods of histochemistry and cytochemistry that we may expect biochem- ical embryology to make its next steps for- ward. REFERENCES Albaum, H., Novikoff, A., and Ogur, M. 1946 The development of the cytochrome oxidase and succinoxidase systems in the chick embryo. J. Biol. Chem., 765:125. Barron, E. S. G. 1943 Mechanisms of carbohy- drate metabolism. Adv. Enzymol., 3:149. 1952 The oxidation pathways of carbo- hydrate metabolism; in Trends in Physiology and Biochemistry. Academic Press, New York. , Nelson, L. and Ardao, M.A. 1948 Regu- latory mechanisms of cellular respiration. II. The role of soluble sulfhydryl groups. J. Gen. Physiol., 32:179. CELLULAR STRUCTURE AND ACTIVITY Bayliss, W. M. 1919 ‘The Nature of Enzyme Ac- tion. Longmans, London. Bodine, J. H. 1950 To what extent is O2 uptake of the intact embryo related to that of its homog- enate? Science, 772:110. Boell, E. J. 1949 The effect of sodium succinate on the development of succinic dehydrogenase in Amblystoma punctatum. Anat. Rec., 105:120. Borei, H. 1948 Respiration of oocytes, unfertil- ized eggs, and fertilized eggs from Psammechinus and Asterias. Biol. Bull., 95:124. Borsook, H., Deasy, C. L., Haagen-Smit, A. J., Keighly, G., and Lowy, P.H. 1949 Uptake of labelled amino acids by tissue proteins in vitro. Fed. Proc., 8:597. Brachet, J., and Chantrenne, H. 1942 Nucleo- proteides libres et combinés sous forme de gran- ules chez l’oeuf d’Amphibiens. Acta biol. belg., 4:451. Bradfield, J. R. G. 1950 The localization of en- zymes in cells. Biol. Rev., 25:113. Caspersson, T., and Thorell, B. 1942 The local- ization of the adenylic acids in striated muscle fibers. Acta Physiol. Scand., 4:97. Chantrenne, H. 1943 Association de la cyto- chrome-oxydase 4 des complexes sedimentables dans le cytoplasme vivant. Acta biol. belg., 3:99. Clark, W. M. 1949 ‘Topics in Physical Chemis- try. Williams & Wilkins Co., Baltimore. Davidson, J. N., and Leslie, I. 1950 A new ap- proach in the biochemistry of growth and develop- ment. Nature, 765:49. Davson, H., and Danielli, J. F. 1943 The Perme- ability of Natural Membranes. The Macmillan Co., New York. Dixon, M. 1949 Multi-enzyme Systems. Cam- bridge University Press, Cambridge, England. Fruton, J.S. 1950 Role of proteolytic enzymes in biosynthesis of peptide bonds. Yale J. Biol. & Med., 22:263-271. Goldinger, J. M., and Barron, E.S.G. 1946 The pyruvate metabolism of sea-urchin eggs during the process of cell division. J. Gen. Physiol., 30: 73. Greenstein, J. P. 1947 Biochemistry of Cancer. Academic Press, New York. ———., Fodor, P. J., and Leuthardt, F. M. 1949 The neoplastic transformation as a_ biological fractionation of related enzyme systems. J. Nat. Cancer Inst., 10:271. Gregg, J. R., and Lovtrup, S. 1950 Biochemical gradients in the axolotl gastrula. Compt. rend. Trav. Lab. Carlsberg., 27 (No. 12): 307. Hassid, W. Z., and Doudoroff, M. 1950 Enzymat- ic synthesis of sucrose and other disaccharides. Adv. Carbo. Chem., 5:29. Heilbrunn, L. V. 1943 An Outline of General Physiology. W. B. Saunders Co., Philadelphia. Herrmann, H., and Nicholas, J. S. 1948 En- zymatic liberation of inorganic phosphate from adenosine triphosphate in developing rat muscle. J. Exp. Zool., 107:177. Holter, H. 1949 Problems of enzyme localization in development. Publ. Staz. Zool. Nap., 2/ (suppl.) :60. CELLULAR METABOLISM Jeener, R. 1949 Distribution of ribonucleic acid in the cytoplasm of growing cells studied with P32, Nature, 163:837. Kalckar, H. 1945 Enzymatic synthesis of nucleo- sides. Fed. Proc., 4:248. Krebs, H. A. 1943 The intermediary stages in the biological oxidation of carbohydrates. Ady. Enzymol., 3:191. Lardy, H. A. (editor) 1950 Respiratory En- zymes. Burgess Publishing Co., Minneapolis. LaValle, J. E., and Goddard, D. E. 1948 The mechanism of enzymatic oxidations and reduc- tions. Quart. Rev. Biol., 23:197. Levy, M., and Palmer, A. J. 1943 Chemistry of the chick embryo. IV. Aminopeptidase. J. Biol. Chem., 150:271. Lipmann, F. 1941 Metabolic generation and util- ization of phosphate bond energy. Adv. Enzymol., 1:99. 1949 Mechanism of peptide bond forma- tion. Fed. Proc., 8:597. Lundegardh, H. 1945 Absorption, tyansport and exudation of inorganic ions by the roots. Arch. Botanik, 32A (No. 12): 1-139. Mazia, D. 1937 ‘The release of calcium in Ar- bacia eggs on fertilization. J. Cell. Comp. Physiol., 10:291. , and Blumenthal, G. 1950 Inactivation of enzyme substrate films by small doses of x-ray. J. Cell. Comp. Physiol., 35: Suppl. 1, 171. Meyerhof, O. 1949 Glycolysis of animal tissue extracts compared with the cell-free fermentation of yeast. Wallerstein Comm., 38:255. Michaelis, L., and Menten, M.L. 1913 Die Kine- tik der Invertinwirkung. Biochem. Z., 49:333. Mirksy, A. E. 1936 Protein coagulation as a result of fertilization. Science, 84:333. Moog, F. 1943 Cytochrome oxidase in early chick embryos. J. Cell. Comp. Physiol., 22:223. 1946 Alkaline and acid phosphomonoes- terase activity in chick embryos. J. Cell. Comp. Physiol., 28:197. 1947 Adenylpyrophosphatase in_ brain, liver, heart, and muscle of chick embryos and hatched chicks. J. Exp. Zool., 105:209. 1950 The functional differentiation of the small intestine. I. The accumulation of al- kaline phosphatase in the duodenum of the chick. J. Exp. Zool., 715:109. 1951 The functional differentiation of the small intestine. II. The differentiation of al- kaline phosphatase in the duodenum of the mouse. J. Exp. Zool., 718:187. 1952 Enzymes in the development of the chick embryo. Ann. N. Y. Acad. Sci., 55:57. , and Steinbach, H. B. 1946 Localization of acid and alkaline phosphomonoesterases in cytoplasmic granules. J. Cell. Comp. Physiol., 28: 209. Needham, J. 1931 Chemical Embryology. Cam- bridge University Press, Cambridge, England, Vol. II, p. 997. Novikoff, A. B., Potter, V. R., and LePage, C. A. 1948 Phosphorylating glycolysis in the early chick embryo. J. Biol. Chem., 773:239. 89 Ochoa, S. 1947 Chemical process of oxidative re- covery. Ann. N. Y. Acad. Sci., 47:835. , and Stern, J. R. 1952 Carbohydrate me- tabolism. Ann. Rey. Biochem., 27:547. Ohman, L. 1944 On the lipids of the sea-urchin egg. Ark. Zool., 36A (No. 7):1. Orstro6m, A. 1941 Uber die chemischen Vor- gange inbesondere den Ammoniakstoffwechsel bei der Entwicklungserregung des Seeigeleis. Z. physiol. Chem., 277;1. , and Lindberg, O. 1940 Uber den Kohle- hydratstoffwechsel bei der Befruchtung des Seeigeleies. Enzymol., 8:376. Pauling, L. 1948 Antibodies and specific bio- logical forces. Endeavour, 7 (No. 26) :43. Philips, F. S. 1940 The oxygen consumption of the early chick embryo at various stages of devel- opment. J. Exp. Zool., 86:257. Pitts, R. F., and Alexander, R.S. 1944 The renal reabsorption of inorganic phosphate in the nor- mal dog. Fed. Proc., 3:37. Potter, V.R. 1944 Biological energy transforma- tions and the cancer problem. Ady. Enzymol., 4: 201. , and DuBois, K. P. 1942 The quantita- tive determination of cytochrome. J. Biol. Chem., 142:417. Ratner, S. 1949 Mechanism of urea synthesis. Fed. Proc., 8:603. Rockstein, M. 1950 The relation of cholin- esterase activity to change in cell number with age in the brain of the adult worker honeybee. J. Cell. Comp. Physiol., 35:11. Rothstein, A., and Meier, R. 1948 The relation- ship of the cell surface to metabolism. I. Phos- phatases in the cell surface of living yeast cells. J. Cell. Comp. Physiol., 32:77. Runnstrém, J. 1949 Some results and views con- cerning the mechanism of activation of the sea- urchin egg. Pub. Staz. Zool. Nap., 27 (suppl.) :9. Sawyer, C. H. 1943a Cholinesterase and the be- havior problem in Amblystoma. I. The relation- ship between development of the enzyme and early motility. J. Exp. Zool., 92:1. 1943b Cholinesterase and the behavior problem in Amblystoma. III. The distribution of cholinesterase in nerve and muscle throughout development. J. Exp. Zool., 94:1. Spiegelman, S. 1948 Differentiation as the con- trolled production of unique enzymatic patterns. Symp. Soc. Exp. Biol., 2:286. ,and Dunn, R. 1947 Interactions between enzyme-forming systems during adaptation. J. Gen. Physiol., 37:153. , and Kamen, M. D. 1947 Some basic problems in the relation of nucleic acid turnover to protein synthesis. Cold Spring Harbor Symp., ae -OAN , and Reiner, J. 1942 A kinetic analysis of potassium accumulation and sodium exclusion. Growth, 6:367. , and Steinbach, H. B. 1945 Substrate- enzyme orientation during embryonic develop- ment. Biol. Bull., 88:254. Spratt, N. T., Jr. 1949 Nutritional requirements 90 of the early chick embryo. I. The utilization of carbohydrate substrates. J. Exp. Zool., 110:273. Stannard, J. N. 1939 The mechanisms involved in the transfer of oxygen in frog muscle. Cold Spring Harbor Symp., 7:394. Steinbach, H. B. 1949 Calcium and apyrase sys- tem of muscle. Arch. Biochem., 22:328. , and Moog, F. 1945 Localization of ade- nylpyrophosphatase in cytoplasmic granules. J. Cell. Comp. Physiol., 26:175. Szent-Gyorgyi, A. 1939 On Oxidation, Fermenta- tion, Vitamins, Health and Disease. Williams & Wilkins Co., Baltimore. 1951 Muscular Contraction. Academic Press, New York. Tatum, E. L. 1949 Amino acid metabolism in CELLULAR STRUCTURE AND ACTIVITY mutant strains of microorganisms. Fed. Proc., 8: 511: Tyler, D. B. 1950 Rate of oxygen uptake of dif- ferently prepared brain suspensions in the pres- ence of 2,4-dinitrophenol. Arch. Biochem., 25: oo ie Werkman, C. H., and Wood, H.G. 1942 Hetero- trophic assimilation of COg. Adv. Enzymol., 2: 1135: Williams, W. T. 1950 Function of urease in Cit- rullus seeds. Nature, 765:79. Wolsky, A. 1950 Changes in the response of silk- worm eggs to rotational force during cleavage. Nature, 765:119. Zamenhof, S., and Chargaff, E. 1949 Studies on the desoxypentose nuclease of yeast and its spe- cific cellular regulation. J. Biol. Chem., 180:727. Section LIT CHAPTERS: Cell Division HANS RIS INTRODUCTION GrowtTH usually is connected with the for- mation of new cells. The ways in which new cells originate have been intensively inves- tigated and much debated ever since the cellular nature of organisms was realized. Though Trembley had already illustrated binary fission of a diatom in 1748, the sig- nificance of this process was first realized by von Mohl in 1837 (Baker, ’51). Yet not until the middle of the century did it be- come generally accepted that cells originate only through division of pre-existing cells, mainly because of the influence of Schleiden and Schwann who vigorously fought for their theory of free cell formation. With the introduction of fixatives and stains, the par- affin sectioning technique and improved mi- croscopes, more detailed study became pos- sible. The first suggestion of the complexity of nuclear division appeared in a paper by Schneider in 1873 who noticed the forma- tion of chromosomes and their separation into two groups which formed the daughter nuclei. In a few exciting years following this the essential features of cell division in plants and animals were discovered through the work of Flemming and Strasburger, van Beneden, Fol and Biitschli, O. and R. Hert- wig and Boveri. In 1879 Flemming described the longitud- inal division of the chromosomes and the separation of the halves into the two daughter cells. The appearance of these basophilic threads in the nucleus and their exact longi- tudinal division was the most striking aspect of cell division and Flemming therefore called it “mitosis” (mitos = thread). The direct simple fission of nucleus and cyto- plasm was thought to be another mode of cell proliferation and named “amitosis.”” With the establishment of the chromosome theory of inheritance at the turn of the century the complexity of nuclear organization was realized and the significance of the beautiful 91 precision of mitosis became clear. More gradually we have become aware of the com- plexity of cytoplasmic organization because it is more subtle and more varied. Today the cell appears as a complex system, a kind of hierarchy of more or less autonomous components. In the chromosomes we have the highest concentration of the factors that are responsible for specificity in the cell, and their irreplaceability is expressed in the ex- treme care taken in their exact duplication and distribution during the reproduction of the cell. It is not surprising that the mecha- nisms by which the cell accomplishes this have been in the center of interest so far. There exist in the cytoplasm, however, sys- tems of at least partial autonomy and the ways of their reproduction and distribution must be more intensively investigated. For the student of development and differentia- tion particularly they may prove of even ereater interest than the behavior of the chromosomes. The division of the nucleus, however, is the central process around which cell re- production is organized and it is better known than the reproduction of the cytoplasmic systems. The following discussion will, there- fore, be organized mainly around nuclear behavior. No attempt will be made to review the large literature in the field. The dis- cussion may be regarded as rather a_ per- sonal and therefore somewhat biased essay on some of the essential problems of cellular reproduction. DESCRIPTION OF MITOSIS IN THE WHITEFISH BLASTULA From algae to orchids and amoeba to man the essential processes of cell division are remarkably similar. We can describe, there- fore, the division of one cell and thereby il- lustrate the fundamental strategy in all. Only bacteria and blue-green algae still 92 CELLULAR STRUCTURE AND ACTIVITY seem to fall outside of the general picture (but cf. DeLamater and Mudd, ’51). The divisions of the blastomeres in the whitefish blastula are useful to describe the morphological aspects of mitosis in an ani- mal cell (Fig. 10). Following Strasburger the process is subdivided into four phases: During prophase the centrosome divides (Fig. : : . ' a) E et, : i Fig. 11. Prophase in blastomere of whitefish, stained with iron-hematoxylin to show centriole. 10A); the centrioles are not visible in these preparations but can be demonstrated in slides stained with iron-hematoxylin (Fig. 11). The chromosomes now appear embedded in eosin- ophilic material that later forms the spindle (Figs. 10A and B) and finally the nuclear membrane breaks down (Fig. 10C). While the spindle takes shape the chromosomes be- come oriented in the equatorial plane (Fig. 10D): metaphase. During anaphase the chromosomes move toward the spindle poles and the spindle elongates (Figs. 10E, F, G). Finally, in telophase the nuclear membrane reappears, in this case around each chomo- some separately, so that karyomeres are formed which later may fuse into a single nucleus. The cytoplasm is subdivided by the cleavage furrow (Fig. 10H). This account of mitosis in the whitefish blastomere has been pieced together from various stages in fixed and stained cells. Is it possible to follow the process in the living cell as it divides? If we place the blastula of a whitefish egg during cleavage in calcium- free Ringer’s solution and gently squeeze it in a rotocompressor,* the cells separate and a single layer of cells can be obtained. For some time the blastomeres continue to divide * Obtained from Biological Institute, Philadel- phia. CELL DIVISION Imin.45sec. £ iG Smin. 45 sec. H _ Simin lose. 1 93 il fe) aR 45sec. C : 4 2min.45 sec. ‘st ree Fig. 12. Mitosis in a living blastomere of the whitefish Leucichthys artidi (Spencer Phase Contrast, dark medium). and the process can be followed in the phase microscope. The anaphase in such a cell is shown in Fig. 12. In addition to chromo- somes the outline of the spindle is clearly visible in the living cell, and sometimes even details such as the centrioles, chromosomal fibers and aster-rays (Fig. 12C). Many other cells of plants and animals have been used to study mitosis in the living cell. With the phase microscope many struc- tures are now visible in life that before were known only from fixed and stained prepara- tions. Photographs of a variety of living cells in division have been published (Table 2). ANALYSIS OF MITOSIS After this brief description of a cell in division we shall now look at some aspects of mitosis in more detail. We can recognize two main events: The division of the nucleus (karyokinesis) and the division of the cyto- plasm (cytokinesis). The nuclear division normally involves duplication of the chromo- some units, the splitting of the chromosome into two equivalent halves (chromatids), their condensation into compact bodies, and a series of chromosome movements resulting in the distribution of the chromosome halves 94 TasLe 2. Illustrations of Living Cells in Mitosis MATERIAL AUTHOR Amblystoma heart tissue Wang, °52 culture Grasshopper sperma- Beélar, °29; Michel, 50 tocytes Frog fibroblast tissue Hughes and Preston, °49 culture Chick embryo osteoblast Hughes and Swann, 48 tissue culture Mouse spleen tissue culture Fell and Hughes, ’49 into the daughter cells. Cytokinesis divides the cytoplasm more or less equally around the daughter nuclei. Cytoplasmic components are generally segregated at random but may become differentially distributed in certain cells. KARYOKINESIS THE CHROMOSOME CYCLE One of the most striking aspects of mitosis is the cyclical change in the chromosomes. ii Microfibril Half chromatid Chromatid Chromosome Fig. 13. Diagram of the subunits in a completely uncoiled chromosome. This involves chromosome _ reproduction, changes in structure (spiralization cycle) and chemical and physiological changes. Structural Changes in Chromosomes. It is generally agreed today that chromosomes consist of a bundle of threads that undergo a cycle of coiling and uncoiling. The exact number of units (chromonemata) has been much debated. There is good evidence that CELLULAR STRUCTURE AND ACTIVITY in many cases the daughter chromosomes at anaphase contain at least two and some- times even four microscopically visible units. Recent studies with the electron microscope in this laboratory, however, have shown that the elementary longitudinal unit is submi- croscopic and in several animals and a plant was found to be of the same thickness (about 500 A). The number of longitudinal subunits in a chromosome therefore must be deter- mined with the electron microscope. We have found evidence that the number of these elementary microfibrils is not the same in comparable chromosomes of different organ- isms (Ris and Kleinfeld, *52). The micro- fibrils and the larger units they compose may be either closely appressed or separate from each other laterally. This fact and the submi- croscopic size of the structural unit are probably responsible for most disagreements and uncertainties of chromosome structure. The unit of reproduction in the chromo- some must therefore be sought in the ele- mentary microfibril. Since chromosomes con- tain more than one microfibril it is, however, not the unit of division during mitosis. This unit is the chromatid, that itself may be visibly separated into two or more bundles of microfibrils (half-chromatids and quarter- chromosomes) (Fig. 13). In early prophase chromosomes show a large number of small gyres. These increase in width through a process of “despiraliza- tion.” While the number of gyres decreases, through this process of “gyre elimination” the chromosomes become shorter and thicker and gain their typical metaphase form (Spar- row, Swanson, Ris; for references see Man- ton, 50). In prophase of meiosis the chromosomes usually condense more than in somatic cells. This is connected with the formation of a double spiral (minor and major coil). The number of gyres in a metaphase chromo- some seems to be remarkably constant for one type of cell under similar conditions, but may vary considerably in different types of cells (somatic mitosis, first and second division of meiosis, for instance). In telo- phase the process of gyre elimination con- tinues so that only a general waviness re- mains in the interphase nucleus (relic spiral). A few workers have attempted to calcu- late the length of the chromonemata during the coiling cycle. In the fern Osmunda (Man- ton) and in Trillium (Sparrow, Wilson and Huskins) considerable variation in the length of the chromonema was found during pro- CELL DIVISION phase of meiosis (see Manton, 50). Whether such changes take place generally during mitosis is not known. What causes the spiralization of chromo- somes? So far we have no satisfactory ex- planation. Much more needs to be known about the chemical changes during coiling and the submicroscopic structure of the chromosome before we can begin to under- stand the mechanism of coiling. During prophase the chromosomes are said to become surrounded by a membrane (pel- licle) and a matrix in which the coiled chromonemata are embedded or which sur- round each individual chromonema (cf. Kauf- mann, °48), but there is no good evidence for this contention. Chemical Changes in the Chromosome. In recent years much progress has been made in our knowledge of chromosome chemistry since it has become possible to isolate nuclei and chromosomes for chemical analysis and with the development of appropriate cyto- chemical techniques. (For a review see Dav- idson, 50; Mirsky, 51; Mazia, 52.) Chromo- somes consist mainly of nucleoprotein. The bulk of the nucleic acid is of the desoxypen- tose type (DNA), which normally occurs only in chromosomes. Variable amounts of pentose nucleic acid (PNA) have been de- tected cytochemically and in isolated chrom- osomes. The basic proteins of chromosomes, the histones and protamines of certain sperm nuclei, have been known for a long time. The presence of more complex, non-histone type proteins was first demonstrated by the Stedmans (43) and by Mirsky and Ris (47). These four groups of compounds are the major components of chromosomes. Lipids and carbohydrates may be present in nu- clei, but have not been localized in the chromosomes themselves. Of inorganic con- stituents besides the phosphorus of nucleic acids, magnesium and especially calcium have been found in chromosomes (Scott, ’43). Quantitative determinations on isolated nu- clei and cytochemical studies have revealed the remarkable fact that the amount of DNA is constant per chromosome set (Boivin, Vendrely and Vendrely; Mirsky and Ris, Swift; for references see Swift, 53). The ab- solute amount of DNA in a diploid nucleus is characteristic for an animal and varies considerably from one taxonomic group to another (Mirsky and Ris, ’51). In contrast to the DNA, the amounts of non-histone protein and PNA seem to vary a great deal with physiological conditions and from one tissue to another (Mirsky and Ris, *49). 95 Changes in the composition of chromo- somes during the mitotic cycle have been studied so far only cytochemically. Quanti- tative determinations by histospectropho- tometry have been hampered by the uneven and changing distribution of the absorbing material. The recently developed methods of Ornstein (52) and of Patau (52), however, have overcome this difficulty and accurate determinations of DNA (with the Feulgen reaction) are now possible through the mi- totic cycle. Thus Patau and Swift (53) have demonstrated that the amount of DNA in the chromosomes does not change from earli- est prophase up to metaphase. If the quan- tity of DNA per chromosome is constant, any increase in this substance would have to be associated with the reproduction of chro- mosomes. How cytochemical determination of DNA can be used to determine the time of chromosome duplication will be shown later. Changes in PNA content of chromosomes during mitosis were reported by Schultz (’44), Brachet (742), Kaufmann (748), Turchini (49), Battaglia and Omodeo (’49), and Jacobson and Webb (’52). These investigators found marked changes in the staining of chromosomes with certain basic dyes before and after digestion with ribonuclease, espe- cially in metaphase and anaphase, and con- cluded that there was an increase in PNA in chromosomes in late prophase followed by a decrease in telophase. According to Jacobson and Webb (52) the loss of ribonu- cleoprotein from anaphase chromosomes is accompanied by an increase of this material in the spindle area between the separating chromosomes, suggesting that chromosomes give off ribonucleoprotein during anaphase movement. The findings are supported by the photographs of chick fibroblasts in ana- phase taken by ultraviolet light (at 2650 A), which show increased absorption at this wave length between the separating chromo- somes (Davies, 52). The ‘chromatin elimina- tion” at anaphase of the first meiotic division in eggs of Lepidoptera (Seiler, 14) and some other insects (Cooper, ’39) involves shedding of nucleoprotein material from chromosomes in discrete bodies that remain in the equa- torial plane of the spindle and later disin- tegrate (Ris and Kleinfeld, 52). It remains to be seen whether the loss of ribonucleopro- tein from anaphase chromosomes is a gen- eral phenomenon and what physiological significance it may have. Changes in the proteins of chromosomes during mitosis were first described by Cas- persson (740) on the basis of ultraviolet spec- 96 trophotometry of spermatocytes from a grass- hopper (as a cell in division) and salivary gland chromosomes of Drosophila (as a rest- ing cell). Unfortunately the quantitative de- termination of different types of proteins in chromosomes by this method is highly ques- tionable. Until confirmed by other methods and by analysis of individual chromosomes through the mitotic cycle (instead of comparing prophase or resting nuclei with metaphase chromosomes), Caspersson’s scheme remains an interesting speculation. A decrease in protein content of chromo- somes from interphase to metaphase was also suggested by Pollister (751) on the basis of comparison of Millon-stained interphase nu- clei and metaphase chromosomes. So far our information on the chemical composition of chromosomes during mitosis is thus very fragmentary. Certainly we are not justified in generalizing from one type of cell. It has long been known that metaphase chromo- somes may vary in size in different tissues of the same animal (cf. Geitler, ’38; Biesele, 46), under different physiological conditions (tissue culture, e.g., Hance, ’26) and in tu- mors (Biesele, 47). The most striking exam- ple is the decrease in the size of metaphase chromosomes during cleavage of many forms (Erdmann, ’08). If we are correct in assum- ing that the amount of DNA is constant, the variations in volume must be due mainly to proteins. The protein content of metaphase chromosomes may thus be just as variable as it was found to be in interphase chromo- somes. A number of cytologists have inferred chemical changes in mitotic chromosomes from their visual appearance after staining and have constructed ambitious theories on such information (Darlington, °42; Serra, 47). The recent quantitative cytochemical studies have shown the dangers of overex- tended deductive reasoning that has been so common in cytology. To be productive, in- ventive speculation needs reliable empirical data as a basis. Chromosome Reproduction. The duplica- tion of the chromosomes is perhaps the fun- damental process of mitosis and one of the most interesting problems of biology. In what stage of the mitotic cycle does it take place? At a time when the chromosome was considered to be a single fiber that divided longitudinally during mitosis it was appro- priate to speak of the time of chromosome splitting and this was often identified with the time of gene reproduction. It appears now that we must distinguish several processes that may be quite independent: (1) Syn- CELLULAR STRUCTURE AND ACTIVITY thesis of the essential chemical components of the chromosome; (2) duplication of the submicroscopic elementary microfibrils. These two processes together shall be called “chromosome reproduction”; (3) subdivision of the bundle of microfibrils into the units that separate at anaphase, or that behave independently with regard to x-ray breakage or genetical exchange (crossing over) or be- come resolved in the UV or light microscope. We might call this splitting of the chromo- some into chromatids, half- or quarter-chrom- atids, etc. (cf. Fig. 13). While the first one involves chemical synthesis and is presum- ably irreversible, the second process is often reversible. In the past the time of chromosome split- ting has mainly been discussed. But this is clearly of secondary interest and the funda- mental process is chromosome reproduction. With the discovery of DNA constancy in chromosomes a study of this question became at least partially possible. Chromosome re- production must involve DNA synthesis and the time in the mitotic cycle when this hap- pens can be determined (for a review see Swift, °53). Ultraviolet spectrophotometry has been used (Caspersson, ’39; Walker and Yates, 52) but since this is not specific for DNA the results are not decisive. The most reliable information comes from absorption measurements on Feulgen stained chromo- somes. Extensive measurements on animal and plant cells by Swift, Alfert, and Patau and Swift have established that the increase in DNA takes place in interphase, before visible changes in the nucleus occur; earlier measurements by Ris (47) must now be con- sidered in error (for references see Swift, 53). Unfortunately there are no good cri- teria to subdivide interphase into the differ- ent physiological phases which we know must exist. In a diploid tissue with dividing cells there are generally two classes of nuclei. One contains twice as much DNA as the other, in preparation for mitosis. The doubling in DNA content, however, may also be asso- ciated with endomitosis resulting in poly- teny or polysomaty. The doubling of DNA in interphase just before mitosis was demon- strated in a very different way by Price and Laird (’50) in an analysis of liver regenera- tion after hepatectomy. The DNA content was determined chemically and from the number of nuclei in the homogenate the amount per average nucleus was calculated. In the first 12 to 24 hours the DNA per average nucleus approximately doubled. Af- CELL DIVISION ter that mitoses appeared and the average amount of DNA per nucleus returned to the value of normal liver. Another interesting property of DNA sug- gested an independent approach to the prob- lem. Experiments with radioactive phos- phorus (P*?) had shown an extremely low turnover of the phosphate of DNA in tis- sues with no or few dividing cells (e.g., Hevesy, 48). Apparently P%? is incorporated into DNA only when this is synthesized dur- ing chromosome reproduction. Howard and Pele (51a) using an ingenious radioauto- graphic technique of high resolution studied the uptake of P*? into the DNA of chromo- somes during mitosis in the root tips of Vicia. Their results are in complete agreement with the cytochemical and chemical studies. P*? is incorporated into chromosomes only during interphase before mitosis, not during mitosis proper or in interphase of differentiated non- dividing cells. In disagreement with the work discussed so far is the view held by Lison and Pasteels (see Swift, °53), who made absorption meas- urements on Feulgen-stained nuclei of the rat, chicken and sea urchin embryo. They claim that DNA doubling takes place in telophase immediately after formation of the nuclear membrane. This may be true in some rapidly dividing cells, but the evidence so far is against this being the general situa- tion. In the sea urchin, tor instance, Mc- Master (see Swift, *53) found that DNA doubling takes place early in interphase during the first cleavage divisions. Later on the DNA synthesis occurs during interphase at rates that vary among different blasto- meres. The DNA content of interphases in different cells thus differs but is intermediate between the telophase and metaphase value. With regard to the synthesis of other com- ponents of chromosomes we are much less fortunate. Is there a specific protein in chromosomes that is synthesized only at the time of chromosome reproduction? Is it made at the same time as DNA? From ex- periments of Howard and Pele (751b) on incorporation of radioactive sulphur (S*°) into chromosomal proteins it looks as if S* is taken up into proteins mainly by dividing nuclei and at about the same time as P*. But no general conclusions can be drawn from these preliminary results. The most interesting aspect of mitosis is, no doubt, the question of how the cell pro- duces an exact copy of the infinitely com- plex and specific organization of the chromo- somal fiber. This involves not only the syn- 97 thesis of specific proteins and nucleic acids, but also the weaving together of the proper components into the right patterns. The syn- thesis of nucleic acids and proteins is today intensively investigated by the biochemist, who may soon give us a better understanding of what goes on during this stage in the life of the chromosome. Models of chromo- some duplication were suggested by Fried- rich-Freksa (’40), Bernal (40) and Delbriick (41). The precursors of the new DNA gen- erally appear to be small molecules. How- ever, in the frog’s egg Zeuthen and Hoff- Jérgensen (752) found that the cytoplasm contains considerable amounts of desoxyribo- sides. Apparently they are stored in the egg cytoplasm to be used during the rapid chromosome synthesis of early cleavage. The total content in desoxyribosides of the egg remains constant until the late gastrula stage. The Chromosomes in the Interphase Nucleus. During telophase the chromosomes swell into optically homogeneous vesicles (cf. Lewis, 47). If these are widely spaced on the spindle each chromosome vesicle may form its own nuclear membrane and as a result a large number of small nuclei appear (karyo- meres). This is common in cleavage divi- sions (Fig. 10H). Karyomeres may contain one or several chromosomes. In general, how- ever, the chromosomes are so close together at telophase that a single nucleus is formed. The nuclear membrane originates at the interphase between chromosome and cyto- plasm and has a complicated structure (Monné, *42; Schmidt, ’39). In the electron microscope two layers can be distinguished (Callan and Tomlin, ’50; Bairati and Leh- mann, °52; Harris and James, ’52). One is mainly lipoid in nature and contains regu- lar perforations. The other has a uniform structure and consists apparently of lamellae of fibrous proteins. The porous layer may be on the outside (amphibian oocyte, Callan and Tomlin, 50) or inside (Amoeba, Harris and James, 52). The changes taking place in the chromo- somes are imperfectly understood. In part a despiralization of the chromonemata takes place. Probably there is also a change in chemical composition, especially an increase in chromosomal proteins. Furthermore, a change in the physical state of chromosomes can be demonstrated. Chromosomes are or- ganized nucleoprotein gels which can swell and contract. In the living interphase nu- cleus they are in the extended state and usu- ally fill the nucleus completely so that no 98 boundaries between chromosomes or chro- monemata are visible. In such nuclei only nucleoli and sometimes heterochromatin are visible (Ris and Mirsky, ’49). This compact nucleus is found in many animal and plant tissues. The granular fixation image is due to the shrinking of the chromosome gel in most fixatives. In some cells the chromonemata appear to be less closely packed and may become visi- ble in the phase microscope (tissue culture of mouse kidney, spleen, heart: Fell and Hughes, ’49). Such nuclei may contain in addition to the extended chromosomes some other material (karyolymph or nuclear sap). The extreme development in this direction is found in the germinal vesicle of many eges (for instance, Amphibia) where the chromosomes finally make up only a small fraction of the nuclear volume. The con- tent of such nuclei may have a _ viscosity little different from that of water (Gray, Dia). At present the following points appear to be established: (1) Chromosomes persist as individuals in the resting nucleus. (2) They have a relatively loose structure: except in the heterochromatin the chromonemata are uncoiled and chromatids and half-chromatids are often less closely appressed (see Mar- quardt, 41). (3) The chromosome material (nucleoprotein gel) is in an extended state and in many nuclei fills the entire nuclear space outside of the nucleoli. (4) Extra-chro- mosomal material (nuclear sap) may be present in the interstices of the swollen chromosome gel, where the chromosomes fill the entire nucleus. In other cells it may separate the chromosomes that now come visible in the living cell (phase microscope) or in rare cases it may increase to make up the bulk of the nucleus (germinal vesicle). THE NUCLEOLUS The nucleolus is an organelle of the rest- ing cell. Chemical and morphological changes indicate that it is involved in cell metabolism but nothing definite is known about its function. During karyokinesis the nucleolus degenerates and is re-formed at telophase in association with a definite re- gion on one or several chromosomes (“nu- cleolar organizer”). Exchange of material between chromosomes and nucleolus has been suggested, but there is no direct evidence for it. The time of dissolution varies from mid- prophase to anaphase or later. If it gets into the spindle it may become divided or moved CELLULAR STRUCTURE AND ACTIVITY to one or the other side and finally into the cytoplasm. In rapidly dividing cells, for in- stance in early cleavage, no nucleolus is gen- erally formed in the interphase nucleus. CHROMOSOME MOVEMENTS Following chromosome reproduction and chromosome splitting normal karyokinesis involves a series of chromosome movements in the course of which the chromosome halves are separated into the daughter nu- clei. Movements inside the Nuclear Membrane. During interphase little movement of the chromosomes occurs. Often they appear in prophase in typical telophase orientation, with the kinetochores of all chromosomes close together (Rabl orientation). In pro- phase, as they become more compact and the amount of extra chromosomal material increases they begin to show some changes in position. In late prophase they often be- come evenly spaced, preferably on the nu- clear membrane. Where the chromosomes are small and compact this spacing is espe- cially clear. We therefore find the best ex- amples in late meiotic prophase. Another type of movement of chromosomes within the nuclear membrane appears to be due to a mysterious relationship between chromosomes (especially chromosome ends) and the centrosome. Some beautiful examples are found in spermatocytes of many insects. In early prophase (leptotene) the ends of all chromosomes come together at one spot opposite the centrosome to form the so-called bouquet stage (cf. Schrader, *53). Later in prophase as the daughter centrosomes move apart, the chromosome ends (or, in case of small chromosomes, entire chromosomes) fol- low along inside the nuclear membrane. The chromosomes are thus separated into two random groups near the centrosome (cf. Hughes-Schrader, ’43a). The same occurs in some somatic cells (whitefish cleavage, Fig. 12; mouse spleen in tissue culture: Fell and Hughes, ’49, Fig. 16). Movements on the Spindle. After the nuclear membrane breaks down the spindle takes shape in the former nuclear area and the further movements of chromosomes are in relation to this cell structure. First the chromosomes are moved into the equatorial plane of the spindle where they become spaced in a regular fashion (metakinesis), then after a certain length of time their halves are moved apart toward opposite spindle poles (anaphase movement). CELL DIVISION The precise and orderly movements of chromosomes during metakinesis and ana- phase have long fascinated cytologists. Many physicochemical explanations have been sug- gested (for a critical discussion see Schrader, 53). Unfortunately they were more often based on artificial models rather than the cell. Before we can attempt to understand what is going on in physical and chemical terms, it is necessary to know just what hap- pens in cytological terms. What are the spe- 99 cial cell structures concerned with the move- ment of chromosomes? What is their history during the mitotic cycle and how do they interact to assure the orderly disjunction of daughter chromosomes? Once these mitotic organelles are recognized they can be investi- gated with regard to their chemical compo- sition, their submicroscopic structure and their function in biochemical terms. We can recognize the following structures as mitotic organelles: cell center (centro- ce) ; =f a pase oe a : SSA ZZ Fig. 14. Mitosis during early cleavage of Drosophila, illustrating the history of the centriole (after Huettner, poo) 100 some, centriole and aster), spindle, kineto- chore, and chromosomal fibers. CENTROSOME. The cells of animals and lower plants contain a self-duplicating cyto- plasmic structure, the centrosome. During mitosis it organizes the aster and plays a part in the polarization of the spindle. In certain cells it also acts as blepharoplast organizing the flagellum of flagellated cells and the axial Fig. 15. Fertilization in Nereis. Amphiaster and spindle of sperm and second maturation spindle of egg. Note the difference between the central spin- dles (C.sp.) and the nuclear spindles (/V.sp.). (After Lillie, ’12.) filament of spermatozoa. It is probably re- lated to the basal body of cilia in ciliated cells (Renyi, ’24) and the kinetosomes of the Ciliata (Lwoff, 50). The centrosome usually consists of a small granule (centriole) that is either round, rod- shaped or V-shaped (for examples see John- son, °31) and is surrounded by a spherical, homogeneous or finely granular area of cyto- plasm. Its history during the mitotic cycle is illustrated in Figure 14, which represents the nuclear division during early cleavage in Drosophila (Huettner, ’33). The centriole divides during mitosis and persists during interphase as a double body (diplosome) (for examples in amphibian tissues see Pol- lister, ’°33). The centrosome divides in pro- phase. Each half contains a centriole and forms an aster. Under favorable conditions the centrosomes and centrioles are visible in the living cell both during mitosis (Fig. 12c) and in the non-dividing cell. The size of centrioles is usually near the limit of resolu- tion of the light microscope (about 0.2 mi- cron) but the rod-shaped centrioles of some insects may be more than 1 micron long (see for instance Johnson, ’31). Changes in size and stainability during the mitotic cycle (for CELLULAR STRUCTURE AND ACTIVITY instance Jorgensen, °13; Johnson, °31; Chick- ering, ’27) in tissue culture and in tumor cells (Ludford, ’25) have been reported. The chemical composition of the centriole is un- known. During prophase the aster develops around the centriole. The aster rays are posi- tively birefringent gel fibers (cf. Inoué and Dan, *51) and are anchored in the cortical gel. They can be moved around and bent with the microdissection needle (Chambers, 17). Often they are visible in the living cell, especially where the cytoplasm contains many granular inclusions that contrast with the clear aster rays, or where filamentous mitochondria become oriented between the aster rays (Fell and Hughes, ’49). According to Chambers, aster rays are hollow canals but this has not been confirmed by other in- vestigators. More likely the observed flow of less viscous cytoplasm occurs between the aster rays. The area around the centriole is usually free of cytoplasmic inclusions and increases in size from prophase to anaphase. Aster-like structures are sometimes present also in the cytoplasm of non-dividing cells such as leukocytes, mesenchyme cells and other cells in which the centriole is near the center of the cell. Accumulation of hyalo- plasm around the centriole in such cells was observed by Lewis (’20). Apparently the centriole influences struc- ture and orientation in the cytoplasm not only during division but also in the resting cell, possibly by setting up diffusion currents as Pollister (41) has suggested. The rays between two asters usually con- nect so that a spindle-like structure is formed (amphiaster). This purely cytoplasmic struc- ture has been called the “central spindle,” but is entirely different from the real mitotic spindle in which chromosome movement takes place. The “central spindle” is a system of aster rays, and cytoplasmic inclusions are free to move through it between the rays (Fig. 15). Chemically the asters contain, in addition to protein, pentose nucleic acids (Brachet, ’42; Pollister and Ris, ’47; Stich, 51a) and polysaccharides (Monné and Slautterback, 90: Stich, -51'b). Supernumerary asters (cytasters) appear in the cytoplasm of some invertebrate eggs after certain experimental treatments (cf. Wilson, 28). This fact has sometimes been used as evidence for a de novo origin of centrioles. However, it has not been estab- lished that cytasters contain centrioles and the possibility thus exists that asters may arise independently of real centrioles. CELL DIVISION SPINDLE. Regular movements of chromo- somes are possible only in the presence of a spindle. This interesting structure originates from the prophase nucleus. In most cells the entire non-chromosomal material of the nu- cleus seems to transform into the spindle, in E 101 others only part of the nuclear material is used (Fig. 10A and B, whitefish blasto- mere). Cytochemical evidence indicates that spindle material appears in the nuclear sap during interphase in cells that are preparing for division (Stich, 51a, b). The size of the Fig. 16. Types of mitotic spindles. A, First maturation metaphase in egg of Artemia (after Gross, 35). B, Cleavage of Artemia (after Gross, ’35). C, First spermatocyte metaphase in a coccid, Llaveia bouvari (after Hughes-Schrader, ’31). D, Oogenesis of Acroschismus (after Hughes-Schrader, ’24). E, Early ana- phase spindle in the radiolarian Aulacantha (after Borgert, 00). F, Second spermatocyte division of Gossy- paria (after Schrader, ’29). 102 spindle is commonly proportional to the vol- ume of the nucleus from which it originates. This is especially clear during cleavage of many organisms (Conklin, °12) where nu- clei and spindles are large in early divisions and get smaller in late ones. In Pediculopsis the spindle volume decreases from the first to the tenth cleavage about 200 fold (Cooper, °39). Other good illustrations are the large and small spermatocytes of Arvelius (Schra- der, 47). Where more than one nucleus is present within a cell, each forms its own spindle. During early cleavage sperm and CELLULAR STRUCTURE AND ACTIVITY to all spindles is the bipolar organization. This bipolarity is independent of centriole and aster as demonstrated by the cases where centrioles are naturally inactive (for in- stance in oogenesis of Ascaris) or experi- mentally inhibited (Bataillon and Tchou Su, 30). In the living cell the spindle is a gelatin- ous semi-solid body that can be moved about in the cell or even dissected out with the micromanipulator (Chambers, ’24; Carlson, 52). Cytoplasmic granules never penetrate the spindle (which differentiates the true Pen sit. a Fig. 17. The birefringence of the spindle. A, Pollen mother cell of Lilium longiflorum (phot. Inoué). B, Amphiaster and spindle isolated from the blastomere of a sea urchin (phot. Inoué, see Mazia and Dan, ’52). ege chromosomes often remain separate and each group forms its own spindle (gonom- ery). Such independent spindles may fuse into one or remain separate through ana- phase (Hughes-Schrader, ’24). The spindle may become organized inside the nuclear membrane or only after the membrane has dissolved. In some cells the membrane does not disappear until anaphase or even persists throughout mitosis (intranuclear spindles; cf. Drosophila cleavage, Fig. 14). The form of the spindle is very variable; it may be shaped like a disc, a barrel or a spindle, or may be flared at either end, or it may be asymmetrical (Fig. 16). Where a centriole is present the spindle is usually pointed at the ends and terminates at the centriole unless the asters are very large compared to the spindle (Fig. 16B). In animal cells where the centriole is inactive the spindle is usu- ally barrel-shaped (Fig. 16A). But common mitotic spindle from the amphiaster or cen- tral spindle, Fig. 15) and in living cells they can be seen bouncing off the spindle (Ris, °43). At metaphase the spindle is firmly anchored to the asters and the whole spindle apparatus (achromatic figure) can be moved about or even isolated from the cell (Mazia and Dan, ’52 (Fig. 17B). After fixation the spindle generally has a fibrous structure (continuous fibers). In the living spindle this is only rarely visible (Cooper, ’41), but Lewis (23) has shown long ago that a change in pH can make it appear reversibly. Recently Inoué (°52) has demonstrated con- tinuous fibers in living spindles with an im- proved polarizing microscope. Even where spindle fibers are not visible in life they can no longer be regarded as artifacts; they are an expression of the basic organization of the spindle. In the polarizing microscope the spindle shows a positive birefringence with CELL DIVISION regard to its long axis (Schmidt, ’39; Swann, 51; Inoué, 52). This indicates that it consists of elongated submicroscopic units, macro- molecules or micelles, that are oriented paral- lel to the spindle axis (Figs. 17A and B). The similarity of the spindle to the tactoids formed in suspensions of elongated macro- molecules (for instance, tobacco mosaic virus) has suggested that the spindle also is a tactoid (Freundlich; Bernal; see Swann, 52). There are, however, fundamental differences between tactoids and the spindle and they were rightly emphasized by Swann (752). In a tactoid the particles are oriented and held together by long-range ionic forces and the antagonizing action of these with surface tension causes their spindle shape. In the mitotic spindle, however, the micelles must be held together also by chemical bonds or else the spindle could not be fixed or iso- lated intact from the living cell. The pres- ence of S—S linkages is suggested by the observations of Mazia and Dan (’52). Ferry (48) has recently reviewed various types of protein gels and the forces involved in their formation. Perhaps the spindle has properties in common with both tactoids and gels of denatured proteins. Electrostatic forces would be mainly involved in the orientation of the micelle into a_ bipolar structure, while chemical bonding at cer- tain points would give it the observed rigid- ity. The appearance of the spindle in the elec- tron microscope depends on fixation (Rosza and Wyckoff, 51; Beams et al., ’50a,b; Sedar and Wilson, *51). After Formalin fixation the spindle looks quite homogeneous, but if acid fixatives are used definite fibers be- come visible. This suggests that the struc- tural units in the spindle are submicroscopic and less than a few hundred A thick, but that they have the property to bunch to- gether, possibly depending on the degree of hydration, and thus form fibers that are visible in the light microscope. The behavior of the spindle under in- creased hydrostatic pressure indicates that it is similar to other protoplasmic gels and myosin, with endothermic gelation reaction and increase in volume upon gelation. It is destroyed by a short exposure to hydro- static pressure of 5000 to 6000 Ibs. per square inch (Pease, °41, ’46; Marsland, ’51). Rather little is known about the chemical composition of the spindle. The most prom- ising advance is the recent development of methods to isolate large cleavage spindles in quantity for chemical study (Mazia and 103 Dan, *52). The bulk of the isolated cleavage spindles and asters of sea urchin eggs was found to consist of a protein that formed a single boundary in the analytical ultracen- trifuge. The molecular weight of the particle was calculated to be approximately 45,000. In addition to protein, cytochemical studies indicate the presence of PNA (Brachet, °42; Pollister and Ris, 47; Stich, *5la) and vari- Fig. 18. Cleavage spindle of Steatococcus, showing diffuse kinetochore. One of the two chromosomes has been broken in two by x-rays. (After Hughes- Schrader and Ris, ’41.) able amounts of polysaccharides (Monné and Slautterback, *50; Stich, *51b) im some but not all spindles. The finding of Brachet that nuclear sap, spindle and aster of amphibian eggs and insect testes contain proteins rich in —SH groups is of special interest in view of Rapkine’s theory on the role of —SH groups and reversible denaturization of pro- teins in the formation of gel structures dur- ing mitosis (reviewed in Brachet, ’50). KINETOCHORE (centromere, spindle attach- ment). Chromosomes do not move in the spindle unless they are attached to it by chromosomal fibers (traction fibers). In most organisms these fibers originate in connec- tion with a specialized region of the chromo- some, the localized kinetochore. In certain animals and plants chromosomal fibers at- tach along the entire length of the chromo- some (diffuse kinetochore, see Fig. 18). Where the kinetochore is localized, chromo- some fragments lacking this organelle fail 104 to form chromosomal fibers and do not move on the spindle (akinetic fragments, Carlson, °38). In the case of the diffuse kinetochore any piece of the chromosome becomes at- tached to the spindle and moves normally (Hughes-Schrader and Ris, 41). The diffuse kinetochore is found in certain insects—the d Fig. 19. Structure of the kinetochore. A, Meiotic chromosomes of Trillium (after Matsuura, ’41); ch. = chromonemata, K.M. = kinetochore matrix, a = metaphase, b-d = early anaphase. B, Meiotic chromosome of Amphiuma (after Schrader, ’39); ch.f. = chromosomal fiber, K.S. = kinetochore spherule. C, Pachytene chromosome of rye (after Lima-de-Faria, 49); K. = kinetochore, c.chr. = centromeric chromomeres. Hemiptera, Homoptera and probably also Odonata (Oksala, ’43) and Lepidoptera (Su- omalainen, ’53) in a myriapod (Ogawa, ’49), a few scorpions (Rhoades and Kerr, °49), and in a group of plants (Malheiros, de Cas- tro and Camara, ’47). The kinetochore in Ascaris, often described as multiple, probably also falls into this category. New insight into the nature of kineto- chores may come from a further study of the accessory kinetochores (neocentric re- gions) that turned up in some strains of rye and maize (reviewed by Rhoades, ’52). During meiotic divisions secondary chromo- somal fibers develop in some of the chromo- somes in addition to those formed by the CELLULAR STRUCTURE AND ACTIVITY regular kinetochores and prematurely pull the chromosome ends to the poles. Especially interesting is the observation that in maize these neocentric regions form chromosomal fibers only if they are in physical connec- tion with the primary kinetochore (Rhoades, 62). The microscopic structure of the kineto- chore is still imperfectly understood. The clearest photographs are those of the kineto- chore in Trillium, published by Matsuura (41). It appears to be a section of the chromonema that remains uncoiled. On the spindle it is surrounded by a hyaline material (kinetochore matrix) that divides in early anaphase (Fig. 19A). In chromosomes of rye, onion and other plants the kinetochore was shown to be an uncoiled region of the chromonema with a pair of “centromeric chromomeres.” It has been suggested that the doubleness is due to an inverse repeat (Fig. 19C) (see Lima-de-Faria, ’49, ’50). It is probable that such special “chromo- meres” or heterochromatic knobs of the ki- netochore region in the chromonema are identical with the “spindle spherule” which has been described in chromosomes of sev- eral plants and animals (Fig. 19B, cf. Schrader, ’39). Normally the kinetochore divides lengthwise like the rest of the chrom- onema. A number of cases are known, how- ever, where it divides transversely (mis- division), giving rise to isochromosomes (for instance, Miintzing, *44). The kinetochore may occasionally be broken into two func- tional fragments (McClintock, °32). Such terminal kinetochores, however, are usually not stable (Rhoades, *40). CHROMOSOMAL FIBERS. Of all the mitotic or- ganelles, chromosomal fibers are most di- rectly involved in chromosome movement. They develop between the kinetochore and the centrosome or spindle pole. In some in- stances they may form in the absence of an organized spindle (Peters, ’46; Rhoades, 52; Scott, ’°36; Pease, ’41). Under favorable con- ditions they are visible in the phase micro- scope in living cells (whitefish, Fig. 12C) (Fell and Hughes, ’49; Tahmisian, °51). After fixation they appear as a bundle or a sheet of fibers thicker near the kinetochore and tapering toward the poles. They stand out distinctly in the polarizing microscope, also in the living cell (Inoué, ’52), owing to their strong birefringence that is positive with respect to their long axis (Fig. 17A). Like the spindle body they have been re- garded as either positive or negative tactoids (Bernal, 40; Ostergren, °49). Just as in the CELL DIVISION spindle, however, we must assume some chemical bonding between the micelles even though these bonds are weak enough to be easily broken and reformed, so that a chrom- osome or a microdissection needle can move through a chromosomal fiber without de- Fig. 20. Pollen mother cell of Lilium, centrifuged (arrow indicates direction of centrifugal force) (after Shimamura, *40). The chromosomal fibers anchor the chromosomes to the spindle in the upper pole. The chromosomes are uncoiled by the centri- fugal force. stroying it permanently (cf. Ostergren, ’49). The centrifuge experiments of Shimamura (40) are impressive evidence for the strength of the chromosomal fibers (Fig. 20). Still too many questions remain completely un- answered. What is the material forming these fibers? What is its origin, and what is the role of the kinetochore? What forces cause their orientation to opposite spindle poles? How are they anchored on the chromosome and on the spindle? How do they function as they move the chromosomes? At present it seems most plausible to consider them as a contractile, fibrous gel reversibly liquefied with hydrostatic pressure (Pease, °46). Some investigators have considered chrom- osomal fibers to be a liquid material se- creted by the chromosomes and adsorbed on the continuous fibers of the spindle. They are thought to pull the chromosomes along the spindle by means of surface forces (Bélat¥ and Huth, ’33; Kupka and Seelich, "48). However, surface forces could hardly keep the chromosomes attached to the spindle with the centrifugal forces used (Beams and King, °36) or against the pull of the microdissection needle (Carlson, °52). After this discussion of the mitotic or- ganelles we must now see how they effect the congression of the chromosomes into the 105 metaphase plate and the anaphase move- ment to the spindle poles. Metakinesis. The chromosome movements of metakinesis result in a regular arrange- ment of chromosomes in the equatorial plane of the spindle (metaphase plate). To accom- plish this the following conditions must be fulfilled: (1) a bipolar spindle must be present; (2) the chromosomes (mitosis) or chromosome pairs (meiosis) must be attached to opposite spindle poles through chromo- somal fibers. If the centrosome fails to divide in pro- phase a single large aster may appear (mon- aster). In Urechis, Bélat and Huth (’33) found that chromosomes orient in the aster and form chomosomal fibers toward and away from the center. But no metaphase plate is formed. In microspore mother cells of certain hybrids or haploids in plants the univalents usually do not form chromosomal fibers and do not congress on the metaphase plate. In Fig. 21A we see the spindle in a pollen mother cell of haploid Datura. No chromosomal fibers are present and the uni- valents are scattered over the spindle. In the diploid the bivalents produce oriented chro- mosomal fibers and a metaphase plate is present (Fig. 21B). Where chromosomal fibers point to one pole only no metaphase plate is formed, as in Sciara (Metz, ’33) and Micromalthus (Scott, ’36). In spermatocytes of many insects (e.g., grasshopper) the uni- valent X-chromosome is attached to one pole only and does not go on the metaphase Fig. 21. First meiotic division im microspore mother cell of Datura. A, Haploid Datura; continu- ous fibers, but no chromosomal fibers are present; no metaphase plate is formed. B, Diploid Datura, chromosomal fibers present; chromosomes have moved into the metaphase plate. (From slide of Dr. Satina. ) plate. In tetraploid spermatocytes of the mantid, Callimantis, the X-chromosomes lie on the metaphase plate if they are paired and have chromosomal fibers to opposite poles. Where pairing is absent, each uni- valent has a single chromosomal fiber to one 106 pole and remains outside the metaphase plate (Hughes-Schrader, ’43b). Congression into the metaphase plate thus seems to be accomplished by the chromosomal fibers. Through them the chromosome or bivalent becomes attached to opposite poles of the spindle and the tension on the chromosomal fibers moves the chromosomes until equi- librium is reached in the metaphase plate (cf. Schrader, ’47). In living cells such a movement of chromosomes back and forth in the long axis of the spindle has often been described and is easily seen in most films of dividing cells. The pre-metaphase stretch ob- served in spermatocytes of several insects SPINDLE LENGTH CHROMOSOME SEPARATION CHROMOSOMAL FIBERS A MIN. CELLULAR STRUCTURE AND ACTIVITY pulsive forces can produce similar patterns, but tell us nothing about the forces involved in the spindle. Most likely a complex inter- action of chromosomal fibers, electrostatic charges on chromosomes, and intermolecular attraction between spindle micelles, tending to crowd out foreign bodies (Ostergren, °51), is responsible for the metaphase arrange- ment. Anaphase Movement. Of all the various aspects of mitosis hardly any has attracted the attention of cytologists more than the strikingly regular movement of chromo- somes at anaphase. For years it has been the subject of much speculation and some experi- SPINDLE LENGTH CHROMOSOME SEPARATION CHROMOSOMAL FIBERS B MIN. Fig. 22. Curves of chromosome separation and spindle elongation. A, In forms with diffuse kinetochore (Hemiptera and Homoptera; after Ris, 43. B, In forms with localized kinetochore (grasshopper spermato- cytes, Ris, °49; chicken fibroblasts, Hughes and Swann, *48; Hughes and Preston, °49). (Hughes-Schrader, 43a) and the metaphase position of multivalents give further sup- port to this hypothesis (Ostergren, ’51). The details and mechanisms of this process, how- ever, are completely unknown. What causes orientation of the kinetochores in the spindle and assures that chromosomal fibers attach to opposite poles? How can we explain the different behavior of kinetochores in mitosis and meiosis? (For a stimulating and interest- ing discussion of these problems see Oster- gren, °51.) Another interesting aspect of metakinesis is the spacing of chromosomes in the metaphase plate. The chromosomes are either all on the periphery of the spindle, the arms of long chromosomes directed radi- ally away from the spindle, or they are evenly spaced in the equatorial plane. Even then the radial arrangement of chromosome arms is often striking. Large chromosomes are usually near the periphery, smaller ones in the center. In different cells even of the same organism this metaphase arrangement shows often striking variations and may show a constant pattern characteristic for the type of cell (Wilson, ’32). Model experiments with floating magnets show that a balance of attractive and re- mental analysis. (For a critical discussion of the various hypotheses, see Schrader, *53.) The initial separation of the chromatids takes place also in colchicine-treated cells and is therefore independent of the spindle apparatus (Levan, °38) and may be due to a swelling and dissolution of some material that holds the chromatids together at meta- phase (cf. Carlson, 52). For the movement to the poles, however, spindle and chromo- somal fibers are indispensable. Analysis of chromosome movement in living cells has shown that two factors are involved. The pole-ward movement is correlated with a shortening of the chromosomal fiber. Just how this occurs is not known but probably some kind of contraction of the organelles is involved (see review by Cornman, ’44). The second factor is the lengthening of the spindle. The spindle suddenly swells in the equator and then stretches in its long axis. Since the chromosomes are attached to the spindle poles they are thus further moved apart. These two components in the ana- phase movement are sometimes separated in time (Fig. 22A)—so far this has been found only in animals with diffuse kinetochore (Ris, °*43)—but they usually occur simul- CELL DIVISION taneously (Fig. 22B) (Ris, ’49; Hughes and Swann, °48; Hughes and Preston, ’49). In this case, it is possible to inhibit spindle stretching without affecting the chromo- somal fibers, demonstrating the relative in- dependence of the two components also where they act simultaneously (Ris, ’49). In some cells spindle stretching may be absent (the common situation in somatic 107 zonal” fibers. Little is known about their origin, but it is possible that their appear- ance is related to the shedding of ribonu- cleoprotein from anaphase chromosomes that was mentioned above (Ris and Kleinfeld, D2 )e The future study of anaphase movement will have to be directed mainly toward an experimental analysis of the mechanisms of Taste 3. Maximum Rates of Chromosome Movement and Spindle Elongation during Anaphase ea ar ns ee CHROMOSOME MOVEMENT * MICRA /MINUTE SPINDLE ELONGATION MICRA /MINUTE AUTHOR SS CELL TEMPERATURE, Or ‘Tamalia Embryonic cell 26 Spermatocyte I 26 Spermatocyte IT 26 Protenor Spermatogonia 26 Spermatocyte I 26 Thelia Spermatocyte I 26 Chortophaga Spermatocyte I 30 Spermatocyte II 17 23 30 Triton fibroblasts, 26 tissue culture Rana fibroblasts, 26 tissue culture Xenopus fibroblasts, 26 tissue culture Gallus osteoblasts, 40 tissue culture 0.7-2 0.3-1.1 Ris, °43 —t 0.3 029=1-2 OFA—len 1.3-1.6 0.3-0.5 Ris, 43 ORS 0), 7/ 0.4 0.5 Ris, °43 15 3 Ris, ’49 0.4 1.4 2 2.4 25 3R6 SS __- Sot Hughes and Preston, °49 4.5f Davie 4 2 Hughes and Swann, ’48 * Due to contraction of chromosomal fibers only. {7 Chromosomal fibers do not contract here. ¢ Combined rate not analyzed into contribution of chromosomal fibers and spindle elongation. cells of plants); in others the chromosomal fibers do not contract and anaphase move- ment is due to spindle stretching alone (Ris, *43). Usually the chromosomes all move simultaneously, but cases of autono- mous movements of chromosomes are known. This independent behavior of chromosomes is based on the autonomy of the kinetochore and chromosomal fibers. As the chromosomes approach the poles the region of the spindle in between may re- main a semisolid structure (for instance, grasshopper spermatocytes) or it may solate and disappear. This is demonstrated by a decrease in viscosity (Carlson, 46), by the penetration of cytoplasmic granules into this space (Ris, 43), and by the disappearance of birefringence (Swann, *51). Chromosomes at anaphase are often connected by “‘inter- chromosomal fiber contraction and spindle elongation. The presence in the spindle of phosphatases that split ATP (Biesele, °49) suggests a possible role of high energy phos- phate bonds and a certain similarity to the contractile processes in muscle and myosin gels (cf. Brachet, 50; Hayashi, *52). The velocity of chromosome movement as determined in living cells is largest in the early part and gradually decreases. The max- imum rates of chromosome movement and spindle elongation in a number of cells are given in Table 3. CYTOKINESIS Following the separation of the chromo- some halves the cytoplasm subdivides to complete mitosis (for a general review see 108 Miihldorf, 51). In animal cells this is com- monly accomplished by the cleavage furrow, a circular groove in the cell surface that gradually deepens and cuts the cell in two. What determines the formation of this fur- row and its position in the cell, and how is the formation accomplished? The experi- mental analysis so far indicates clearly that several factors are involved and that different ones predominate in different cells. Elongation of the Spindle. As a general rule cleavage depends on the elongation of the cell (mitotic elongation, cf. Churney, 36), and the furrow is formed at a right angle to this elongation. In most tissue cells where asters are relatively small this elonga- tion is dependent on the stretching of the spindle. Elongation by itself does not assure a cleavage furrow, but where it is sup- pressed the cell does not divide. If the spindle is prevented from stretching in its long axis, for instance through sticking of chromo- somes, it may spread out laterally and stretch the cell at a right angle to the normal axis. In this case a furrow appears vertically to the new direction of elongation and splits off an anuclear bud (Ris, 49). Cell elonga- tion may be accomplished independent of the spindle by centrifugation. Irrespective of the orientation of the cell the cleavage fur- row again cuts through the narrow region (Harvey, 35). Function of the Amphiaster. In cells with large asters, such as blastomeres of many eggs, mitotic elongation and cleavage may occur in the absence of a spindle (Fank- hauser, ’34; Harvey, ’36; Briggs et al., 51). In such cells the growing amphiaster is probably responsible for mitotic elongation (Gray, ’27b). The role of the aster for the initiation of the cleavage furrow was stressed by Dan (48). By studying movements of the cell surface with kaolin particles he found that the surface gradually stretches during cleavage except in the region of the furrow, where it first contracts and then ex- pands. He explains this by assuming that aster rays are anchored in the cell cortex and cross in the equatorial region. As the asters move apart in anaphase, owing to spindle elongation, the aster rays pull in the surface of the equatorial ring, thus initiating the furrow. The increased birefringence dur- ing anaphase of the aster rays crossing in the equator lends support to this hypothesis (Inoué and Dan, °51). Function of the Cell Cortex. Certain observations indicate that the cleavage fur- row may be formed independently of both CELLULAR STRUCTURE AND ACTIVITY spindle and aster and thus suggest a defi- nite autonomy ot this structure. Painter (718) round that in sea urchin eggs treated with phenyl urethane cleavage occurs in the ab- sence of asters. Harvey (735) displaced the amphiaster to one side by centrifugation and Carlson (752) did the same by micromanipu- lation, without affecting the position of the cleavage furrow. According to Marsland (51) the cell cortex increases in viscosity before cleavage and the cleavage furrow is part of the cell cortex that is particularly thick and more solidified. The study of plasmolysis in the sea-urchin egg by Mon- roy and Montalenti (47) also indicates that the viscosity of the cortex is low in metaphase and high before cytokinesis. Like other cy- toplasmic gels the cortex can be solated by increased hydrostatic pressure which thus inhibits cleavage or reverses it, if in prog- ress. Wilson (’51), on the other hand, claims that in Chaetopterus the cell cortex decreases in viscosity during division. The gel nature of the cleavage furrow was demonstrated dramatically by Chambers (738), who de- stroyed one half of the dividing egg and found the furrow remaining intact. He also showed that only egg fragments containing cortical material can divide. In the am- phibian egg Schechtman (737) has studied the cleavage furrow in detail and concluded that it originates as a localized growth of the cortex toward the egg interior. His view is supported by Waddington’s experiments on the frog egg (52). Here the furrow can grow and deepen even if it is isolated from the egg interior by a cellophane strip. Changes in the cell cortex are also in- dicated by the “bubbling” so evident in many films of dividing cells. It is most pronounced near the spindle poles and may be due to a thinning of the cortex in that region. The progress of the cleavage furrow is accom- panied by actual contraction of the ring of cortical gel (Lewis, *51). Such contraction has also been observed in the furrow of the sea urchin (Scott, ’46). The Cleavage Substance. Cornman and Cornman (’51) have suggested that as the membrane dissolves a substance is released from the nucleus that spreads to the cor- tex and initiates the furrow. Similar ideas have been expressed by others, for instance Dalcq, Costello, Beams, and Conklin (for references see Cornman and Cornman, 751). The observation that cytasters develop only after breakdown of the germinal vesicle (Yatsu, 05), and the fact that asters are necessary for division of anucleate cells, sug- CELL DIVISION gest that nuclear material somehow en- hances gelation in the aster forming material and in the cortex. The whole question of what is released by the germinal vesicle into the cytoplasm needs further investiga- tion. A particulate fraction that facilitates cleavage has been demonstrated in the egg of the Dendraster (Moore, ’38), in the sea urchin (Harvey, 36), and in the ascidian egg (Reverberi, ’40). These are the major factors that determine cytokinesis. The development of the furrow is a property of the cortical gel, influenced by mitotic elongation, aster rays and per- haps some nuclear substance. The site of the furrow is mainly determined by the dividing nucleus. If the spindle is in the center and the asters of equal size, cytokinesis is equal. Where the asters are of different size (Conk- lin, °17) or the spindle is placed asymmetric- ally in the cell, cytokinesis is unequal. What the detailed mechanisms are by which the cell insures that both daughter cells receive a nucleus whether the spindle is in the middle of the cell or near the surface to one side, we hardly can surmise today. Differential Mitosis. The significance of mitosis is often sought in the formation of two equivalent cells. Yet the problem whether mitosis can produce two funda- mentally different daughter cells may be of equal if not greater interest to the student of development. Is cell division a mechanism for cellular differentiation? Many examples are known where the two offspring of a mitosis have an entirely different and dis- tinctive fate. Often morphogenesis and dif- ferentiation are associated with a specific and constant number of cell divisions which give rise to a determined number of cells, each with its own specific fate. This is well illustrated in the development of the various cells of the lepidopteran wing (Henke, 747; Henke and Pohley, 52). Other examples are found in the determinate cleavage of an- nelids and molluscs. Such cell divisions as- sociated with differentiation are often called differential mitoses. However only in few cases has it been demonstrated that differen- tiation took place at cell division and not through some environmental factors after- wards. A cell division should be called dif- ferential only if it can be shown that it leads to a qualitatively unequal distribution of nuclear or cytoplasmic elements and that this is responsible for the different develop- ment of the daughter cells. Differential divi- sion of the nucleus in somatic cells occurs as an accident (e.g., nondisjunction) or a 109 special adaptation (chromatin-elimination in Ascaris and some Diptera) but has been dis- counted as a general mechanism of differen- tiation. During meiosis the chromosomes are of course segregated differentially, generally in a random fashion. (For a recent discussion of preferential segregation, see Rhoades, 752.) Differential distribution of cytoplasmic ma- terial, on the other hand, has been described in a number of cases, for instance in the determinative cleavage of annelids, molluscs and ascidians. The segregation of cytoplasmic constituents must, however, occur before di- vision, and spindle orientation has to be spe- cific. Nothing is known about the mecha- nisms involved. The recent emphasis on autonomous cytoplasmic units (plasmagenes) has revived the idea of differentiative mitosis (cf. Ephrussi, 51). A decrease in the relative rate of reproduction of autonomous cyto- plasmic units may result in a loss of these from the cell and can thus alter the char- acteristics of the cell (Lwoff and Dusi, ’35; Sonneborn, *46; Ephrussi, 51; Spiegelman, Delorenzo and Campbell, 751). A similar process has been suggested as a possible ori- gin of tumors (Potter et al., ’50). Where identical plasmagenes or their precursors exist in large numbers the random separa- tion during cytokinesis assures their dis- tribution to both cells. They may reproduce either in the interphase or during mitosis. If a specific plasmagene occurs in small num- bers its reproduction has to be synchronized closely with cell division in order not to get lost (plastids in lower plants). CYTOPLASMIC CHANGES DURING MITOSIS Marked physical and chemical changes have been observed in the cytoplasm during mitosis. Rhythmic changes in viscosity were demonstrated in many types of cells by dif- ferent techniques through the work of Heil- brunn on eggs, Carlson in grasshopper neuro- blasts, Zimmermann and Kostoff in plant cells (references in Heilbrunn, 52a). There is general agreement that the viscosity is high in prophase (in eggs it increases after activation: mitotic gelation of Heilbrunn). It then decreases to a minimum in metaphase and anaphase, to increase again before cyto- kinesis. Related to the changes in viscosity is the rounding off observed in elongated epithelial cells and in irregularly shaped cells such as fibroblasts. According to Lettré (51) this is the result of the lowered level of ATP in dividing cells and can be pro- 110 duced experimentally with respiratory poi- sons. In the irregularly shaped cell the cor- tical proteins are in a state of chronic con- traction that needs a high level of ATP. With less ATP irregular amoeboid move- ments result and the cell takes on a spherical shape. Cytoplasmic streaming is often pronounced during mitosis. Especially during anaphase and cytokinesis vortical currents are visible in many cells (cf. Bélat, ’29). Often these currents carry along pigment granules, yolk platelets, mitochondria and other inclusions that accumulate in the equatorial plane or along the spindle surface (e.g., Nussbaum, 02). Considering these changes in the phys- ical state of the protoplasm and the mixing up through cytoplasmic currents we can understand that mitosis is generally antag- onistic to cytoplasmic differentiation and specific functioning of the cell. Not only is there interruption in nuclear functions, but also more or less severe changes in cyto- plasmic organization. Cytoplasmic organelles such as cilia, brushborder, and ergastoplasm often disappear and specific function is in- terrupted (cf. Berrill and Huskins, ’36; Peter, °40). Some cells divide, however, without visible simplification in cytoplasmic organi- zation (Dawson, 40), and many highly dif- ferentiated cells are able to divide mitot- ically. Differentiation is usually accompanied by a decrease in the rate of mitosis, yet the factors responsible for cessation of cellu- lar proliferation are independent of differen- tiation as such. An interesting change in the cytoplasm is the marked decrease in PNA in late prophase and metaphase (Brachet, ’42; Montalenti et al., °50; Battaglia and Omodeo, 49). This may be related to an interruption in nu- clear function during division in view of the report of Brachet (’50b) that RNA in the cytoplasm decreased in enucleated halves of amoebae. According to Mazia and Hirshfield (50), incorporation of P%? into the cyto- plasmic RNA is under nuclear control and it would be interesting to study how this is affected during mitosis. MITOSIS AND METABOLISM In the past the study of the metabolic characteristics of the dividing cell has lagged behind the analysis of the mechanisms of mitosis. Only in recent years have some sig- nificant advances been made in the under- standing of the metabolic processes associated with cell reproduction. The various aspects CELLULAR STRUCTURE AND ACTIVITY of the recent work have been summarized by Brachet (50a), Krahl (50), Zeuthen (751) and Bullough (752). Much of the older work on the metabolism of mitosis is quite mean- ingless because mitosis was treated as a uni- tary process. Mitosis is a chain of individual processes that are quite different in char- acter, and metabolic studies of cell division will make sense only when they are related to the various components of cellular repro- duction. We shall want to know, for in- stance, what changes in metabolism are as- sociated with the initiation of mitosis; are there metabolic pathways specific for divid- ing cells? What processes furnish the energy for chromosome synthesis, for spindle for- mation and chromosome movements, for cytokinesis and synthesis of cytoplasmic com- ponents? How do the metabolic processes during the various phases of mitosis differ in different types of cells? Eggs during cleavage have been a favorite material for metabolic studies, especially those of echinoderms (Krahl, *50). In these eggs division occurs only in the presence of oxygen. Inhibition of respiration to less than 30 per cent of normal blocks mitosis. A cyanide-sensitive system containing an iron- porphyrin catalyst (probably cytochrome c) is involved. Mitosis is also blocked by inter- fering with generation and transfer of en- ergy-rich phosphates. The substrate oxidized during cleavage has not been conclusively characterized. In some eggs, however, oxygen is not necessary for cleavage. For instance in the frog, the toad, Fundulus and Ilyanassa (references in Bullough, 52; Brachet, ’50a), cleavage may continue in the absence of respiration. Several investigators have measured the rate of respiration during cell division. The most careful and extensive determinations have been made by Zeuthen (751), both on single eggs and on large numbers of eggs dividing synchronously. In the eggs of the frog, Urechis and several echinoderms a definite increase in respiration was demon- strated with each mitosis. Zeuthen, further- more, showed that the rise in respiration occurred during interphase and not during the actual division in Psammechinus eggs (Fig. 23) and in the ciliate Tetrahymena gelei. This agrees with observations of Forster and Orstrom and Bataillon (refer- ences in Brachet, *50a) that respiration is necessary only in the first part of mitosis, before metaphase. In the presence of col- chicine, chromosome movements and cyto- kinesis are suppressed, yet the rhythmic CELL DIVISION change in respiration associated with rhyth- mic disappearance and regeneration of the nuclear membrane persists (Zeuthen, °51). The increase in respiration is therefore not associated with chromosome movement or cytokinesis but with some process occurring before visible prophase. If eggs have the advantage of being easily available in large numbers in the same stage, they have the considerable drawback of being full of storage material and equipped to go on largely independently of the en- vironment. Because of this they are not very useful for the study of the metabolic requirements of mitosis. A few years ago Medawar and Bullough and Johnson (cf. Bullough, 52) found an excellent material for physiological studies on dividing cells in adult mammalian epi- dermis, a tissue that can be cultured quite easily. In a series of beautiful studies Bul- lough has analyzed the metabolic conditions for mitosis. He showed that in the epidermis mitosis is directly proportional to the con- centration of glucose and the oxygen ten- sion. Glucose is metabolized through the citric acid cycle. The inhibition of mitosis by 2,4-dinitrophenol suggests that oxidative phosphorylation is involved in the reactions providing the energy for division. Any agent that interferes with the citric acid cycle and oxidative phosphorylation prevents mitosis. These metabolic inhibitors affect the cell only during the part of interphase preced- ing mitosis, a stage which Bullough named antephase. Once the cell has passed this stage it can go on in the absence of respira- tion. In other tissues mitosis occurs also in the absence of respiration. Embryonic cells, for instance, either in vivo or in vitro, can divide without oxygen and are not inhibited by respiratory poisons (Parker, Pomerat and Willmer). These cells therefore may depend on glycolysis alone for mitosis (Laser; O’Con- nor). They are inhibited by fluoride and iodoacetate (Hughes). (References in Bul- lough, *52.) The observations on cultured epidermis are corroborated by studies on intact mice (Bullough, 52). The diurnal rhythm of mi- tosis in several tissues of mice was shown to be related to the deposition of glycogen in the cells during rest. When the animals are active the amount of carbohydrates avail- able is decreased and the number of mitoses goes down. Injection of carbohydrates and phosphate raises the number of mitoses to a maximum. Inhibition of phosphorylation by Pat phloridzin, on the other hand, blocks division completely. Though the biochemical analysis of mi- tosis is still in its beginning, some generaliza- tions are already possible. In order to divide, the cell requires energy. This energy is de- rived from breakdown of carbohydrates, through glycolysis in some cells and respira- tion in others. Oxidation of other foodstuffs may also occur, especially in eggs. The energy is then trapped in high-energy phos- phate bonds. All this takes place in the antephase, before any visible changes occur "yO (o—e) GO, Jin. SC een = aoe 4 (x—*) log “whole” nucleifembr (e—s) same, eggs from diver ' SS hr 13 14 15 Fig. 23. Oxygen uptake and mitosis in consecutive cleavages of Psammechinus. “Whole” nuclei indi- cate interphase and prophase, “open” nuclei meta- phase and anaphase. The increase in oxygen uptake occurs before each division when the nuclei are “Whole.” (After Zeuthen, ’51.) in the cell. During this phase the “batteries” of the cell are charged up and from then on mitosis proceeds without further energy up- take from the environment and only direct interference with mitotic organelles, or death of the cell, can block division. Further work will have to show how these metabolic processes of antephase are related to chro- mosome reproduction, how the energy is stored and made available for the various reactions during actual division. MITOGENESIS What are the conditions, both internal and external, that cause a cell to divide? And what conditions are responsible for the cessa- tion of cell reproduction? A free living cell such as an amoeba grows until it reaches a certain size and then it divides. A similar relationship between size and mitosis is found in other protozoans, the rate of divi- sion under optimal conditions being char- acteristic of a species. If the size of an 112 amoeba is kept small by cutting off a piece every day the cell may survive for months without ever dividing (Hartmann, ’26). We can consider this situation the simplest and most primitive relationship between growth (cytoplasmic synthesis) and cell division. R. Hertwig (03) based his theory of mitosis on it. He believed that the change in nu- clear-cytoplasmic ratio due to cell growth was the primary stimulus to cell division. It soon became evident, however, that cyto- plasmic growth and division are not neces- sarily connected; the stimulus for division therefore must be sought outside of a simple quantitative relationship between nucleus and cytoplasm. In certain algae division can be suppressed while growth continues, and giant cells are formed. Release of the inhibi- tion is followed by a series of rapid divisions until the usual size is restored again. Under different circumstances several divisions fol- low each other without interphasic growth and dwarf cells are produced (Hartmann, 33). By exchanging nuclei between amoebae that had just divided and cells entering mi- tosis, Commandon and de Fonbrune (’42) analyzed the function of cytoplasmic and nuclear growth in mitosis. They found that the nuclear-cytoplasmic volume ratio is not important, but that both the nucleus and cytoplasm have to be in a special state of maturity for mitosis to occur. Eggs have often been used to study mito- genesis, either through fertilization or ar- tificial parthenogenesis. Some physiologists expected to find a common factor in mitotic stimulation and the stimulation of nerve and muscle. According to Lillie increased permeability of the cell membrane was such a factor. Heilbrunn substituted the increased cytoplasmic viscosity as a common factor and suggested that the mitotic gelation was the primary stimulus to division (see Heilbrunn, ’62b). This hypothesis, however, is hardly more satisfactory as a theory of mitosis than the previous ones. It deals with certain phenomena that accompany the division of the cell but not with the fundamental proc- ess leading to the complex chain reaction from chromosome synthesis to formation of mitotic organelles, chromosome movements and division of the cell. In recent years more emphasis has been placed on biochemical changes in the cell during mitogenesis. The cytologist has failed so far in the understanding of mitogenesis because the essential changes in the cell take place before any microscopically visible man- ifestations appear. A better understanding of CELLULAR STRUCTURE AND ACTIVITY mitogenesis has to be based on a study of the shifts in enzyme systems, metabolic path- ways and synthetic mechanisms during ante- phase following the mitotic stimulus. A cell about to enter mitosis needs energy that is obtained mainly from a breakdown of carbohydrates through glycolysis or respira- tion. Any factor that increases the amount of substrate available to the cell therefore in- creases the number of cells in division. Daily rhythms in the number of dividing cells are explained by the variable amounts of car- bohydrate available to the cells (Blumenthal, 50; Bullough, ’52). Changes in vasculariza- tion are responsible for waves of mitoses and certain hormones (estrogen, testosterone) stimulate mitosis, apparently by increasing the carbohydrate supply to the cells affected (cf. Bullough, ’52). It has long been known that starvation suppresses cell division. With renewed feeding there is commonly a great increase in the number of divisions, followed by a new minimum. Apparently, in dividing tissues, energy sources affect the length of antephase, but not the entrance into a “ready state.” Starvation thus leads to a piling up of “ready” cells and when feeding starts again all these cells go through mitosis at once until all the “ready” cells are ex- hausted (cf. Kornfeld, ’22). It was shown above that DNA is synthe- sized during antephase. Continuation of mi- tosis depends, therefore, also on the avail- ability of precursors, enzymes and coenzymes involved in nucleic acid synthesis. Addition of such factors stimulates cell division (cf. Norris, *49). The relationship of thiols to cell division seems to be well established through the work of Hammett, Voegtlin, Chalkley and espe- cially Rapkine. The pertinent literature has been reviewed by Brachet (’50a). Sulfhydryl compounds stimulate cell division in plants and in animals, in tissue culture and in vivo. Compounds combining with —SH, on the other hand, inhibit mitosis reversibly. Rapidly growing tissues are especially rich in —SH. An attractive hypothesis of the role of —SH has been suggested by Rapkine and expanded by Brachet (50a). The changes in —SH in the cell are related to a reversible denaturation of proteins. Denaturation of proteins and associated conversion of globu- lar into fibrous proteins is thought to occur during the formation of asters, spindles and other mitotic gels. The —SH groups freed upon denaturation would then reduce oxi- dized glutathione in the cell. This could account for the observed increase in free CELL DIVISION —SH in dividing cells. S—S linkages, on the other hand, are thought to play a role in the formation of the protein gels of mitotic organelles. Indeed, Brachet has found that asters, spindle and the nuclear sap from which it is formed are especially rich in —SH after denaturation. Since —SH groups are also essential to the functioning of many enzymes (cf. Barron, 51), the effect of thiols on mitosis may be quite diverse. As mentioned above, the original state of affairs may be illustrated by the amoeba that grows to a certain size and then divides. Cells of metazoa no longer behave that way; they are dependent on checks and balances originating in the tissue, organ or organ- ism as a whole. Single cells in tissue culture are said to be unable to divide except under special conditions (Likely et al., 52); usu- ally there is a minimal number of cells below which growth does not occur (Fischer, 46). In the developing embryo, during re- generation, in buds of asexually reproducing animals this control of cellular proliferation has been demonstrated many times. Cell division in an organ stops when the proper relative size is reached and only where con- tinuous replacement is necessary, as in the intestinal mucosa or in the skin, do we find mitoses. If the balance is upset by removal of tissues or by other injuries to the cells, proliferation sets in again until the normal functional balance is restored. What factors are responsible for the con- trol of proliferation and what causes renewed mitotic activity? With the introduction of tissue culture techniques it became possible to study the factors responsible for initiation of mitosis directly and under controlled conditions. The mitosis-stimulating effect of embryo extracts and adult tissue extracts was discovered and the search was on for the nature and mode of action of “growth factors.” Much interest- ing information has been collected in the years since (see Fischer, 46), but it still remains unknown how these extracts act on the cells of the explant and how they stimulate mitosis. The facts indicate that substances released by injured cells and tis- sue extracts contain a whole spectrum of factors, substrates, coenzymes, building blocks and possibly self-reproducing enzyme systems (cytoplasmic particulates; Shaver, 53). It remains for future work to separate these factors and study their specific mode of action on the cell. The mitosis-stimulating effect of injury substances and tissue extracts has been stud- m3 ied not only in explants but also in the in- tact animal. Tissue extracts accelerate the healing of wounds (Auerbach and Doljanski, 45). Tumor extracts increase the growth of tumors (Annau, Manginelli and Roth, ’51) and other tissues. Even more interesting is the organ-specific action of tissue extracts. Injury of a certain tissue was shown to re- lease factors that stimulate mitosis in homol- ogous tissues, but not in others. These fac- tors are carried in the blood stream (Teir, 62). Similar results were obtained by Weiss (52) in embryos. Homologous tissue ex- tracts delay differentiation and stimulate cell proliferation. These experiments so far sup- port Weiss’ hypothesis of self-regulation of organ growth by diffusible products. Accord- ing to Weiss a cell produces “templates” in- volved in the production of new specific cytoplasm and diffusible units that move into the humoral pool. If these reach a certain concentration further proliferation of the particular cells is inhibited, since the dif- fusible units combine with and inactivate the “templates.” By injecting tissue extracts the concentration of the homologous humoral units is reduced and this leads to renewed proliferation in the particular tissue. Superimposed on the self-regulation of organs is the activity of hormones. Bullough (52) has recently reviewed the effect of hor- mones on mitotic activity in vertebrates. Both androgens and estrogens stimulate mitosis, except in nerve cells and striated muscle. Adrenal hormones, on the other hand, de- press mitotic activity. This hormonal con- trol of mitosis apparently operates by in- fluencing the availability of carbohydrate to the cell. Another kind of tissue interaction was discovered by Carrel (22). Fibroblasts can grow in serum alone in the presence of leukocytes. He called the substance liberated by leukocytes ‘‘trephones.” The role of such trephocytes has been investigated recently, es- pecially by Liebman (’47). The nature of trephones is unknown; they may be nutrients, vitamins or enzymes. Bacteria in tissue cul- ture or even in vivo may act as trephocytes (Lasfargues and Delaunay, *49). Wooley (53) has shown that tumors may act simi- larly. Mouse embryos deficient in B12 could erow if the mother carried certain tumors. These tumors synthesized Bie which then be- came available to the embryonic tissues, en- abling them to grow. In this connection it is of interest that certain tumors can grow in serum alone, while normal tissues need spe- cial growth factors in addition. It appears 114 that some cells retain or regain the ability to synthesize specific compounds that are essen- tial for mitosis, while other tissues remain unable to do so. Sometimes differentiation is connected with a permanent loss of mitotic activity. In vertebrates this applies to nerve cells and striated muscle. In some invertebrates the development of the entire organism or of certain organs is a closed system involving a specific and constant number of cell divi- sions. When morphogenesis is complete all cells have lost the ability to divide. An important aspect of mitogenesis is the “latent period.” Whether in tissue culture or in vivo, no matter what the stimulus, there is invariably a period of many hours be- tween application of the stimulus and ap- pearance of the first mitosis. In tissue culture it lasts from 20 to 24 hours (Fischer, ’46). It depends to some extent on the type of medium (Jacoby, 49). In explants of liver it is not the same in different types of cells (Abercrombie and Harkness, *51). It is gen- erally longer than the interphase under op- timal conditions. Growth factors act during this time and once mitosis is under way they are no longer required until the next ante- phase (Jacoby, ’37). No doubt mitotic stimu- lation produces some essential changes in the cytoplasm, as is borne out by the many observations of synchronous division of nu- clei in the same cell or in cells connected by cytoplasmic bridges. Whatever the nature of this change it does not go beyond the cell membrane. In other cells synchronous divisions are most likely the result of an inherent rhythm of mitosis with a fixed length of the various phases characterizing the particular type of cell (spermatocytes of insects, early cleav- age). The rate at which the number of cells increases in a tissue depends on the time for mitosis, the length of the interphase and the rate of removal of cells from the proliferating population by death or differentiation. The sum of mitotic time and interphase has been called generation time. Mitotic time and length of interphase for some cells can be determined directly by observation (eggs, tissue culture). In the intact animal the time for mitosis has been determined for several tissues by making use of the fact that moderate doses of x-rays inhibit the ante- phase so that no new cells enter mitosis, but those in division continue normally (Knowl- ton and Widner, 50; Widner et al., ’51). From the mitotic index (number of dividing CELLULAR STRUCTURE AND ACTIVITY cells/number of interphases) and the mitotic time the average length of interphase can be calculated (Table 4). The shortest generation time is found in early cleavage. Probably the shortest on record was reported for early cleavage of Drosophila (Huettner, ’33). A short inter- phase is characteristic also for certain em- bryonic cells such as the neuroblasts in the grasshopper (Carlson and Hollaender, ’48). In vertebrates the average mitotic time is remarkably similar from one tissue to an- other. Even in embryonic cells in tissue cul- ture and in various tumors the time for mitosis is about the same. The average length of interphase, however, is very variable. As shown in tissue culture it differs even in daughter cells (Jacoby, ’49; Fell and Hughes, 49). In other cells, for instance in spermato- cytes (Ris, ’49) or in cleaving eggs, mitotic time and interphase may be remarkably con- stant under the same conditions. The generation time is determined by in- trinsic factors and by external conditions. In the eggs of sea urchins (Moore, °33), am- phibians (Porter, *42) and fishes (Moenk- haus, 04), the rate of cleavage is specific for a species and is determined by the cyto- plasm. In later cleavage blastomeres may have different rates, specific for each cell and independent of the size of the cells (Chen and Pai, *49). Both mitotic time and interphase are in- fluenced by environmental factors, for in- stance, temperature (Barber, 739), pH, con- centration of embryo extract in tissue cul- ture (Jacoby, °37), tonicity of the medium (Cornman, *43), and the presence of certain ions (Moellendorff, 38) and hormones (Bul- lough, 52). The mitotic index has often been used as a measure for the proliferative ac- tivity of a tissue. To draw any conclusion, however, more than just the index must be known. The relationship of mitotic time and interphase to the mitotic index under vari- ous conditions was discussed by Hoffmann (49). We may conclude that many factors in- fluence in one way or another the number of cells in division or change the mitotic time or length of interphase and in a few cases we have some information on the mech- anisms of these effects. Other factors are known that are truly mitogenetic, that induce cells to enter mitosis that would not normally divide, even though energy sources and building blocks may be available. So far little is known of how this change in cells is brought about. CELL DIVISION P15 TasLe 4. Some Examples of Mitotic Time and Length of Interphase in Animal Cells CELL TEMPERATURE, MITOTIC TIME, INTERPHASE, AUTHOR “EL MINUTES HOURS DIRECT DETERMINATIONS Chick heart fibroblasts 40 25 10 or more Willmer, ’33 tissue culture (longer in old cultures) Chick macrophages 40 28-39 Jacoby, *49 tissue culture Chick embryo 40 34-52 Hughes, ’49 tissue culture Chick embryo 40 20 Olivo and tissue culture Slavich, ’30 Chick myocardium 40 1A Olivo and tissue culture Delorenzi, 233, Mouse spleen 40 43-90 8-18 Fell and tissue culture Hughes, *49 Frog fibroblast 26 90 Hughes and tissue culture Preston, ’49 Newt fibroblast 26 120 Hughes and Preston, ’49 Chortophaga 38 181 27 min.! Carlson and neuroblast Hollaender, 48 Rat cornea 70 200 Friedenwald, 50 Rabbit cleavage 9-10 8=9 Pincus, °39 cee at > Ae Drosophila cleavage 23 10 min. Huettner, ’33 Echinus miliaris cleavage WW 33-36 min. Gray, ’27b Psammechinus micro- 13 33-39 min. Callan, ’49 tuberculatus cleavage Sphaerechinus granularis 13 52-59 min. Callan, °49 cleavage INDIRECT DETERMINATIONS Mouse, female 150 Bullough and epidermis, castrate van Oordt, 50 with estrogen 45 Rat tissues jejunum PAY 335) 55 Widner et al., myelotic series 51 marrow Dail 32.4 nucleated red cells 24.6 61.0 Walker rat carcinoma 24.8 11.4 Jensen rat sarcoma 26.6 W258) Mouse tissues myelotic series 353 155 Knowlton and Widner, ’50 jejunum 2S)9) 43 erythrocytic series 295 99 lymph node PE) 5 O2 100 epidermis 30.2 670 adrenal 14.4 1090 * Values for mitotic time by indirect determinations are averages. 116 INHIBITION OF MITOSIS Many physical and chemical agents in- hibit cell division without killing the cell. They have been called “mitotic poisons.” Besides being useful tools for dissecting mi- tosis into its component processes and for the biochemical characterization of these components, they have great practical in- terest as potential inhibitors of pathological growth. Some inhibitors were discovered empirically to block various phases of mitosis (x-rays, colchicine, nitrogen mustards, etc.). The specific ways in which they affect the cell are now studied in many laboratories. Other inhibitors are well known tools of the biochemist, specific inhibitors of certain en- zymes or antimetabolites that block meta- bolic processes at definite points. They have revealed some of the enzymatic reactions that play a role in cellular reproduction. Reviews of the more recent work on mitotic inhibitors have been published by Lettré (51, °52) and Lehmann (751). Two phases in the course of cell division are especially sensitive to external influences. One is the antephase, the time when energy is produced and stored for mitosis and when chromosome reproduction takes place. The second is characterized by the formation of oriented gels, of which the swelling and contraction play a role in chromosome move- ment and cytokinesis. Antephase inhibitors prevent cells from entering prophase, but do not affect those already in division. They include factors that interfere with glycolysis or respiration and with the formation of high-energy phos- phates (Bullough, 52; Krahl, 50), and those that inhibit chromosome reproduction. Of special interest in this second group are the inhibitors of DNA synthesis. DNA is gen- erally restricted to chromosomes and its syn- thesis to chromosome reproduction. Synthesis of DNA may be accomplished by different pathways in different cells, tissues or organ- isms, and in normal cells and tumor cells; hence the hope for cell specific inhibitors. The better known inhibitors of chromosome reproduction are x-rays (Hevesy, ’48: Skip- per, 51), nitrogen mustard (Bodenstein and Kondritzer, ’48; Friedenwald and Siegelman, 53: Goldthwait, 52), folic acid antaconists such as aminopterin and amethopterin (cf. Petering, 52) and certain purines and pyrim- idines (Lettré, °51). Cells arrested in antephase are extremely labile and easily undergo deeeneration. In rapidly proliferating tissues the majority of cells can be accumulated in this phase and CELLULAR STRUCTURE AND ACTIVITY then destroyed (Gillette and Bodenstein, 46; Friedenwald, *51). Chemical agents that in- terfere with chromosome reproduction have cytological effects that are similar to those of ionizing radiations, and therefore have been called radiomimetic (cf. Loveless and Revell, *49). In addition to the antephase block they produce chromosome clumping and chromosome breaks. Another group of “chromosome poisons” that inhibit antephase and also later stages are acridine derivatives (trypaflavin, pro- flavin), investigated especially by Bauch (47) and Lettré (51). They seem to act by forming complexes with nucleic acids, interfering mainly with polymerization of nucleic acid. The movements of chromosomes and cell cleavage are dependent on nuclear and cy- toplasmic gels (asters, spindle, cortical gel of cleavage furrow). Beginning with meta- kinesis, mitosis can be blocked by agents that interfere with gelation of these organelles, or with contraction of chromosomal fibers and cortical gel. The relative sensitivity of these structures is often somewhat different so that cleavage, for instance, can be sup- pressed without halting chromosome move- ments, or spindle stretching may be _ in- hibited without affecting the contraction of chromosomal fibers (Ris, 749). Complete in- hibition of these organelles is produced by certain anticoagulants, for instance, heparin (Heilbrunn, *52a), by hydrostatic pressure (Marsland, 51), by high or low temperature, and by hypotonic media (Lewis, ’34). The best known specific poison of the spindle is colchicine. In the presence of this alkaloid the spindle does not form, or, if it is present, is destroyed together with asters and the cleavace furrow. Chromosome reproduc- tion, spiralization and breakdown of nuclear membrane and the initial parallel senara- tion of chromatids are not affected and nu- clei may undergo several cycles of repro- duction in colchicine (Zeuthen, *51). Often the chromosomes clump, or are scattered, giving rise to micronuclei with variable numbers of chromosomes. Tnoué (752) has used an improved polariz- ins microscope to study the effect of col- chicine on the structure of the spindle. He found that colchicine primarily destroys the orientation of the spindle micelles. The spin- dle material may become scattered or remain in the cell as a spherical mass (Gaulden and Carlson, *51), devending on the colchicine concentration and the type of cell. Lettré (51, °52) has studied a great num- CELL DIVISION ber of colchicine derivatives and other alka- loids in search for the molecular structure responsible for the colchicine effect. He found that the stilbylamine group is es- sential, though the type of substitution too is important. It is not yet clear how col- chicine acts on the spindle micelles. Lettré suggested that it interferes with the utiliza- tion of ATP for spindle contraction, since ATP counteracts colchicine. However, there is more to it than inhibition of contraction, since the structure of the spindle and asters is actually destroyed. Ostergren (44) made a comparative study of many chemically unrelated substances that have a colchicine-like action. He called attention to the fact that the more lipoid- soluble the substance the lower the threshold concentration for colchicine-like effect (C- mitosis). He proposed a protein chain folding theory according to which the active mole- cule attracts the lipoid side chains of the proteins, causing a folding up which would explain the breakdown of the cytoplasmic gels and spindle and perhaps also the more than normal shortening of the chromosomes typical of colchicine-mitosis. Colchicine destroys the spindle almost universally in both animals and plants. It is very interesting that there are cells and organisms that are resistant to it. In the hamster colchicine is without effect (Orsini and Pansky, ’52) and Lettré (52) has found a strain of ascites tumor in the mouse that is resistant. Another group of spindle poisons have in common that they combine specifically with —SH groups. Quinones were discovered by Lehmann (751) to affect specifically spindle and cytokinesis. Organometallic compounds were studied by Klages and Lettré (for ref- erences see Lettré, 51). Their action is re- versed by cysteine and other —SH com- pounds. Chloroacetophenone, another —SH poison, blocks metaphase in vitro (Hughes, 50) and in vivo (Beatty, *51). Sulfhydryl poisons can block mitosis in two ways, either by inhibition of —SH enzymes (Barron, ’51) or by combining with —SH groups of the proteins in spindle and asters and thus in- terfering with the formation and function of these gel structures. It is interesting that some —-SH poisons are spindle poisons in low concentrations and inhibitors of ante- phase in higher concentration, possibly by interfering with carbohydrate metabolism (Meier and Allgéwer, ’45; Hughes, 50). Folic acid, as mentioned above, is essential in antephase and aminopterin prevents cells EVA from entering prophase. Hughes (750) and Jacobson (51) found that aminopterin and amethopterin block mitosis also at metaphase. Folic acid is thus necessary also for ana- phase movements of chromosomes, but noth- ing is known about how it functions here. MODIFICATION OF MITOSIS The division of cells involves a complex series of events that normally follow each other in a definite order. It can be inter- rupted experimentally by a variety of in- hibitors that affect one or several links in the chain. Often it is modified in connection with specializations in the growth and func- tion of tissues or in the life cycle of the organism. Thus the mitotic chain may be broken at one point or the other, or the nor- mal sequence of events is changed, or one of the components is altered in some way. Completion of cytoplasmic division is most easily affected and so we find many exam- ples of division without cytokinesis. In mam- malian tissues, for instance, binucleate cells are common, and they are the result of such failure of cytoplasmic division (Beams and King, ’42). In spermatogenesis of coccids cell division is omitted regularly in the second meiotic division (Hughes-Schrader, °48). If mitosis is interrupted at an earlier point chro- mosome movements and therefore nuclear division are absent. Reproduction of nuclear elements that is not followed by chromo- some movements and cytoplasmic division has been called “endomitosis” (cf. Geitler, 48). Endomitosis may occur with a typical prophase, including increase in nuclear vol- ume and chromosome spiralization. The nu- clear membrane, however, persists and the cell becomes polyploid (somatic tissues of Hemiptera and many other insects). Poly- ploid cells may later again divide by reg- ular mitosis. In. other cases prophase is suppressed too, only antephase is left, and there is no visible change in the structure of the interphase nucleus. Where chromo- some reproduction is not followed by chro- mosome splitting and separation of chro- matids, endomitosis does not change the num- ber of chromosomes but increases the number of units in each chromosome, giving rise to polytene chromosomes. They are best known from larval tissues of Diptera, where through many consecutive endomitoses the number of chromonemata per chromosome may increase a hundredfold. The volumes of polytene and polyploid nuclei are usually multiples of the original 118 size, and the amount of DNA per nucleus in- creases by a factor of two with each endomi- tosis. The doubling in nuclear size is fol- lowed by an increase in cytoplasm. Endomitosis is a major factor in the erowth of many differentiated tissues. Ap- parently it interferes less with the function- ing of differentiated cells than does a com- plete mitosis. Although the number of cells remains the same, it accomplishes an in- crease in the functional elements of the cell from chromosomes and nucleolus to the vari- ous cytoplasmic elements, and therefore aug- ments the functional capacity of the tissue as a whole. The relationship of endomitosis and polysomaty (polyploidy of tissue cells in a diploid organism) to differentiation has been discussed in detail by Geitler (41) and by Huskins (47). In mammalian tissues polyploid cells orig- inate not only through endomitosis but also through fusion of spindles during mitosis of binucleate cells (Beams and King, °42; Fell and Hughes, *49). Excellent discussions of polysomaty in mammals in connection with the problem of nuclear size classes have been published by Teir (44) and Helweg-Larsen (52). Mitosis may be modified also by a change in the sequence of mitotic processes. An interesting example was described by Berger (738) and by Grell (’46) in the mid- gut of the mosquito. In these cells several cycles of chromosome reproduction occur in the larva and polytene chromosomes are formed. During metamorphosis these cells then divide repeatedly without chromosome reproduction until the chromatids produced by endomitosis in the larva are divided up into the newly formed cells and the diploid condition is restored. Chromosome reproduc- tion thus takes place in the larva, chromo- some separation and cell division in the pupa. More commonly karyokinesis is separated in time from the division of the cytoplasm. Repeated nuclear divisions give rise to a multinucleated cell which is subdivided later simultaneously into the appropriate number of uninucleated cells (insect cleavage; other examples in Miihldorf, 51). The cytological literature is rich in de- scriptions of interesting modifications in the behavior of mitotic organelles. The elimina- tion of certain chromosomes during cleavage in some dipterans, for instance, is accom- plished through failure of the chromosomal fibers of the eliminated chromosomes to con- tract during anaphase. These chromosomes are therefore left behind and are not in- CELLULAR STRUCTURE AND ACTIVITY cluded in the daughter nuclei. Thus somatic cells with a reduced chromosome number are formed (DuBois, °33; Reitberger, 40; White, 46). Then there is the puzzling behavior of the chromosomes in the primary spermato- cytes of Sciara (Metz, ’33). The chromo- somes having failed to synapse in prophase, all form chromosomal fibers toward the sin- gle aster (monocentric mitosis). No meta- phase plate is formed. At anaphase the maternal chromosomes move toward the active center while the paternal chromo- somes move away from it, and backward too, since they are still attached to the center by chromosomal fibers that appear to be under tension and attenuate the chro- mosomes. How can we account for this unorthodox chromosome movement? I think it is unnecessary to introduce any special mechanisms, since it can be understood on the basis of slight modifications in the behavior of the known mitotic organelles. First, the chromosomal fibers contract only in the maternal chromosomes, pulling them to the poles. Secondly, the spindle elongates and carries the paternal chromosomes along passively away from the single aster. The presence of a spindle is indicated by the distribution of the mitochondria and spindle stretching is suggested by the elongation of the cell precisely in the direction in which the paternal chromosomes move. The observation that low temperature, which is known to destroy the spindle, inhibits the backward movements of the chromosomes and also the elongation of the cell supports our interpretation. The most genera] and most important modifications of mitosis are found in the meiotic divisions during gametogenesis. In the first division the most significant modifi- cation is the pairing of homologous chromo- somes during prophase. So far there is no satisfactory explanation of this phenomenon. Theories involving changes in timing (the precocity theory of Darlington, for instance) have no factual basis since Swift and Klein- feld (53) have shown that chromosome re- production (DNA doubling) takes place be- fore prophase as in somatic mitoses. Since the association of homologous chromosomes continues into metaphase, either because of chiasmata or owing to a localized or general persistence of the “pairing force,” homolo- gous kinetochores are co-oriented in meta- kinesis instead of kinetochores of sister chro- matids. As a result this division segregates homologous kinetochores and, depending on CELL DIVISION the degree of crossing over, certain segments of the homologous chromosomes. The second division proceeds without chromosome re- production. The kinetochores orienting on the spindle are therefore the sister kineto- chores of the first division. The fundamental features of meiosis are thus pairing in the first division followed by a separation of the four chromatids in each chromosome pair by two karyokineses without any further chromosome reproduction. The four nuclei resulting from these divisions are therefore haploid. A somewhat different way in distributing the four chromatids of each bivalent during the two meiotic divisions has been described in some insects with diffuse kinetochore. It was discussed by Ris (°42), Oksala (’43) and Hughes-Schrader (’48). AMITOSIS According to Remak the nucleus divides by simple constriction. After the discovery of mitosis many cytologists continued to believe that such direct division, or “amitosis” as Flemming called it, was an alternate method for nuclear and cell reproduction. With the establishment of the chromosome theory of inheritance it became very improbable that cells could proliferate by amitosis. A critical study of most cases of amitosis disclosed that they were either modified or abnormal mi- toses (pseudo-amitoses, cf. Politzer, *34) or based on faulty observations. There remained, however, a number of observations of con- striction and fragmentation of nuclei into two or more parts, especially in certain highly specialized tissues that no longer divide mitotically. Only very rarely is nu- clear fragmentation followed by division of the cytoplasm (Schrader, *45), and then there is no evidence that these cells persist or are able to multiply. Nuclear fragmenta- tion is especially common in cells that have become polyploid through endomitosis (Heid- enhain, °19) or inhibited mitosis (Pfuhl, ’39; Bucher, °47). It is, however, never followed by cell division. Direct division of the nucleus, or ‘“‘ami- tosis,” is not a method of cellular prolifera- tion but of nuclear fragmentation, generally in polyploid cells that are no longer able to divide. As suggested already by Flemming, it appears to be related to increased physio- logical activity of a cell. Especially in large polyploid cells it would result in a_ better distribution of nuclear material throughout the cytoplasm. 119 The term “amitosis” is misleading, since it suggests an alternative to mitosis. Reproduc- tion of cells can take place only through mi- tosis. Reproduction of nuclear material within a cell may occur through mitosis or endo- mitosis. The endomitotically enlarged nu- cleus may remain single or break up into two or more fragments. I would suggest that this process be called “nuclear fragmenta- tion” and that the term “amitosis” be elim- inated. REFERENCES Abercrombie, M., and Harkness, R. D. 1951 The growth of cell population and the properties in tissue culture of regenerating liver of the rat. Proc. Roy. Soc. London. B138:544-561. Annau, E., Manginelli, A., and Roth, A. 1951 In- creased weight and mitotic activity in the liver of tumor bearing rats and mice. Cancer Res., 17: 304-306. Auerbach, E., and Doljanski, L. 1945 Effect of cell growth activating tissue extracts, parenter- ally applied, on experimental skin wounds. Proc. Soc. Exp. Biol. & Med., 58:111-114. Bairati, A., and Lehmann, F. E. 1952 Ueber die submikroskopische Struktur der Kernmembran bei Amoeba proteus. Experientia, 8:60-61. Baker, J. 1951 Remarks on the discovery of cell- division. Isis, 42:285-287. Barber, H. N. 1939 The rate of movement of chromosomes on the spindle. Chromosoma, 7:33— 50. Barron, E.S.G. 1951 Thiol groups of biological importance. Advances Enzymol., 77:201-266. Bataillon, E., and Tchou Su 1930 Etudes analy- tiques et experimentales sur les rythmes ciné- tiques dans l’oeuf. Arch. de Biol., 40:441-540. Battaglia, B., and Omodeo, P. 1949 Ricerche is- tochimiche sugli acidi nucleinici nella spermato- genesi dei Lumbricidi. Caryologia, 2:1—12. Bauch, R. 1947 Trypaflavin als Typus der Chro- mosomengifte. Naturwiss., 34:346—-347. Beams, H. W., Evans, T. C., Baker, W. W., and van Breemen, V. 1950a Electron micrographs of the amphiaster in the whitefish blastula (Core- gonus cluperformis). Anat. Rec., 107:329-346. , Evans, T. C., van Breemen, V., and Baker, W. W. 1950b Electron microscope studies on structure of mitotic figure. Proc. Soc. Exp. Biol. & Med., 74:717-720. , and King, R. L. 1936 The effect of ul- tracentrifuging upon chick embryonic cells, with special reference to the “resting” nucleus and the mitotic spindle. Biol. Bull., 77:188-198. ,and King, R. L. 1942 The origin of binu- cleate and large mononucleate cells in the liver of the rat. Anat. Rec., §3:281-297. Beatty, R. A. 1951 Effects of chloracetophenone and di-isopropylfluorophosphonate on amphibian eggs. Proc. Roy. Soc. London, B138:575-599. Bélar, K. 1929 Beitrage zur Kausalanalyse der Mitose. II. Roux’ Arch. Entw.-mech., 178:359- 484. 120 Bélar, K., and Huth, W. 1933 Zur Teilungsauto- nomie der Chromosomen. Z. Zellf., 77:51-66. Berger, C. A. 1938 Multiplication and reduction of somatic chromosome groups as a regular de- velopmental process in the mosquito, Culex pi- piens. Contrib. to Embryol. No. 167, Carnegie Institution of Washington, pp. 211-232. Bernal, J. D. 1940 Structural units in cellular physiology; in the Cell and Protoplasm, edited by F. R. Moulton. AAAS Pub. No. 14, Washington, D. C., pp. 199-205. Berrill, N. J., and Huskins, C.L. 1936 The “rest- ing” nucleus. Am. Nat., 70:257-260. Biesele, J. J. 1946 The size of somatic chromo- somes at different ages in the rat. J. Gerontol., 7: 433-440. 1947 Chromosomes in lymphatic leuke- mia of C58 mice. Cancer Res., 7:70-77. 1949 Phosphatases of the mitotic appara- tus in cultured normal and malignant mouse cells; in Proceedings of the First National Cancer Conference, pp. 34-41. American Cancer Society, Washington, D. C. Blumenthal, H. T. 1950 The nature of cycle variations in mitotic activity: the relation of ali- mentation and nutrition to this phenomenon. Growth, 74:231-249. Bodenstein, D., and Kondritzer, A. A. 1948 The effect of nitrogen mustard on nucleic acids during embryonic amphibian development. J. Exp. Zool., 107:109-121. Borgert, A.,1900 Untersuchungen iiber die Fort- pflanzung der tripyleen Radiolarien, speziell von Aulacantha scolymantha, H. Zool. Jahrb., Anato- mie, 74:203-276. Brachet, J. 1942 La localization des acides pento- senucléiques dans les tissus animaux et les oeufs d’amphibiens en voie de développement. Arch. de Biol., 53:207—-257. 1950a Chemical Embryology. Intersci- ence Publishers, New York. 1950b Une étude cytochimique des frag- ments nucléés et énucléés d’amibes. Experientia, 6:294-295. , and Shaver, J. R. 1950 The injection of embryonic microsomes into early amphibian em- bryos. Experientia, 5:204-205. Briggs, R., Green, E. U., and King, T.S. 1951 An investigation of the capacity for cleavage and differentiation in Rana pipiens eggs lacking “functional” chromosomes. J. Exp. Zool., 116: 455-500. Bucher, O. 1947 Division nucléaire amitotique dans des cultures de fibrocytes aprés administra- tion de colchicine. Acta Anat., 4:60-67. Bullough, W.S. 1952 Energy relations of mitotic activity. Biol. Rev., 27:133-168. , and van Oordt, G. J. 1950 The mitogenic actions of testosterone propionate and of oestrone on the epidermis of the adult male mouse. Acta Endocrin., 4:291—305. Callan, H. G. 1949 Cleavage rate, oxygen con- sumption and ribose nucleic acid content of sea- urchin eggs. Biochim. Biophys. Acta, 3:92-102. , and Tomlin, S. G. 1950 Experimental studies on amphibian oocyte nuclei. I. Investiga- CELLULAR STRUCTURE AND ACTIVITY tion of the structure of the nuclear membrane by means of the electron microscope. Proc. Roy. Soc. London, B137:367-378. Carlson, J. G. 1938 Mitotic behavior of induced chromosomal fragments lacking spindle attach- ments in the neuroblasts of the grasshopper. Proc. Nat. Acad. Sci., 24:500-507. 1946 Protoplasmic viscosity changes in different regions of the grasshopper neuroblast during mitosis. Biol. Bull., 90:109-121. 1952 Mlicrodissection studies of the divid- ing neuroblast of the grasshopper, Chortophaga viridifasciata (de Geer). Chromosoma, 5:199- 290. , and Hollaender, A. 1948 Mitotic effects of ultraviolet radiation of the 2250 A region with special reference to the spindle and cleavage. J. Cell. Comp. Physiol., 37:149-174. Carrel, A. 1922 Growth promoting function of leucocytes. J. Exp. Med., 36:395-392. Caspersson, T. 1939 Ueber die Rolle der Desoxy- ribosenukleinsaure bei der Zellteilung. Chromo- soma, 7:14:7-156. 1940 Ueber Eiweiss-stoffe im Chromo- somengertist. Naturwiss., 28:514-515. Chambers, R. 1917 Milicrodissection studies, II. J. Exp. Zool., 23:483-505. 1924 The physical structure of proto- plasm as determined by microdissection and in- jection; in General Cytology, edited by E. V. Cowdry. University of Chicago Press, Chicago, pp. 269-276. 1938 Structural and kinetic aspects of cell division. J. Cell. Comp. Physiol., 72:149-165. Chen, C. L., and Pai, S. 1949 Furchungsge- schwindigkeit und Furchungsrhythmus bei Bra- chionus pala und Polyarthra platyptera. Exp. Cell Res., Suppl., 7:540-541. Chickering, A. M. 1927 Spermatogenesis in the Belastomatidae, II. J. Morph., 44:541-607. Churney, L. 1936 The quantitative determina- tion of mitotic elongation. Biol. Bull., 70:400-407. Commandon, J., and de Fonbrune, P. 1942 In- fluence des stades évolutifs du cytoplasme et du noyau greffé d’Amoeba sphaeronucléus sur leurs volumes respectifs et sur le déclenchement de la caryocinése. Compt. rend. Soc. Biol., 136:763-764. Conklin, E. G. 1912 Cell size and nuclear size. J. Exp. Zool., 72:1-98. 1917 Effects of centrifugal force on the structure and development of the egg of Crepi- dula. J. Exp. Zool., 22:311-420. Cooper, K. W. 1939 The nuclear cytology of the grass mite, Pediculopsis graminum (Reut.) with special reference to karyomerokinesis. Chromo- soma, 7:51-103. 1941 Visibility of the primary spindle fibers and the course of mitosis in the living blasto- meres of the mite, Pediculopsis graminum, Reut. Proc. Nat. Acad. Sci., 27:480-483. Cornman, I. 1943 Acceleration of cleavage of Arbacia eggs by hypotonic sea water. Biol. Bull.. 84:244.951. 1944 A summary of evidence in favor of the traction fiber in mitosis. Am. Nat., 78:410- 429. CELL DIVISION Cornman, I., and Cornman, M. E. 1951 The ac- tion of podophyllin and its fractions on marine eggs. Ann. N.Y. Acad. Sci., 57:1443-1488. Dan, J. K. 1948 On the mechanism of astral cleavage. Physiol. Zool., 27:191-218. Darlington, D. C. 1942 Chromosome chemistry and gene action. Nature, 749:66-69. Davidson, J. N. 1950 Biochemistry of Nucleic Acids. Methuen & Co., London. Davies, H. G. 1952 The ultra-violet absorption of living chick fibroblasts during mitosis. Exp. Cell. Res., 3:453-461. Dawson, A. B. 1940 Cell division in relation to differentiation. Growth Suppl., 2:91-106. DeLamater, E. D., and Mudd, S. 1951 The oc- currence of mitosis in the vegetative phase of Bacillus megatherium. Exp. Cell. Res., 2:499- 512. Delbriick, M. 1941 A theory of autocatalytic synthesis of polypeptides and its application to the problem of chromosome reproduction. Cold Spring Harbor Symp. Quant. Biol., 9:122-126. DuBois, A. M. 1933 Chromosome behavior dur- ing cleavage in the egg of Sciara coprophila (Dip- tera) in relation to the problem of sex determina- tion. Z. Zellf., 19:595-614. Ephrussi, B. 1951 Remarks on cell heredity; in Genetics in the 20th Century, edited by L. C. Dunn. The Macmillan Co., New York, pp. 241- 262. Erdmann, R. 1908 Experimentelle Untersuchun- gen uber Massenverhaltnisse von Plasma, Kern und Chromosomen in dem sich entwickelnden Seeigelei. Arch. f. Zellf., 2:76-136. Fankhauser, G. 1934 Cytological studies on egg fragments of Triton. IV. J. Exp. Zool., 67:349- 395. Fell, H. B., and Hughes, A. F. 1949 Mitosis in the mouse: a study of living and fixed cells in tissue culture. Quart. J. Micr. Sci., 90:355-380. Ferry, J.D. 1948 Protein gels. Advances Protein Chem., 4:2-79. Fischer, A. 1946 Biology of Tissue Cells. Cam- bridge University Press, Cambridge, England. Friedenwald, J.S. 1950 Recent studies on corneal metabolism and growth: a review. Cancer Res., 10:461-466. 1951 The action of nitrogen mustards and related substances on cell division. Ann. N. Y. Acad. Sci., 57:1432-1442. , and Sigelman, S. 1953 The influence of ionizing radiation on mitotic activity in the rat’s corneal epithelium. Exp. Cell. Res., 4:1-31. Friedrich-Freksa, H. 1940 Bei der Chromoso- menkonjugation wirksame Krafte und ihre Be- deutung fiir die identische Verdoppelung von Nucleoproteinen. Naturwiss., 28:376-379. Gaulden, M. E., and Carlson, J. G. 1951 Cyto- logical effects of colchicine on the grasshopper neuroblast in vitro with special reference to the origin of the spindle. Exp. Cell. Res., 2:416—-433. Geitler, L. 1938 Chromosomenbau. Protoplasma- monographien No. 14. 1941 Das Wachstum des Zellkerns in tierischen und pflanzlichen Geweben. Erg. Biol., 18:1-54. 21 1948 Ergebnisse und Probleme der En- domitoseforschung. Oesterr. Bot. Zeitschr., 95: 277-299. Gillette, R., and Bodenstein, D. 1946 Specific de- velopmental inhibitions produced in amphibian embryos by a nitrogen mustard compound. J. Exp. Zool., 103:1-32. Goldthwait, D. A. 1952 Effect of nitrogen mus- tard on nucleic acid metabolism. Proc. Soc. Exp. Biol. & Med., 80:503-504. Gray, J. 1927a The mechanism of cell-division, IV. The effect of gravity on the eggs of Echinus. Brit. J. Exp. Biol., 5:102-111. 1927b The mechanism of cell division, III. The relationship between cell division and growth in segmenting eggs. Brit. J. Exp. Biol., 4:313-321. Grell, S. M. 1946 Cytological studies in Culex. I. Somatic reduction divisions. Genetics, 37:60-76. Gross, F. 1935 Die Reifungs-und Furchungstei- lungen von Artemia salina im Zusamenhang mit dem Problem des Kernteilungsmechanismus. Z. Zellf., 23:522-566. Hance, R. T. 1926 A comparison of mitosis in chick tissue cultures and in sectioned embryos. Biol. Bull., 50:155-159. Harris, P., and James, T. W. 1952 Electron microscope study of the nuclear membrane of Amoeba proteus in thin sections. Experientia, 8: 384-385. Hartmann, M. 1926 Ueber experimentelle Un- sterblichkeit von Protozoen Individuen. Natur- wiss., 14:433-435. 1933 Allgemeine Biologie. Springer, Ber- lin. Harvey, E. B. 1935 The mitotic figure and cleavage plane in the egg of Parechinus micro- tuberculatus as influenced by centrifugal force. Biol. Bull., 69:287-297. 1936 Parthenogenetic merogony or cleavage without nuclei in Arbacia punctata. Biol. Bull., 77:10-121. Hayashi, T. 1952 Contractile properties of com- pressed monolayers of actomyosin. J. Gen. Phys- iol., 36:139-152. Heidenhain, M. 1919 Uber die Noniusfelder der Muskelfaser. Anatom. Hefte, 56:321—402. Heilbrunn, L. V. 1952a An Outline of General Physiology, 3d ed. W. B. Saunders Co., Philadel- phia. 1952b The physiology of cell division; in Modern Trends in Physiology and Biochemistry, edited by E. S. G. Barron. Academic Press, New York. Helweg-Larsen, H. F. 1952 Nuclear class series. Contr. Univ. Institute f. Human Genetics, Copen- hagen, 27:1-139. Henke, K. 1947 Einfache Grundvorgange in der tierischen Entwicklung. Naturwiss., 34:149-157, 1947, , and Pohley, H. J. 1952 Differentielle Zellteilungen und Polyploidie der Schuppenbil- dung der Mehlmotte Ephestia kiihniella Z. Af. Naturforsch., 7b:65-79. Hertwig, R. 1903 Ueber Korrelation von Zell und Kerngrésse und ihre Bedeutung fiir die ge- 122 schlechtliche Differenzierung und die Teilung der Zelle. Biol. Centralbl., 23:49-62. Hevesy, G. 1948 Nucleic acid metabolism; in Advances Biol. Med. Physics, 7:409-454. Hoffman, J.G. 1949 Theory of the mitotic index and its application to tissue growth measurement. Bull. math. Biophys., 77:139-144. Howard, A., and Pelc, S. R. 1951a Nuclear in- corporation of P32 as demonstrated by autoradio- graphs. Exp. Cell. Res., 2:178-187. , and Pelc, S. R. 1951b Synthesis of nu- cleoprotein in bean root cells. Nature, 167:599- 600. Huettner, A. F. 1933 Continuity of the centrioles in Drosophila melanogaster. Z. Zellf., 19:119-134 Hughes, A. F. W. 1949 The effect of iodoaceta- mide upon cell division in tissue cultures on the chick embryo. J. Roy. Micr. Soc., 69:215-224. 1950 The effect of inhibitor substances on cell division. Quart. J. Micr. Sci., 97:251-278. 1952 The Mitotic Cycle. Academic Press, New York. , and Preston, M. M. E. 1949 Mitosis in living cell of amphibian tissue cultures. J. Roy. Micr. Soc., 69:121-131. , and Swann, M. M. 1948 Anaphase movements in the living cell. J. Exp. Biol., 25: 45-70. Hughes-Schrader, S. 1924 Reproduction in Acro- schismus. J. Morph. 39:157-207, 1924. 1931 A study of the chromosome cycle and the meiotic-division figure in Llaveia bouvari —a primitive coccid. Z. Zellf., 13:742-769. 1943a Polarization, kinetochore move- ments, and bivalent structure in the meiotic chro- mosomes of male mantids. Biol. Bull., 85:265—300. 1943b Meiosis without chiasmata—in diploid and tetrapoid spermatocytes of the mantid Callimantis antillarum Saussure. J. Morph., 73: 111-140. 1948 Cytology of coccids. Advances Gen- etics, 2:127-203. , and Ris, H. 1941 The diffuse spindle at- tachment of coccids, verified by the mitotic be- havior of induced chromosome fragments. J. Exp. Zool., 87:429-451. Huskins, C. L. 1947 The subdivision of the chro- mosomes and their multiplication in nondividing tissues: possible interpretations in terms of gene structure and gene action. Am. Nat., 87:401-434. Inoué, S. 1952 The effect of colchicine on the microscopic and submicroscopic structure of the mitotic spindle. Exp. Cell. Res. Suppl., 2:305-314. , and Dan, K. 1951 Birefringence of the dividing cell. J. Morph., 89:423-456. Jacobson, W. 1951 Biology of mitosis; in Trans- actions of the 13th Conference on Problems of Ageing, edited by N. W. Shock. Josiah Macy, Jr. Foundation, New York. , and Webb, M. 1952 The two types of nucleoproteins during mitosis. Exp. Cell. Res., 3: 163-183. Jacoby, F. 1949 A quantitative analysis of the growth of pure populations of fowl macrophages in vitro. Exp. Cell. Res. Suppl., 7:454455. CELLULAR STRUCTURE AND ACTIVITY , Trowell, O. A., and Willmer, E. N. 1937 Studies on the growth of tissues in vitro. J. Exp. Biol., 14:255-266. Jorgensen, M. 1913 Zellenstudien II. Die Ei- und Nahrzellen von Piscicola. Arch. Zellf., 10: 127-160. Johnson, H. H. 1931 Centrioles and other cyto- plasmic components of the male germ cells of the Gryllidae. Z. wiss. Zool., 740:115-166. Kaufmann, B. P. 1948 Chromosome structure in relation to the chromosome cycle. Bot. Revy., 74, 57-126. , McDonald, M., and Gay, H. 1948 En- zymatic degradation of ribonucleoproteins of chromosomes, nucleoli and cytoplasm. Nature. 162:814. Knowlton, N. P., and Widner, W. R. 1950 The use of x-rays to determine the mitotic and inter- mitotic time of various mouse tissues. Cancer Res., 10:59-63. Kornfeld, W. 1922 Ueber den Zellteilungsrhyth- mus und seine Regelung. Roux’ Arch. Entw.- mech., 50:526-592. Krahl, M. E. 1950 Metabolic activities and cleavage of eggs of the sea-urchin, Arbacia punc- tulata. Biol. Bull., 98:175-217. Kupka, E., and Seelich, F. 1948 Die anaphasische Chromosomenbewegung. Chromosoma, 3:302- 3272 Lasfargues, E., and Delaunay, A. 1949 Sur les trephones d’origine microbienne. Exp. Cell. Res. Suppl., 7:452-453. Lehmann, F. E. 1951 Der Kernapparat tieri- scher Zellen und seine Erforschung mit Hilfe von Antimitotica. Schweiz. Zeitschr. f. Allg. Path. Bakt., 74:487-508. Lettré, H. 1951 Zellstoffwechsel und Zelltei- lung. Naturwiss., 38:490-4.96. 1952 Some investigations on cell be- havior under various conditions: a review. Can- cer Res., 12:847-860. Levan, A. 1938 The effect of colchicine on root mitosis in Allium. Hereditas, 24:471-486. Lewis, M. R. 1923 Reversible gelation in living cells. Bull. Johns Hopkins Hosp., 34:373-379. 1934 Reversible solation of the mitotic spindle of living chick embryo cells studied in vitro. Arch. Exp. Zellf., 76:159-160. Lewis, W. H. 1920 Giant centrospheres in de- generating mesenchyme cells of tissue cultures. J. Exp. Med., 37:275-292. 1947 Interphase (resting) nuclei, chro- mosomal vesicles and amitosis. Anat. Rec., 97: 433-446. 1951 Cell division with special reference to cells in tissue culture. Ann. N. Y. Acad. Sci., 51:1287-1294. Liebman, E. 1947 The trephocytes and their function. Experientia, 3:442-451. Likely, G. D., Sanford, K. K., and Earle, W. R. 1952 Further studies on the proliferation in vitro of single isolated tissue cells. J. Nat. Cancer Inst., 73:177-184. Lilie, F. R. 1912 Studies of fertilization in Nereis. J. Exp. Zool., 12:413-476. CELL DIVISION Lima-de-Faria, A. 1949 The structure of the cen- tromere of the chromosomes of rye. Hereditas, 35: 77-85. 1950 The Feulgen test applied to centro- meric chromosomes. Hereditas, 36:60-74. Loveless, A., and Revell, S. 1949 New evidence on the mode of action of “mitotic poisons.” Na- ture, 164:938-944. Ludford, R. J. 1925 The general and experi- mental cytology of cancer. J. Roy. Micro. Soc., 45:249-299. Lwoff, A. 1950 Problems of Morphogenesis in Ciliates. John Wiley & Sons, New York. , and Dusi, H. 1935 La suppression ex- périmentale des chloroplastes chez Euglena mes- nili. Compt. rend. Soc. Biol., 779:1092-1095. Malheiros, N., de Castro, D., and Camara, A. 1947 Chromosomas sem centromero localizado. O case da Luzula purpurea Link. Agron. Lusitana, 9: 51-71. Manton, I. 1950 The spiral structure of chromo- somes. Biol. Revy., 25:486-508. Marquardt, H. 1941 Untersuchungen iiber den Formwechsel der Chromosomen im generativen Kern des Pollens und Pollenschlauchs von Al- lium und Lilium. Planta, 37:670-725. Marsland, D. 1951 The action of hydrostatic pressure on cell division. Ann. N. Y. Acad. Sci., 51:1327-1335. Matsuura, H. 1941 The structure and behavior of the kinetochore. Cytologia, 77:369-379. Mazia, D. 1952 Physiology of the cell nucleus; in Modern Trends in Physiology and Biochem- istry, edited by E. S. G. Barron. Academic Press, New York. ,and Dan, K. 1952 The isolation and bio- chemical characterization of the mitotic apparatus of dividing cells. Proc. Nat. Acad. Sci., 38:826- 838. , and Hirshfield, H. I. 1950 The nucleus- dependence of P?2 uptake by the cell. Science, 112:297-299. McClintock, B. 1932 A correlation of ring- shaped chromosomes with variegation in Zea mays. Proc. Nat. Acad. Sci., 78:677-681. Meier, R., and Allggwer, M. 1945 Zur Charac- terisierung zellteilungswirksamer Substanzen an der Gewebekultur. Experientia, 7:57-61. Metz, C. W. 1933 Monocentric mitosis with seg- regation of chromosomes in Sciara and its bear- ing on the mechanism of mitosis. Biol. Bull., 64: 333-347. Michel, K. 1950 Das Phasenkontrastverfahren und seine Eignung fiir zytologische Untersu- chungen. Naturwiss., 37:52-57. Mirsky, A. E. 1951 Some chemical aspects of the cell nucleus; in Genetics in the 20th Century, edited by L. C. Dunn. The Macmillan Co., New York, pp. 127-154. , and Ris, H. 1947 The chemical composi- tion of isolated Chromosomes. J. Gen. Physiol., 31:7-18. , and Ris, H. 1949 Variable and constant components of chromosomes. Nature, 763:666— 667. 123 , and Ris, H. 1951 The desoxyribonucleic acid content of animal cells and its evolutionary significance. J. Gen. Physiol., 34:451—462. Moellendorff, V. W. 1938 Ueber regulierbare Einwirkungen auf die Zahl und den Ablauf der Mitosen. Arch. Exp. Zellf., 27:1-66. Moenkhaus, W. J. 1904 The development of the hybrids between Fundulus heteroclitus and Menidia notata with especial reference to the behavior of the maternal and paternal chromatin. Am. J. Anat., 3:29-65. Monné, L. 1942 Ueber die Doppelbrechung der Kernhiillen. Ark. f. Zool., 34B, No. 2:1-81. , and Slautterback, D. B. 1950 Differen- tial staining of various polysaccharides in sea- urchin eggs. Exp. Cell. Res., 7:477-491. Monroy, A., and Montalenti,G. 1947 Variations of the submicroscopic structure of the cortical layers of fertilized and parthenogenetic sea-ur- chin eggs. Biol. Bull., 92:151-160. Montalenti, G., Vitagliano, G., and de Nicola, M. 1950 The supply of ribonucleic acid to the male germ cells during meiosis in Asellus aquaticus. Heredity, 4:75-87. Moore, A. R. 1933 Is cleavage rate a function of the cytoplasm or of the nucleus? J. Exp. Biol., 70: 230-236. 1938 Segregation of “cleavage sub- stance” in the unfertilized egg of Dendraster ex- centricus. Proc. Soc. Exp. Biol. & Med., 38:162- 163. Miihldorff, A. 1951 Die Zellteilung als Plas- mateilung. Springer, Berlin. Miintzing, A. 1944 Cytological studies of extra fragment chromosomes in rye. I. Iso-fragments produced by misdivision. Hereditas, 30:231-248. Norris, E. R., and Majnarich, J. J. 1949 Vitamin By4 and cell proliferation. Science, 109:32-33. Nussbaum, M. 1902 Ueber Kern-und Zelltei- lung. Arch. Mikrosk. Anat., 57:637-684. Ostergren, G. 1944 Colchicine mitosis, chromo- some contraction, narcosis and protein chain fold- ing. Hereditas, 30:429-467. 1949 Luzula and the mechanism of chromosome movements. Hereditas, 35:445-468. 1951 The mechanism of co-orientation in bivalents and multivalents. Hereditas, 37:85—156. Ogawa, K. 1949 Chromosome studies in the Myriapoda. Report 1. The chromosomes of Thereuonema hilgendorfi Verhoeff (Chilopoda), with special regard to the post-reductional sepa- ration of the sex chromosomes. Jap. J. Genet., 25:106-111. Oksala, T. 1943 Zytologische Studien an Odona- ten. Ann. Acad. Scient. Fenn., 4(ser. A. IV) :1-64, 1943. Olivo, O. M., and Delorenzi, E. 1933 Ricerchi sulla velocita di accrescimento delle cellule e degli organi. Arch. Exp. Zellf., 13:221-257. , and Slavich, E. 1930 Ricerche sulla velo- cita dell’accrescimento delle cellule e degli organi. Roux’ Arch. f. Entw-mech., 7217:96-110, 408-429. Ornstein, L. 1952 The distributional error in microspectrophotometry. Labor. Invest., 7:250- 262. 124 Orsini, M. W., and Pansky, B. 1952 The natural resistance of the golden hamster to colchicine. Science, 715:88-89. Patau, K. 1952 Absorption microphotometry of irregular-shaped objects. Chromosoma, 5:341- 362. , and Swift, H. 1953 The DNA content (Feulgen) of nuclei during mitosis in a root tip of onion. Chromosoma, 6:149-169. Painter, T. 1918 Contributions to the study of cell mechanics, II. J. Exp. Zool., 24:445-498. Pease, D. C. 1941 Hydrostatic pressure effects upon the spindle figure and chromosome move- ment, I. J. Morph., 69:405-434. 1946 Hydrostatic pressure effects upon the spindle figure and chromosome movement, II. Biol. Bull., 97:145-169. Peter, K. 1940 Die indirecte Teilung der Zelle in ihren Beziehungen zu Tatigkeit, Differenzie- rung und Wachstum. Z. f. Zellf., 30:721-750. Petering, H. G. 1952 Folic acid antagonists. Physiol. Rev., 32:197-213. Peters, J. J. 1946 Fig. 33. a, A haploid/diploid, white/dark mosaic axolotl] from a cross between a white (recessive) female and a heterozygous dark male, at age of six months. Eye, gills, and limbs are smaller on haploid side (after Humphrey and Fankhauser, 43). b and c, Two possible explanations of origin of this mosaic. b, “Partial fertilization”; egg nucleus carrying factor for white first divided alone; left-hand cell produced haploid, white side of body; in right-hand cell, the other descendant of the egg nucleus fused with a sperm nucleus carrying the gene for dark and gave rise to the diploid, dark side. c, Dispermy; one sperm nucleus carrying gene for white divided independently and furnished the nuclei for the haploid, white side of the body; a second sperm nucleus with gene for dark fused with the egg nucleus giving rise to the diploid nuclei of the right, dark side. the meiotic divisions is suppressed so that tetraploid embryos are produced. In contrast to most balanced polyploids, aneuploids have a greatly reduced viability and are usually abnormal in appearance. The addition of as few as one or two chro- mosomes to the diploid complement disturbs the genic balance sufficiently to make normal development a rare event. Of 377 embryos with known aneuploid chromosome number only thirteen lived for periods of three months or more. The great majority showed typical abnormalities at an early stage. Most commonly the circulation was either not established at all or remained subnormal, with frequent stasis and hemorrhages in the eral effect on the development and main- tenance of the circulation which in turn creates the fluid imbalance. As might be expected, hypertriploid or near-tetraploid larvae are more viable and normal, since the unbalance created by the addition of single chromosomes to a multiple complement is less severe. None of the more viable hyperdiploids have reached sexual maturity so far. How- ever, offspring have been obtained from non- viable embryos by transplantation of gonad preprimordia to normal diploid embryos in the tail-bud stage. Chromosome Mosaics. In most of the species of amphibians investigated, embryos appear, THE ROLE OF NUCLEUS AND CYTOPLASM either spontaneously or in various experi- ments (polyspermy, parthenogenesis, cold or heat treatment of eggs), which exhibit two or sometimes three different chromosome numbers in different regions which may vary in extent from a few cells to one-half of the body. Most frequent are haploid/diploid and haploid/triploid embryos, but several other combinations have been observed. 139 The haploid side of a mosaic generally shows a much more normal development than may be seen in completely haploid animals (compare Fig. 33a with Fig. 29c). Since mosaics possess a normal circulation, with both hapoid and diploid blood cells cir- culating through the haploid organs, the bet- ter performance of the latter may possibly be explained on this basis. However, parabio- Fig. 34. Development of dispermic sea urchin egg. a, Fusion of both sperm nuclei with egg nucleus, each containing a haploid set of 18 chromosomes; both sperm asters have divided. b, Random distribution of 54 chromosomes among spindles connecting the four poles of the mitotic figure. c, Anaphase; chromosomes from two adjacent spindles move toward each pole. d, Telophase; division of egg into four cells with different numbers of chromosomes. e, “Stereoblastula,” final stage of development. Cells of right-hand half disinte- grating. (After Boveri, 07.) Haploid/diploid mosaics are easily dis- covered among the offspring of crosses be- tween white and dark axolotls, if the haploid area is derived from the white (recessive) parent. Figure 33a shows an unusually regu- lar lateral mosaic of this type with a sharp line of demarkation almost exactly in the median plane. The organs on the haploid (left) side are smaller than those on the right. Dissection of the animal at the age of eight months further revealed that this in- dividual started out as a gynandromorph, with a testis on the left and an ovary on the right side. However, at the time of the autopsy the ovary had been almost com- pletely transformed into a testis under the influence of the male sex hormone produced by the testis on the haploid side. The origin of this lateral mosaic may be explained in either one of two ways which are described in Figures 33b and ce. sis between haploid and diploid embryos of T. pyrrhogaster, which also accomplishes a joint circulation, did not result in the ex- pected improvement of the haploid partner, aside from a reduction of the edema (Kaylor, 40). The presence of a partly diploid nervous system may be another factor favoring nor- mal functioning of the haploid side of the body. Finally, the proximity of normal dip- loid tissues may have a stimulating influence on adjacent haploid tissues, as has been shown by Hadorn (’35, ’37) in chimaeras composed of halves of diploid and haploid embryos. The Effects of Irregular Distribution of Chromosomes in Multipolar Mitosis. In sea urchin eggs that are fertilized by two sper- matozoa, both sperm nuclei fuse with the egg nucleus (Fig. 34a) so that three sets of chromosomes are present at the first division of the egg. In the majority of such eggs, both 140 THE NucLEUS AND CYTOPLASM IN DEVELOPMENT sperm asters divide to form a mitotic figure with four poles connected by spindles among which the chromosomes are distributed at random (Fig. 346). During anaphase, the four poles thus receive different numbers of chromosomes which are incorporated into four nuclei (Figs. 34c and d). Since these nuclei divide normally in subsequent mi- tity of chromosome material that causes the standstill of development, but the presence of abnormal combinations of individual chro- mosomes in one or more quarters of the blastula or pluteus. Normal development be- yond the gastrula stage requires the pres- ence of nuclei containing at least one com- plete set of chromosomes. Fig. 35. Diagram of development of polyspermic eggs of the newt, following removal of egg nucleus; frequent disengagement of nuclear and centrosomal cycles producing abnormal mitoses. a, Five sperm nuclei present, two of which are about to fuse. b, First mitotic cycle; tetrapolar mitosis at upper right, two tripolar mitoses at left, monaster at lower right. c, Second mitotic cycle; several cells again contain abnormal mitotic figures. d, Section through final stage of development reached; blastula with wide variety of chromo- some numbers in different areas. toses, the abnormal chromosome sets are re- produced and handed on to all the cells descended from each of the first four blasto- meres. A blastula results which consists of four areas differing in chromosome num- ber as is shown by differences in the size of the nuclei; the individual cells are too small to allow actual chromosome counts at this stage. Some of these eggs never develop be- yond a “stereo-blastula’” in which one or more quarters disintegrate, causing the death of the whole embryo (Fig. 34e). Others may form abnormal plutei in which again a regional breakdown of the cells takes place. Boveri (’02, ’07), in a masterly analysis of the development of dispermic eggs, showed that it is not a deficiency in the mere quan- In polyspermic eggs and egg fragments of newts from which the egg nucleus has been removed, multipolar and other abnormal mi- toses occur frequently. Because of the large size of the eggs, several mitotic figures may be present in the same egg and cause the formation of multiple cleavage furrows (Figs. 35a and b). The cytological picture is com- plicated still further by the fact that multi- polar or monocentric mitoses may again occur in individual blastomeres during the second or a subsequent mitotic cycle (Fig. 35c). Counts of chromosomes in the blastula re- veal a highly complex pattern of abnormal, mostly aneuploid, chromosome numbers (Fig. 35d). Since, in Triturus, most of the 12 chromosomes of the haploid set may be rec- THE ROLE OF NUCLEUS AND CYTOPLASM ognized individually by differences in size and shape, it is possible to analyze the ab- normal chromosome complements in detail (Fankhauser, °34) and to demonstrate di- rectly their random composition and ex- treme unbalance. As would be expected, all of these eggs die in the blastula stage. Recently Morgenthaler (’48) transplanted pieces from androgenetic Triturus blastulae, of constitution presumably similar to that shown in Fig. 35d, to the flank of normal neurulae. Sections of the host embryos fixed in an early tail-bud stage showed the pres- ence of mitotic figures in the graft, some of which contained fewer than twelve chromo- somes. This indicates that subhaploid cells may continue to divide and survive for some time in the normal environment of a diploid embryo. QUALITATIVE CHANGES: EXPERIMENTS INVOLVING HYBRIDIZATION BETWEEN TWO SPECIES The introduction of a nucleus of one spe- cies into the egg cytoplasm of another offers an opportunity to study the effects of the combination of two components that presum- ably differ in a qualitative way. A variety of unusual modifications of the processes of development are produced that have been analyzed most intensively in echinoderms and amphibians. DIPLOID HYBRIDS Echinoderms. The development of most echinoderm hybrids has not been followed beyond the pluteus stage. The internal phe- nomena of fertilization may be completely normal, as in the cross Echinus esculentus 9 darkest I I I IZ Ww A Instars THE NUCLEUS AND CYTOPLASM IN DEVELOPMENT different rates of gene-controlled, qualita- tively identical reactions and has shown how such a rate concept fits a great variety of phenomena in developmental genetics. In some cases, different rates of production of pigments or of growth, controlled by differ- ent alleles, have been demonstrated by direct observation (Fig. 38A and B). These different rates have been reported by Goldschmidt (27) for larval pigmentation in Lymantria Head length (mm.) aay ol 30 Sexual maturity Sexual maturity j of normal males of dwarf males begins begins LAS) iO 20° 30) 40) 50: GO: 70/780 B Time in days Fig. 38. A, Pigmentation curves of sevendifferent geographic races of the gypsy-moth Lymantria dispar (after Goldschmidt, ’27). B, Growth curves for a normal (GG) and a “dwarf” (gg) strain of Gammarus chevreuxi (after Ford and Huxley, ’29). actions in the bread mold Neurospora, are under immediate gene control in the sense that specific alleles impress enzymatic spec- ificity on cellular constituents (Beadle, °45; Bonner, °48; Horowitz, °50). Attempts to prove this hypothesis have met with great difficulties (Emerson, °50). The examples of gene-controlled antigens and enzymes indicate that the primary prod- ucts of the action of different alleles may be qualitatively different. There are other examples which suggest quantitative differ- ence in allelic action. A series of multiple alleles of a given gene usually affects the same phenotypic trait, in a graded manner. In Drosophila, for instance, different alleles of the vestigial locus control the appearance of a fully formed wing, of a slightly nicked wing, a deeper notched one, a short vestig- ial wing stump, and a wingless condition (Mohr, *32). Goldschmidt (38) has given an interpretation of these cases in terms of dispar, and by Ford and Huxley (29) for rate of growth in Gammarus chevreuxi. Different rates of apparently identical processes do not need to depend on different rates of identical primary genic reactions. Different alleles may control the production of qualitatively different primary products which in turn may influence identical re- actions in quantitatively graded fashion (Wright, *45a; Stern, °48). The theory that different alleles of a gene cause their different developmental effects by influencing the rate of gene-controlled reactions has led Goldschmidt to the pre- diction that the different phenotypes can be produced purely environmentally in devel- oping animals of identical genotype. External agents, particularly temperature, should change differentially the many gene-con- trolled reactions and thus act, at appropriate stages, essentially like shifts in reactions due to different alleles. This expectation has been GENE ACTION verified over a wide range of phenotypes in various organisms, e.g., Lepidoptera (Kuhn and Henke, ’29-36), Drosophila (Gold- schmidt, ’35a,b, 737; Henke, v. Fink, and Ma, 41), chickens (review by Landauer, *48) and mice (Russell and Russell, *48). In Drosophila, for instance, heavy shock of short duration with high temperatures, or less striking temperature increases for longer periods, produce modifications, in genetically normal individuals, of wings and halteres, eyes and aristae, legs and body wall, and bristles. The modifications are of numerous % of modification ine) ro) Hours: 2 4 6 8 10 12 Stage: — L___] Egg Blastoderm Segmentation 155 that the interference with development by an external agent may well occur at the same time, and perhaps in the same manner, as the developmental shift under the in- fluence of the mutant which is phenocopied. In other instances, it has been demonstrated that the sensitive period during which a phenocopy may be induced is later than the phase at which the mutant first shows its developmental effect. Obviously in these latter cases the phenocopy does not actually imitate the course of events initiated by the mutant allele whose effects it simulates. | | | 3rd Instar Prepupa Pupa Fig. 39. Sensitive periods of production of developmental anomalies in Drosophila melanogaster: tetraptera- like wing modification of the halteres; abnormal abdomen; abnormal bristle growth on halteres (after Henke and Maas, ’46). types, as in the case of wings, where they affect shape and size, venation and distri- bution of bristles. Specific relations have been demonstrated between the stage in development of the egg, larva, or pupa, and the production of the various modifications (see Fig. 39). Thus, a period sensitive to moderate temperature ef- fects occurs between two and four hours after fertilization—before completion of the blastoderm—resulting in the later morpho- genesis of the metathoracic segment ap- proaching that typical for the mesothorax. Shortly later an embryonic sensitive period may be used to initiate processes leading to abnormal abdominal segmentation. Modifi- cations of imaginal bristles may be induced in newly hatched larvae, of wing shape in larvae of the second instar, and of the eye surface in those of the third instar. Likewise, all these may be induced by treatment of prepupae and pupae (Henke and Maas, *46). Goldschmidt introduced the name pheno- copy for these modifications, since they seem to imitate in genetically normal strains the phenotypes known to be caused by mutant allelic substitutions. Comparisons between the development of mutant phenotypes and of phenocopies have shown in some cases There is, however, an arbitrariness in choos- ing a single specific mutant and comparing it with a given phenotypic modification. It is known that many different mutant loci may cause the development of similar or identical final phenotypes and it seems pos- sible that every environmentally induced developmental modification may have its true genetic counterpart in one of the various known, or not yet discovered, mutants. While it is necessary to be aware of the usually somewhat arbitrary nature of desig- nating a modification as the phenocopy of a given mutant, it seems likely that a pheno- copy will be found for each mutant, and a mutant for each phenotypic modification. Therefore, the terms “true” and “false” phenocopies, which suggest that an environ- mentally induced phenotype either corre- sponds in its developmental origin to a mu- tant phenotype (“true”), or only simulates it (“false”), have only limited validity. TIME OF GENIC ACTION Gene-controlled events are known to occur at various stages of development. One of the earliest evidences for genic action are “size genes” brought into the egg by the sperm in 156 crosses between large and small races of rabbits (Gregory and Castle, ’31). The rate of cleavage of eggs, which is correlated with adult body size, is increased under the in- fluence of paternal genes as early as at the third or fourth cell division. Slightly later effects of genes are known in the early em- bryology of Drosophila (Poulson, ’40, ’45), in the morphology of sea urchin larvae derived MUTATIONS gfgh FAULTY BONE DEVELOPMENT wWhANEMIA UU YUROGENITAL ABNORMALITIES, SdSd\SPINA BIFIDA, GCLOACA TT\FAILURE OF NOTOCHORD, SOMITES, POSTERIOR TRUNK REGION AND UMBILICAL CIRCULATION Ki Ki ALLANTOIC DERIVATIVES tPtTAFAILURE OF ORGANIZATION AY AYVFAILURE OF IMPLANTATION IMPERFORATE ANUS DUPLICATIONS, ABNORMALITIES OF AND MESODERM FORMATION THE NUCLEUS AND CYTOPLASM IN DEVELOPMENT much earlier, invisibly preparing a situa- tion which is a prerequisite for later effects. There is evidence for both types of activity, the immediate and the distant. Examples of the latter, in purest form, are presented by the so-called maternal effects, in which genes present in the cells of a female, including her immature, premeiotic egg cells, cause developmental effects in her offspring even NORMAL STAGES 20 BIRTH ESTABLISHMENT OF MATERNAL CIRCULATION 8 TURNING OF EMBRYO CLOSURE OF NEURAL FOLDS SOMITES ALLANTOIS NOTOCHORD 7 PRIMITIVE GUT (TRANSITORY)\EGG CYLINDER PRIMITIVE STREAK a MESODERM 5 IMPLANTATION 4 BLASTULA {tPFAILURE OF PREIMPLANZ3 MORULA TATION STAGES 2 CLEAVAGE FERTILIZATION DAYS Fig. 40. Some of the lethal mutations affecting embryonic processes in mice (from Dunn, *49). from species crosses (Moore, °43), at the morula stage of mice homozygous for cer- tain tailless alleles (Dunn, ’49; see Fig. 40), and at the blastocyst stage of mice homo- zygous for the yellow-lethal gene (Robert- son, 42). Figure 40 illustrates other em- bryonic processes in mice affected by lethal mutations. While it is obvious in these examples that genic action occurs within a_ short initial period, namely that preceding the time of an observable effect, it is more difficult to make statements when observable effects occur late in development. Here, genic action may have immediately preceded the observable event or it may have taken place though these genes may not have been in- cluded in the zygotic nuclei. This is the well known explanation for the inheritance of dextrality and sinistrality in the snail Lim- naea (Boycott, Diver, and Garstang, ’30; Sturtevant, ’23). The type of configuration which an individual will attain is determined by a pair of alleles, but not by those present in his own genotype. Rather, the genotype of the maternal tissue in which the oocyte is formed fixes the fate of the next generation. This is accomplished by imparting to the cytoplasm of the egg the property, unknown in its essence, of initiating the sequence of its spiral cleavages in either a clockwise or a counterclockwise direction. GENE ACTION In other examples of maternal (prefertil- ization) effects the path from genic action to external expression is much longer. Thus, the sexual differentiation of the genital sys- tem in the adult moth Lymantria dispar de- pends on the genotype of the female parent (Goldschmidt, ’34), and so does adult testis size in hybrids between Drosophila pseudo- obscura and D. persimilis (Dobzhansky, ’35). Another rather well understood situation con- cerns pigmentation of the larval eyes of the moth Ephestia kiihniella (Caspari, °33; Kiihn, Caspari, and Plagge, ’35). Here, the gene A is concerned with the elaboration of kynurenin, a colorless precursor of certain pigments. Usually caterpillars of the geno- type aa do not form pigment unless the mother was Aa. Furthermore, implantation of AA tissues into aa females results in a situation such that the newly hatched cater- pillars from a mating of these females to aa males have pigmented eyes and skin. Gradu- ally, however, the pigment disappears again during growth of the caterpillars. Appar- ently, the effect of the gene A in the mother, or the implantation of AA organs, leads to storage of kynurenin or of a derived pig- ment precursor in the egg, its presence be- coming evident only when ocelli and skin have been differentiated and its later disap- pearance being due to depletion of the stored precursor. In contrast to these examples in which genic action precedes the developmental ef- fect by a strikingly long period are others in which genic action can be shown to occur late in development, relatively shortly prior to the appearance of a specific trait. A means for studying such late gene activity is avail- able in certain genetic mosaics (Sturtevant, 32). If, in the course of development, somatic mutation or chromosomal changes occur in a cell and if, as a consequence, a new pheno- type arises, then the conclusion is justified that the developmental processes in this cell and its progeny tissue were not independent of, but susceptible to, control by the changed genotype, although the processes themselves may have previously been initiated geneti- cally. Most cases of mosaics in Drosophila and other insects, as well as some rare somatic mosaics in birds and mammals, be- long to this category. These phenotypic mo- saics indicate either that the genic action sets in normally after the time of the “muta- tional” accident, or that even a shortened active period is sufficient to produce an ef- fect. Often this effect is as strong as if it had been present from the beginning of develop- 157 ment. Sometimes a weakened or otherwise modified phenotype results. Here we may witness the outcome of abbreviated gene ac- tivity, or we may wonder whether prior action of the original genotype has left a shortage of some specific substrate on which the new genotype has to act. Still another cause for a modified phenotype may lie in some interaction between the genetically dif- ferent tissues of such a mosaic organism (see pp. 159-160). TYPE OF GENIC ACTION Most genes produce multiple effects in development (‘‘pleiotropy”). For instance, in the so-called “Frizzle” fowl, animals ho- mozygous for the Frizzle allele show not only the abnormal feather curliness which has given them their name but also the follow- ing traits: “The heart rate is accelerated and both ventricles of the heart are hypertro- phied. The blood volume is greater than normal, and the spleen is enlarged. Weight (relative to the body) of crop, gizzard, pan- creas and kidneys, length of the coeca, ca- pacity of the intestinal tract (from duo- denum to large intestine) are in excess of what is normal. The adrenals are hypertro- phied. The white cells of the blood may show a relative agranulocytosis and relative gran- ulopenia. The metabolic rate is above nor- mal, food consumption increased, water va- porization reduced. The thyroids may be re- duced in size or much enlarged. . . . The ovaries mature late and may remain atrophic; the testes frequently show interstitial edema and tubular damage. General body growth falls below normal. Viability is re- duced.” (Landauer, ’46.) Another example of multiple genic effect, in a mammal, is furnished by the gene for myelencephalic blebs in mice. Homozygous mb individuals frequently have (1) defec- tive eyes, (2) polydactylous or syndactylous feet, and (3) abnormal hair growth (Grine- berg, °43). In Drosophila, the most striking examples for multiple effects of genes come from studies specifically designed to investigate the question of the frequency of such mul- tiple effects (Dobzhansky, ’27; Dobzhansky and Holz, ’43; Schwab, ’40). Mutant strains which had been characterized by means of observed effects on such external traits as eye pigment or body coloration were ana- lyzed for the type of growth of an internal organ, the female spermatheca. In several instances it was discovered that the external 158 THE NUCLEUS AND CYTOPLASM IN DEVELOPMENT effects of the mutant genes were accom- panied by internal effects. Since the sperma- thecae had been selected arbitrarily as pos- sible indicators of multiple genic effects, it may be concluded that many other traits would likewise exhibit specific phenotypes as the result of different mutant alleles. The phenomenon of multiple effects of a gene poses the important problem of mul- tiple primary gene action versus unitary primary gene action with secondary, mul- tiple developmental results. In several in- stances, it has been possible to trace back developmentally the multiple observed ef- fects to a single primary cause. Thus, in the Frizzle fowl, the abnormality of the feathers, which wears off quickly, is the initial event which seems to account for all others. The loss of plumage results in loss of body heat, this in turn in subnormal body temperature and increased metabolic rate. “From here, one cHain of adaptations leads to accelera- tion of heart rate, ventricular hypertrophy, enlarged blood volume, and excessive size of spleen. In order to sustain the accelera- tion of these vital processes more food is consumed and this, in turn, produces en- largement of the intestinal parts, hypertro- phy of the pancreas and of the kidneys. A heavy burden falls on the adrenals and thy- roid. . . . Reduced vaporization is another adjustment useful to the body economy.” (Landauer, 46.) In the “myelencephalic blebs” (mice) the primary event is the appearance, during em- bryonic stages, of large blisters under the skin of the head (Bonnevie, 34). These blisters are filled with fluid of cerebrospinal origin which has escaped through a foramen in the myelencephalon. At first, these blebs move under the skin to different parts of the body. Finally, they come to rest and gradually disappear before the termination of gestation. Since the blisters are under in- ternal pressure, it is not surprising that dis- turbances of development result in their neighborhood. Those blebs which come to rest in the region of the developing eye may produce eye defects, while others whose final position is at the limb buds cause the various types of foot abnormalities. Other developmental “pedigrees of causes” of multiple genic effect have been worked out, in both rats and mice, by Griineberg (38). He is inclined to postulate that all pleiotropic genic action depends on a single primary effect. It can, indeed, not be doubted that developmental studies of multiple gene effects not only repeatedly have succeeded in tracing back the pedigree of causes to a single basic event but also that more such results can be expected in cases which still defy such an analysis. However, there seems to be no a priori reason why primary gene action may not sometimes be of multiple nature. Genic action has often been com- pared to that of enzymes, and intrinsic iden- tity of genic action and gene-controlled en- zymatic action has been suggested (Beadle, °45; Emerson, °45). It is significant, from this point of view, that at least one enzyme, tyrosinase, seems to have two very different activities (Nelson and Dawson, 44). There- fore, the existence of multiple primary ac- tion of a gene would be compatible with concepts which consider genic action within the frame of protein chemistry. Actually, it will be very difficult to decide between multiple primary action and early diversifi- cation of unitary action unless the details of the earliest gene-controlled cellular phys- iological processes are fully understood. An example, from Neurospora, may illustrate this statement. A certain mutant of this fungus is unable to synthesize adenine; in- stead, it makes a purple pigment not pres- ent normally (Mitchell and Houlahan, *46). Although a mutant allele which can syn- thesize reduced amounts of adenine and at the same time forms the purple pigment has not yet been described, we may assume its existence for purposes of this discussion. There would, then, exist an allele with two intracellular effects, adenine synthesis and pigment formation. However, these two ef- ects would still be found to depend on a single activity, since it is probable that the pigment is nothing but a derivative of a precursor of adenine. In a cell with a normal allele the precursor is transformed into aden- ine, but in the mutant cell with its impeded transformation to adenine, the precursor accumulates and becomes observable as a “second” gene effect. In a multicellular organism the problem of multiple versus unitary gene action refers not only to the primary action at the molecu- lar level but also to the site of action in the organism as a whole. The abnormality of feathers in the Frizzle fowl, which is the primary event in the development of the various symptoms, has a single site, although extensive in space, namely, all cells of the epidermis concerned in the production of feathers. Similarly, in mice with myelen- cephalic blebs, there is a primary site of developmental abnormality, the region of the embryonic brain. On the other hand, in GENE ACTION Drosophila it is not easy to discern, at pres- ent, a primary site from which effects radi- ate out both to the formation of pigment in the eye and to specific shape of the sperm- athecae. While such a developmental site may still be found, it is perhaps easier to conceive of genic action occurring independ- ently at two different sites, the eye anlagen and the spermathecae. A similar interpreta- tion has been considered in respect to the gene locus W in mice, which controls both pigmentation and hematopoiesis (E. S. Rus- sell, °49). More than one site may also be involved in the ‘““myelencephalic blebs” mice, since the effect of the mb gene on hair erowth has not been related to the appear- ance of blebs but seems to be separate in origin. The action of the same gene at different sites (“repetitive genic action”) (Stern and Schaeffer, ’43) raises the problem of multi- ple or unitary primary action in a new form. Assuming a single activity at each site, it would be possible for this activity to be identical in the several sites or to be differ- ent from site to site. Some further comments on these problems will be made when gen- eral aspects of genic mechanisms of differ- entiation are considered (see pp. 163-166). DEPENDENT AND AUTONOMOUS DEVELOPMENTAL PROCESSES The site of genic action may or may not be the site where its final developmental effect becomes apparent. The example of myelencephalic blebs in mice has shown that the earliest known developmental site is in the region of the brain, whereas the later phenotypic effects are on eyes and _ feet. The embryological analysis of the sequence of events suggests strongly that the appear- ance of blebs is the cause of the defects in the distant eyes and feet. It would be pos- sible, however, to argue that the gene mb which leads to bleb formation is also an agent in the anlagen of eyes and limbs which makes them react in their abnormal way to the migrating blebs. Such an argument, while unlikely, can best be tested by experi- ment. A classic experiment of this type re- lates to a genetically dwarf strain of mice (Smith and MacDowell, ’30). In this strain, a histological study of the pituitary gland showed deficiencies in the anterior lobe. It was obvious to suggest that the growth re- tardation of the animals was caused by means of a deficiency in the growth hor- mones normally produced in the pituitary. v9 It could be shown that hormonal deficiency indeed is the essential factor and _ that the slowly growing tissues which were of the same abnormal genetic constitution as the pituitary nevertheless had normal growth potencies: artificial supply of growth hor- mones by implantation of normal pituitaries leads to successful growth. Such dependent, gene-controlled, differen- tial development is undoubtedly frequent. Particularly striking examples have been provided by some lethal genotypes. Cultures of some tissues from lethal brachyuric mouse embryos (Ephrussi, ’35), and from the homo- zygous lethal Creeper fowl (David, ’36), have been successfully grown beyond the normal life span of the doomed donor. Vari- ous tissues of lethal Drosophila larvae have survived and undergone development if trans- planted to normal hosts (Hadorn, ’48), or present in small mosaic patches on nonlethal individuals (Demerec, ’34, 736). The death of the whole, in these lethals, must be due to lethal action in some particular region or regions, an action which unavoidably leads to death of the potentially viable tis- sues. On the other hand, the existence of “‘cell- lethals” discussed earlier (p. 152) shows that the specific differentiation of cells often de- pends on the action of their own genes. This is obviously true even in dependent differentiation, where a primary site of gene action, as perhaps within the pituitary cells of dwarf mice, must form the starting point for secondary dependent events. In many cases the final phenotypic effects are the re- sults exclusively of genic action within the cells concerned in the phenotype. Studies of genetic mosaics (see p. 157) as well as of transplants have demonstrated autonomous differentiation in coat colors of mammals (Bhat, ’49), in eye and body pigmentation, bristle shape, wing type, and many other traits of insects. Particularly the gynanders of Drosophila, and of many other insects, are evidence of a cellular autonomy for they show sharp phenotypic lines of demarcation corresponding to the genetically diverse tis- sue areas (Sturtevant, ’32). There is no general rule as to when de- pendent and when autonomous differentia- tion may be expected. While most eye color mutants in Drosophila behave autonomously, the difference between vermilion and nor- mal red eyes, or between cinnabar and nor- mal red eyes, is due to the presence or absence of circulating or diffusing substances, kynurenin and hydroxykynurenin, respec- 160 THE NUCLEUS AND CYTOPLASM IN DEVELOPMENT tively, which are elaborated by such organs as the Malpighian tubes and the fat bodies (Beadle, ’37; Beadle, Tatum, and Clancy, 39). In another Drosophila case a combina- tion of autonomous and dependent processes has been shown to exist. The difference be- Fig. 41. Mosaic eye of Habrobracon juglandis. Upper left: ivory (o'C); lower right: cantaloupe (Oc). Note the wild type coloration at the junc- tion of the two tissues and the decreasing wild type coloration within the cantaloupe region. (After A. R. Whiting, 734.) tween the narrow bar eye of Drosophila and the normal round eye is caused in a twofold way: (1) an embryonic or very early larval reduction in the size of the eye disc which may be an autonomous result of the geno- type, and (2) a later larval interference with the formation of a specific diffusible sub- stance which is instrumental in transform- ing cells of the eye disc into ommatidia (Chevais, *43). Genic action at a distance may be brought about by substances whose transfer from one part of the body to another is mediated by a circulatory system. In some instances action at a distance results from short range diffu- sion of gene-dependent substances rather than from equal distribution of such substances throughout the body. A striking example has been described in the parasitic wasp Habro- bracon juglandis (A. Whiting, ’34). (See Fig. 41.) The normal stock possesses a black eye pigment while the coloration in two non- allelic mutant stocks is ivory and cantaloupe, respectively. We may signify the phenotype black as the genotype “not-ivory, not-canta- loupe,” the phenotype ivory as genotype “ivory, not-cantaloupe,” and the phenotype cantaloupe as genotype ‘“not-ivory, canta- loupe.” From double nucleus eggs individ- uals have been raised which carry in one part of an eye the genetic constitution “ivory, not-cantaloupe” and in the other part ‘“not-ivory, cantaloupe.” On the whole there is autonomy of phenotypic expression as wit- nessed by a large ivory and a large canta- loupe area. However, at the border zone between the two a strip of cells is neither ivory nor cantaloupe but black. This finding can readily be explained if one assumes that a substance produced in only one of the two genetically different tissues diffuses into the other and thus reconstitutes the full effect of a “not-ivory, not-cantaloupe” genotype. The black zone shows a sharp border with the cantaloupe region but gradually grades over an orange coloration into the ivory part. The orange is known as an intermediate step between full black and ivory pigmenta- tion. From these facts and others it has been inferred that the diffusible substance is pro- duced in the “not-ivory, cantaloupe” region by the dominant not-ivory gene. An unusually interesting nonautonomous differentiation, also in mosaics of Habro- bracon, concerns the external genitalia (P. W. Whiting, Greb, and Speicher, ’°34; Whiting, *40). Here, eggs develop occa- sionally in which the mature egg nucleus as well as a polar nucleus take part in parthen- ogenetic cleavage. Since the diploid females Fig. 42. Ventral view of genitalia of a male com- posed of tissues of two different sex genotypes. The right outer clasper is feminized, resembling a mi- nute sensory female gonapophysis. (From P. W. Whiting, 40.) are heterozygous for a pair of genetic sex ‘factors,’ and their sons are usually hemi- zygous for either one or the other of the sex factors, the offspring developing from the binucleate eggs may be genetically mo- saic, pessessing one of the sex factors in some of its nuclei and the other sex factor in GENE ACTION the remainder. Either constitution typically determines male differentiation. In accord- ance with expectation the mosaic individuals were males but frequently showed some feminized genital structures (Fig. 42). Ap- parently, the two alternative kinds of sex factors in the nuclei of the mosaics control the production of two kinds of diffusible male determining substances which are com- plementary in such a fashion as to cause, jointly, female differentiation. Restricted diffusion of genetically con- trolled substances is also involved in the specific shape of the adult testes of different species of Drosophila (Stern, *41). In some species these organs are ellipsoidal, in others spiral or helical. Normally, the uncoiled larval testes attain the specific adult shape af- ter having become attached to the vasa defer- entia which originate independently. An in- ductor-reactor relation in regard to spiraliza- tion exists between vas and testis in “spiral” species with transplanted ellipsoidal testes. Interspecific transplantations of larval testes show that the typical different adult testes shapes are not autonomous responses to the stimulus emanating from the vasa but that the vasa of the different species cause dif- ferential growth of testes. The vas of a “non- spiral” species causes limited, more or less equally distributed, growth of an attached testis of even a spiral species, and the vas of a spiral species causes unequally distributed growth of an attached testis of even a non- spiral species. These growth differentials are, presumably, due to diffusion into the react- ing testicular tissue of specifically distributed growth promoting substances in the vas. An analysis in terms of individual genes of the differences between the different noninter- breeding species is not possible. Therefore, in the present state of our knowledge one may designate the substances involved as “genom dependent” rather than as gene dependent (Hammerling, *46). HOMEOTIC MUTANTS In D. melanogaster, a number of mutant genes have been found which direct the de- velopment of embryonic anlagen of a body segment into new channels so that they dif- ferentiate into organs normally characteristic of other segments (see Villee, *42). Such mutant genes thus produce the kind of changes which fall under the term homeosis (Bateson, 94) and which have been dealt with extensively by earlier students of mor- phogenesis (Herbst, 1896-1901; Przibram, 161 10). Examples of homeotic mutant genes are aristapedia, which changes the arista of the antenna to a leg, proboscipedia, which changes mouth parts into leglike organs or aristae, and bithorax, which changes halteres into winglike appendages (Fig. 43A-D); these and other homeotic mutants are listed by Herskowitz (49). It is not known how these simply and typically inherited genes produce striking morphogenetic effects. The appendages of adult insects develop from separate embryonic imaginal discs, and the nature of the stimuli which normally cause the differentiation of an antenna from a head disc and of a leg from a thoracic disc has not been established. It could be that the differ- ent regions of the embryo or larva determine the specific differentiation of originally toti- potent discs by means of specific evocators. It could also be true that more general prop- erties imposed on the discs in the different regions, for instance, different rates of growth or of metabolism, lead to autonomous chan- nelling of development within the disc into one or another of the alternative courses. Similarly, in a homeotic mutant, it is con- ceivable that the unusual development of a disc is determined by an evocator which is not typical for the region of the specific disc, or that a change in intrinsic general proper- ties of the disc results in the homeotic dif- ferentiation. A third scheme, proposed by Goldschmidt (?38), assumes that at different developmental periods different kinds of evo- cators are present in the embryo or larva and that only discs in specific stages of de- velopment are able to react to these evoca- tors. A mutant which shifts the rate of de- velopment of a still totipotent disc so that it becomes reactive at an earlier or later time than normal will expose it to the evocator of that period and thus cause the production of a homeotic organ. For the mutant arista- pedia, studies by Vogt (46) have not sup- ported this interpretation. GENETIC ASYMMETRIES A special group of gene-controlled pheno- types is that of hereditary asymmetries. Many animals typically show specific asymmetries, for instance, of the viscera in vertebrates (Ludwig, ’32). The geneticist cannot con- tribute directly to the understanding of these basic asymmetries. However, specific genes are known which change the type of asym- metry normally present or produce asym- metries in structures which usually have a symmetrical arrangement. 162 Among the latter is a class of asymmetries which are not fixed in relation to the main axis or secondary axes of the animal; thus both right-left and left-right arrangements occur in a group of affected individuals. For instance, eye defects in the myelencephalic strain of mice may be present on only one THE NUCLEUS AND CYTOPLASM IN DEVELOPMENT asymmetrical ones. Two types of symmetrical individuals are found in each instance, those with both sides normal, and those with both sides affected. The problem of asymmetry is thus bound up with that of variable expres- sion of the genotype from one specimen to the next. Basically the genes involved are Fig. 43. Homeotic mutants of Drosophila melanogaster. A, Frontal view of normal head; a, antenna with arista (ar); pb, ventral section of proboscis (extended). B, Antenna of “aristapedia” (after Balkaschina, ’29). C, Two examples of proboscis of “proboscipedia” (after Bridges and Dobzhansky, ’33). D, Bithorax; the halteres are changed into wing-like structures. (Original, courtesy of E. M. Wallace.) side, but the two types “right eye normal, left eye abnormal” and “right eye abnormal, left eye normal” occur with equal frequency (Bonnevie, 734). Similarly, in Drosophila, in the Dichaete strain, there are often present only three of the four normally occurring thoracic dorsocentral macrochaetae, but the two types “two right bristles, one left” and “one right, two left” again are equally nu- merous (Plunkett, ’26). The developmental interpretation of these asymmetries is easily suggested by the occurrence of symmetrical individuals of the same genotypes as the such that they do not always actually pro- duce the phenotype by means of which their presence can be recognized. Depending on variations in the general genetic background provided by the rest of the genotype, or on variation in the external or internal environ- ment of the developing system, the action of a given gene may not always “penetrate to the surface” of observable phenomena. In the two examples given, the existence of mice or flies which are normal on both sides shows complete lack of penetrance of the po- tentially abnormal genotype, while the ex- GENE ACTION istence of symmetrically abnormal animals is evidence of presence of penetrance of de- velopmental action on both sides. Since the developmental mechanism which brings about an eye defect or absence of a bristle on one side is independent of that which acts on the other side it is not surprising that the phenomenon of variable penetrance often leads to phenotypic asymmetries. Sometimes the two types of right-left and left-right asymmetries do not occur with equal frequency, as for instance in harelip and cleft palate in man (Fogh-Andersen, 43). The genetic constitutions which under- lie these abnormalities often are not pene- trant, so that some individuals may appear fully normal. Some other individuals show a bilateral defect but in the majority the de- fect is asymmetrical, occurring either on the right or left side. Among the asymmetrical individuals those with left harelip or cleft palate are two to three times as frequent as those with right defects. No specific explana- tion can be given for this preference. In a general way it is, however, obvious that the typical asymmetry of the human embryo fur- nishes internal environmental differences be- tween the left and right side and that it is not surprising if the variable penetrance of the harelip—cleft palate genotypes responds developmentally to these differences. It might well be that cases exist where a mechanism of variable penetrance always responds to the basic asymmetry by showing an effect on one specific side and never on the other. However, no well established example seems to be known. Some types of inherited asymmetries can- not easily be interpreted by the hypothesis of variable penetrance, which rests heavily on the observation of symmetrically affected and symmetrically nonaffected individuals side by side with the asymmetrical ones. The clearest divergent case is that of an asym- metrical spotting pattern in the beetle Bruchus quadrimaculatus (Breitenbecher, °25). The normal females possess two black spots bilaterally located on each elytrum, but in a recessive mutant strain the females have two black spots on one elytrum and two red ones on the other. The asymmetry is present in all individuals of the relevant genotype but the two types “black spots at right, red spots at left’? and vice versa, are not fixed—half of the females belong to one and half to the other type regardless of the phenotype of their ancestors. Breitenbecher suggested that “probably some delicately ad- justing mechanism shifts this asymmetry 163 right or left according to chance.” Dahlberg (43 and earlier) has elaborated a hypothesis according to which in the cells of the em- bryo ‘a particular gene may promote un- equal distribution of cytoplasmatic material without deciding the orientation of the dis- tribution in a specified relation to particular morphological axes. .. . The result is to dif- ferentiate one cell lineage with an excess from a second with a deficiency of the in- clusion.” If an excess inclusion or its de- ficiency is decisive in determining the manifestation of a phenotypic trait, as for instance, position of pigmentation in Bruchus, an explanation of its always asymmetrical presence is provided. Dahlberg suggests that his hypothesis should even be applied to the interpretation of asymmetries expressed by traits which often occur also symmetrically. Such an extension of the scheme seems un- necessary at present. A unique situation of genetically caused asymmetry is given by the position of the ex- ternal male genitalia in D. melanogaster. Several mutants have been found which cause a rotation of the genitalia. ‘““Rotated-penis,” for instance, consists of a counterclockwise rotation, to a variable degree. This externally appearing asymmetry in a typically sym- metrically arranged system of parts is in reality evidence of an approach to internal symmetry which is absent in nonmutant males. In a normal nonmutant male the ex- ternal genitalia may be described as rotated clockwise through 360 degrees as compared with the position of the genitalia in the fe- male. The rotation is not apparent externally but is evidenced internally by a spiral coun- terclockwise looping of the sperm duct about the intestine. In the males of the mutant the internal rotation is partly undone by coun- ter-rotation of the external genitalia which now appear in an abnormal, asymmetrical sidewise position, or upside-down (Bridges, in Morgan, Sturtevant and Bridges, ’29). The genetic determination of direction of coiling in snails has been mentioned earlier (p. 156). It would be tempting to speculate on the causes of the two types of coiling in terms of asymmetrical arrangements at the molecular level! DIFFERENTIATION The most general problem related to ge- netics and development is posed by the fact of embryonic differentiation “in spite of” sup- posedly equal genic endowment of the cells of all differentiating parts of the embryo. A 164 THE NUCLEUS AND CYTOPLASM IN DEVELOPMENT general concept of genic action in develop- ment is available which gives an intelligible picture of the type of events involved and thus overcomes the apparent contradiction which is expressed in the phrase “in spite of.” Before discussing this concept it may be pointed out that the assumption of equal genic endowment of all somatic cells cannot be proven rigidly. It rests on the observable equal distribution of the chromosomes, by means of mitosis, to all cells of an embryo and, therefore, presumably on the equal dis- tribution of the constituent genes in the chro- mosomes. It rests, furthermore, on experi- ments on regeneration, particularly in plants, which show that at least some differentiated cells still contain all potentialities to form a whole organism with all types of differen- tiated tissues and organs. These experiments are supplemented by others, in which cleav- age nuclei of early developmental stages have been shown to be totipotent (review in Spemann, °38). None of these results can be taken as actual evidence for an unchanged genic content of differentiated cells. While they seem to disprove schemes involving a kind of mechanical sorting out of genic ele- ments in development which would give one cell one part of the initial total gene assort- ment and another cell a different part, they do not exclude a hypothesis according to which the “same” gene in different cells may possibly become modified in different specific ways. If such a hypothesis were en- tertained it would only be necessary to as- sume that “modifications” of genes are re- versible in cells which, in regeneration or similar experiments, can give rise to new dif- ferentiations. In this sense the hypothetical gene modifications would be basically differ- ent from the gene changes brought about by mutations whose very nature is their perma- nent reproducibility. As will be shown be- low, the hypothesis of genic modification is not required by the facts. It is, however, not inconsistent with them. Quantitative differences in the genic con- tent of different cells and tissues have been well established in recent years. The process of endomitosis, by means of which the chro- mosomes of a nucleus multiply without nu- clear division, leads to the presence of poly- ploid constitutions characteristic for various tissues of diverse organisms (for a review and discussion see Huskins, *47). Undoubtedly this polyploidy is part of the differentiation of the tissues concerned but one wonders whether it is more than a concomitant of this differentiation. Furthermore, there are many more differentiations than states of ploidy. The concept which reconciles the apparent contradiction between essential genetic equal- ity of all cells and their differentiation from one another rests on the existence of regional cytoplasmatic differences in the egg cell and its cellular derivatives. The concept implies a differential response of the genes which during cleavage come to lie in different cyto- plasmic surroundings. The relatively few re- gional differences of the early egg are as- sumed to become increased when cleavage produces new topographic diversity which may lead to physiologic diversity. Each new type of regional diversity sets up a new sys- tem of gene-substrate interaction which ex- presses itself in developmental differentia- tion. For a more detailed discussion of the concept, based on a historical analysis and on new experiments, see Stern (754). The assumed differential response of the genes to the regional extragenic, primarily extranuclear, differences may be thought of in two different but not really mutually ex- clusive ways. On the one hand it is possible that different types of cytoplasm call forth the action of different genes, leaving other genes inactive. If in a region 1 genes A, B, and C find suitable substrates or a suitable milieu, while genes D, E, and F are not pro- vided with suitable conditions for their ac- tion, a differential activity would result in contrast to the situation in a region 2 where A, D, and E might find suitable conditions but not B, C, and F. On the other hand, it is possible that all genes, A, B, C, D, E, F, etc., are active in the cells of all regions but that their reaction products vary quantitatively, as a result of differential kinetic properties in different cells, or qualitatively, by depend- ing on different substrates. This latter as- sumption is compatible with high specificity of genic action if it is kept in mind that such specificity does not need to be restricted to a single substrate but may include a whole class of substrates. Such a scheme of differential genic effects in different parts of a developing system is not only a logically consistent derivative from knowledge on the variability of gene effects as dependent on variations in the genetic background and in the environment. Some phenomena are also known where cellular properties can be changed without genic alterations, in a way which may be regarded as a model for developmental dif- ferentiation. Experiments on asexually and GENE ACTION sexually reproducing microorganisms have yielded changed phenotypes whose under- lying mechanisms proved to be transmissible even though no changes in the genotype had occurred. One of these phenomena is enzy- matic adaptation as studied in yeasts and other microorganisms. When a population of these organisms is placed in contact with some substrate to which it does not respond typically on account of (absolute or relative) lack of suitable enzymes, it may gradually acquire the necessary enzymes. Once having acquired them, and as long as the specific substrates are present, the enzymes are re- produced and descendents of the enzymati- cally adapted cells continue to possess high enzyme activity. However, after removal of the specific substrates, the cells soon lose their adaptive enzymes. The ability of cells for enzymatic adaptation is controlled by genes (Winge and Laustsen, °39; Linde- eren, Spiegelman, and Lindegren, ’44). These genes, then, do not control in an absolute sense the presence of specific enzymes in the cell but the potentiality of enzyme forma- tion. Monod (742) and Spiegelman and Dunn (47) have shown that the formation of one enzyme frequently occurs in competition with formation of another enzyme so that enzymatic adaptation may involve not only the rise of one type of enzyme but the de- cline in quantity of others. In analogy with these findings, Spiegelman (’48) views dif- ferentiation ‘“‘as the controlled production of unique enzymatic patterns.” Parallel studies on Paramecium aurelia are concerned with antigenic properties (Son- neborn, *47). It has been found that various strains all contain four different antigens, designated as 1, 2, 3, and 4, but that in any one strain only one antigen, the “primary,” either 1 or 2 or 3 or 4, produces a very high titer of antibody in homologous antiserum and that only the primary antigen is con- cerned in the immobilization reaction when the animals are subjected to antiserum. There exist, then, four different types of strains as characterized by their primary antigen. Each strain is highly stable, breeding true during asexual fission to its specific antigenic pat- tern of a given primary and the correspond- ing three secondary antigens. It is, how- ever, relatively easy to induce changes from one type to another: by exchange of cyto- plasm between conjugants derived from dif- ferent types, or, in type 3 animals, by the mere fact of conjugation with type 1 or 2 animals, or by exposure to antisera, or after x-raying. Sometimes the changes go spe- 165 cifically from one type to another given type, sometimes two or more diverse antigenic strains may be isolated from the asexually produced progeny of a single altered indi- vidual which apparently exists in a tempo- rarily unstable state. Genetic analyses in- volving crosses or autogamy show clearly that the changes in antigenic type are purely cytoplasmatic, taking place within the frame- work of an unchanged nuclear genotype. Similarly as in enzymatic adaptation of mi- croorganisms, a differentiation of antigenic systems, perhaps depending on the establish- ment of one or the other of several alternative equilibria, can serve as a model for develop- mental differentiation of multicellular or- ganisms. The mechanisms by means of which the nuclear genes control the potentiality of enzyme formation or of antigenic types are still obscure. Hypotheses have been proposed which assume the production by the nuclear genes of partial replicas of themselves which reach the cytoplasm and are endowed with the genic property of reproducing themselves or the cytoplasmic complexes of which they become part. Proofs for the existence of “nlasmagenes” of this type are not available at present (Schultz, *50). Weiss (review in 50) has outlined a concept which involves interaction of self-perpetuating cytoplasmatic elements with primordial gene products. “Ac- cording to this concept, differentiated proto- plasmic units would owe their origin and their specific shapes to two entirely different processes, occurring in different places. They would be propagated in the nuclear center, and be remodeled in the cytoplasm.” An interpretation of gene action in de- velopment when transcending the level of intracellular states may be fitted into the concept of embryonic fields. Theoretically, specific alleles may either change the extent or type of a field or they may change the reactivity of cells to an existing unchanged field. These alternatives may be illustrated by the example of the sex-comb in Dro- sophila melanogaster. It consists of a row of heavy spines located on the first tarsal seg- ment of the foreleg of males (Fig. 444). The sex-comb is absent in females. The different developmental fate of the foreleg could be due either to the fact that the male genotype of an anterior leg disc leads to the existence of a male type field which calls forth sex- comb formation in the appropriate region, in contrast to the female genotype which does not lead to a sex-comb—inducing field; or to the presence of an appropriate field 166 THE NUCLEUS AND CYTOPLASM IN DEVELOPMENT in both male and female anlagen but a response by formation of a sex-comb only of cells with the male genotype. An ex- perimental embryologist would attack the problem by transplantation of male or fe- male embryonic tissue into the prospective sex-comb region of female or male embry- onic discs. The geneticists may accomplish equivalent results by studying gynandro- morphs in which, early in development, genetically male patches of tissue of varying Fig. 44. First tarsal joint of Drosophila melano- gaster. A, Normal male; distal (upper right) sex- comb consisting of a row of ten teeth. B, Gynander; sex-comb consisting of four teeth, interrupted by a bristle of female genotype. C, Gynander; sex-comb consisting of two teeth located in a region normally not occupied by sex-comb structures. (After Stern and Hannah, ’50.) extent may originate in the appropriate re- gions of otherwise genetically female tissues (Stern and Hannah, *50). The observations indicate that male and female genotypes alike lead to the emergence of a sex-comb field but that the response of the cells in the region of the prospective cells depends on their own genotype, male cells forming teeth of the sex-comb even if surrounded by fe- male tissue and female cells being unable to form teeth even if in the appropriate region of a preponderantly male disc (Fig. 44B, C). In a secondary fashion, however, the dif- ferent genotypes affect the field itself. When the cells of the typical sex-comb region are female in constitution and therefore unable to form a sex-comb, cells of male constitu- tion near by may differentiate into teeth of a sex-comb even though they are in a region which normally lies outside of the sex-comb region. It seems that the development of the sex-comb in the typical region modifies the surroundings so that no differentiation into further teeth occurs. Absence of tooth forma- tion in the typical region results in a modi- fication of the extent or intensity of the out- lying regions of the field such that male cells can respond to it which would not have done otherwise. This analysis of the origin of the sex-comb may serve as an illustration for many other cases where specific genotypes change the course of development both by varying the response of localized regions and by remolding more general developmental configurations. REFERENCES Balkaschina, E.I. 1929 Ein Fall der ErbhomGosis (die Genovariation “‘Aristopedia’’) bei Drosophila melanogaster. Roux’ Arch. Entw.-mech., 175: 448-463. Bateson, VW. 1894 Materials for the Study of Evolution. The Macmillan Co., London. Beadle, G. W. 1937 Development of eye colors in Drosophila: fat bodies and malpighian tubes in relation to diffusible substances. Genetics, 22: 587-611. 1945 Biochemical genetics. Chem. Rey., 37:15-96. latimenh. ls and «Clancys |G) VWWel 9359 Development of eye colors in Drosophila: pro- duction of v* hormone by fat bodies. Biol. Bull., 78:407-414, Bhat, M. R. 1949 A dominant mutant mosaic house mouse. Heredity, 3:243-248. Bonner, David M. 1948 Genes as determiners of cellular biochemistry. Science, 108:735-739. Bonnevie, K. 1934 Embryological analysis of gene manifestation in Little and Bageg’s abnormal mouse tribe. J. Exp. Zool., 67:443-520. Boveri, Th. 1902 Uher mehrpolige Mitosen als Mittel zur Analyse des Zellkerns. Verh. phys. med. Ges. Wiirzburg, N. F. 35:67-90. 1907 Zellenstudien. VI. Die Entwicklung dispermer Seeigeleier. Ein Beitrag zur Befruch- tungslehre und zur Theorie des Kernes. G. Fischer, Jena. Boycott, A. E., Diver, C., Garstang, S. L., and Turner, F. M. 1930 The inheritance of sin- istrality in Limnaea peregra (Mollusca, Pul- monata). Trans. Roy. Soc. London Phil., B279: 51-131. Brehme, K. 1941 Development of the Minute phenotype in Drosophila melanogaster: a com- parative study of the growth of three Minute mu- tations. J. Exp. Zool., 88:135—160. Breitenbecher, J. K. 1925 The inheritance of sex- limited bilateral asymmetry in Bruchus. Genetics, 10:261-277. Bridges, C. B. 1922 The origin of variations in sexual and sex-limited characters. Amer. Nat., 56-51-63. , and Brehme, K. 1944 The mutants of Drosophila melanogaster. Carnegie Inst. Wash- ington Publ. No. 522. GENE ACTION Bridges, C. B., and Dobzhansky, Th. 1933 The mutant “proboscipedia” in Drosophila melanogas- ter—a case of hereditary homodsis. Roux’ Arch. Entw.-mech., 727:575—590. Caspari, E. 1933 Uber die Wirkung eines pleio- tropen Gens bei der Mehlmotte Ephestia kiihni- ella Zeller. Roux’ Arch. Entw.-mech., 730:353- 381. Chevais, S. 1943 Déterminisme de la taille de Voeil chez le mutant Bar de la Drosophile. Bull. Biol., 78:1-108. Dahlberg, G. 1943 Genotypic asymmetries. Proc. Roy. Soc. Edin., 63:20-31. David, P. R. 1936 Studies on the Creeper fowl. X. A study of the mode of action of a lethal factor by explantation methods. Roux’ Arch. Entw.- mech., 735:521-551. Demerec, M. 1934 Biological action of small de- ficiencies of X-chromosome of Drosophila mel- anogaster. Proc. Nat. Acad. Sci., 20:354-359. 1936 Frequency of “cell-lethals’” among lethals obtained at random in the X-chromosome of Drosophila melanogaster. Proc. Nat. Acad. Sci., 22:350-354:. Dobzhansky, Th. 1927 Studies on the manifold effect of certain genes in Drosophila melano- gaster. Z. I. A. V., 43:330-388. 1935 Maternal effect as a cause of the differences between the reciprocal crosses in Drosophila pseudoobscura. Proc. Nat. Acad. Sci., 21:443-446. ———.,, and Holz, A. M. 1943 A re-examina- tion of the problem of manifold effects of genes in Drosophila melanogaster. Genetics, 28:295- 303. Dunn, L. C. 1949 Some relations between muta- tions and abnormal development; in 20th An- niversary Lectures, Jackson Memorial Labora- tory, Bar Harbor, Maine. Emerson, S. 1945 Genetics as a tool for studying gene structure. Ann. Missouri Bot. Garden, 32: 243-249. 1950 Competitive reactions and antago- nisms in the biosynthesis of amino acids by Neurospora. Cold Spring Harbor Symp. Quant. Biol., 14:40-48. Ephrussi, B. 1935 The behavior in vitro of tis- sues from lethal embryos. J. Exp. Zool., 70:197- 204: Fogh-Andersen, Paul 1943 Inheritance of hare- lip and cleft palate. Opera ex Domo biol. hum. hered., 4:266 pp. Ford, E. B., and Huxley, J.S. 1929 Genetic rate- factors in Gammarus. Roux’ Arch. Entw.-mech., 117:67-79. Goldschmidt, R. B. 1920 Die quantitativen Grundlagen, von Vererbung und Artbildung. J. Springer, Berlin. 1927 Physiologische Theorie der Verer- bung. J. Springer, Berlin. 1931 Die sexuellen Zwischenstufen. J. Springer, Berlin. 1934 Lymantria. Bibliogr. genet., 7/7: 1-186 1935a Gen und Ausseneigenschaft, I-II. Z. 1. A. V., 69:38-131. 167 — 1935b Gen und Aussencharakter, III. Biol. Zentralbl., 55:535-554. 1937 Gene and character, IV—VIII. Univ. California Publ. Zool., 47:277-333. 1938 Physiological Genetics. McGraw- Hill Book Co., New York. Gregory, P. W., and Castle, W. E. 1931 Further studies on the embryological basis of size inher- itance in rabbits. J. Exp. Zool., 59:199-211. Griineberg, H. 1938 An analysis of the “pleio- tropic” effects of a new lethal mutation in the rat (Mus norvegicus). Proc. Roy. Soc. London, B125: 123-144. 1943 The Genetics of the Mouse. Cam- bridge University Press, Cambridge, England. Hadorn, E. 1948 Gene action in growth and dif- ferentiation of lethal mutants of Drosophila Symp. Soc. Exp. Biol., No. II, Growth, pp. 177- 195. Haldane, J. B. S. 1937 The biochemistry of the individual; im Perspectives in Biochemistry, edited by J. Needham and D. E. Green. Cam- pridge University Press, Cambridge, England, pp. 1-10. Hammerling, J. 1934 Uber formbildende Sub- stanzen bei Acetabularia mediterranea, ihre raumliche und zeitliche Verteilung und ihre Her- kunft. Roux’ Arch. Entw.-mech. 737:1-81. 1946 Neue Untersuchungen iiber die physiologischen und genetischen Grundlagen der Formbildung. Naturw., 33:337-342, 361- 365. Harvey, E. 1940 A comparison of the develop- ment of nucleate and non-nucleate eggs of Ar- bacia punctulata. Biol. Bull., 79:167-187. Henke, K. E., Finck, E.v., and Ma, S. Y. 1941 Uber sensible Perioden fiir die Auslésung von Hitze-Modifikationen bei Drosophila und die Beziehungen zwischen Modifikationen und Muta- tionen. Z. I. A. V., 79:267-316. , and Maas, H. 1946 Uber sensible Peri- oden der allgemeinen K6rpergliederung yon Drosophila. Nachr. Akad. Wiss. Gottingen Math. Physik. Klasse, 7:3-8. Herbst, C. 1896-1901 Uber die Regeneration von antennenahnlichen Organen an Stelle von Augen. I. Roux’ Arch. Entw.-mech., 2:544-558. II. Vier- teljahrsschrft. naturf. Ges. Ziirich, 41:435-454. III-V. Roux’ Arch. Entw.-mech., 9:215-292; 13: 436-447. Herskowitz, Irwin 1949 Hexaptera, a homoeotic mutant in Drosophila melanogaster. Genetics, 34: 10-25. Horowitz, N. H. 1950 Biochemical genetics of Neurospora. Advances in Genetics, 3:33-71. Huskins, C. L. 1947 The subdivision of the chro- mosomes and their multiplication in non-dividing tissues: possible interpretations in terms of gene structure and gene action. Amer. Nat., 87:401- 434, Irwin, M.R. 1947 Immunogenetics. Advances in Genetics, 7:133-160. Jollos, V., and Peterfi, T. 1923 Furchung von Axolotleiern ohne Beteiligung des Kernes. Biol. Zentralbl., 43:286-288. Kiihn, A., Caspari, E., and Plagge, E. 1935 Uber 168 THE NUCLEUS AND CYTOPLASM IN DEVELOPMENT hormonale Gen-Wirkungen bei Ephestia kiihniel- la Z. Nachr. Ges. Wiss. Gottingen, 2:1—30. , and Henke, K. 1929-36 Genetische und entwicklungsphysiologische Untersuchungen an der Mehlmotte Ephestia kithniella Zeller, I-XIV. Abh. Wiss. Gottingen, N.F. 75:1-121, 127-219, 995-972. Landauer, W. 1946 Genetic aspects of physiol- ogy; in Howell’s Textbook of Physiology, edited by J. F. Fulton. 15th ed. W. B. Saunders Co., Philadelphia, pp. 1232-1247. 1948 Hereditary abnormalities and their chemically induced phenocopies. Growth, 12 (Suppl.) :171-200. Li, J.C. 1927 The effect of chromosome aberra- tions on development in Drosophila melanogaster. Genetics, 72:1-58. Lindegren, C. C., Spiegelman, S., and Lindegren, G. 1944 Mendelian inheritance of adaptive en- zymes in yeast. Proc. Nat. Acad. Sci., 30:346-352. Ludwig, W. 1932 Das Rechts-Links Problem im Tierreich und beim Menschen. J. Springer, Ber- lin. Mitchell, H. K., and Houlahan, M. B. 1946 Adenine-requiring mutants of Neurospora crassa. Federation Proc., 3:370-375. Mohr, O. L. 1932 On the potency of mutant genes and wild-type allelomorphs. Proc. 6th Int. Cong. Genet., 7:190-212. Monod, J. 1942 Recherches sur la croissance des cultures bactériennes. Actualités Scientifiques et Industrielles, No. 911. Hermann and Co., Paris. Moore, A. R. 1943 Maternal and paternal inher- itance in the plutei of hybrids of the sea urchins Strongylocentrotus purpuratus and Strongylocen- trotus franciscanus. J. Exp. Zool., 94:211-228. Morgan, T. H., Bridges, C. B., and Sturtevant, A. H. 1925 The genetics of Drosophila. Bibliogr. genet., 2:1-262. , Sturtevant, A. H., and Bridges, C.B. 1929 The constitution of the germinal material in re- lation to heredity. Carnegie Inst. Year Book, 28: 338-345. Nelson, J. M., and Dawson, C. R. 1944 Tyro- sinase. Advances in Enzymology, 4:99-152. Plunkett, C. R. 1926 The interaction of genetic and environmental factors in development. J. Exp. Zool., 46:181-244. Poulson, D. F. 1940 The effects of certain X- chromosome deficiencies on the embryonic de- velopment of Drosophila melanogaster. J. Exp. Zool., 83:271-325. 1945 Chromosomal control of embryo- genesis in Drosophila. Amer. Nat., 79:340-363. Przibram, H. 1910 Die Homoeosis bei Arthro- poden. Roux’ Arch. Entw.-mech., 29:587-615. Robertson, G. G. 1942 An analysis of the devel- opment of homozygous yellow mouse embryos. J. Exp. Zool., 89:197—231. Russell, E. S. 1949 Analysis of pleiotropism of the W-locus in the mouse: relationship between the effects of W and W” substitution on hair pig- mentation and on erythrocytes. Genetics, 34:708- 723. Russell, L. B., and Russell, W. L. 1948 The pro- duction of phenocopies in the mouse by means of x-ray treatment of embryos. Genetics, 33:627. Schultz, J. 1950 The question of plasmagenes. Science, 777:403-407. Schwab, J. J. 1940 A study of the effects of a random group of genes on shape of spermatheca in Drosophila melanogaster. Genetics, 25:157- vate Sinnott, E. W., Houghtaling, H., and Blakeslee, A. F. 1934 The comparative anatomy of extra- chromosomal types in Datura stramonium. Car- negie Inst. Washington Publ. No. 431. Smith, P. E., and MacDowell, E. C. 1930 An hereditary anterior-pituitary deficiency in the mouse. Anat. Rec., 46:249-257. Sonneborn, T. M. 1947 Developmental mecha- nisms in Paramecium. Growth Symp., 77:291- 307. Spemann, H. 1938 Embryonic Development and Induction. Yale University Press, New Haven, Connecticut. Spiegelman, S. 1948 Differentiation as the con- trolled production of unique enzymatic patterns. Symp. Soc. Exp. Biol., No. II: Growth, pp. 286- 325. ,and Dunn, R. 1947 Interactions between enzyme-forming systems during adaptation. J. Gen. Physiol., 37:153-173. Standfuss, M. 1914 Mitteilungen zur Verer- bungsfrage. Mitt. schweiz. entomol. Ges., 72:238- 252: Stern, C. 1940 Recent work on the relation be- tween genes and developmental processes. Growth Supplement, pp. 19-36. 1941 The growth of testes in Drosophila. I. The relation between vas deferens and testis within various species. II. The nature of inter- specific differences. J. Exp. Zool., 87:113-158, 159-180. 1948 The effects of changes in quantity, combination, and position of genes. Science, 108: 615-621. 1954 Two or three bristles. Amer. Sci- entist, 42:212-247. , and Hannah, A. M. 1950 The sex-combs in gynanders of Drosophila melanogaster. Portug. Acta Biol., Ser. A; R. Goldschmidt Volumen, pp. 798-812. , and Schaeffer, E. W. 1943 On wild-type iso-alleles in Drosophila melanogaster. Proc. Nat. Acad. Sci., 29:361-367. Sturtevant, A. H. 1923 Inheritance of direction of coiling in Limnaea. Science, 58:263—270. 1932 The use of mosaics in the study of the developmental effects of genes. Proc. 6th In- ternat. Congr. Genetics, 7:304—307. Villee, C. A. 1942 The phenomenon of homoeo- sis. Amer. Nat., 76:494-506. Vogt, M. 1946 Zur labilen Determination der Imaginalscheiben von Drosophila, II. Die Um- wandlung présumptiven Fiihlergewebes in Bein- gewebe. Biol. Zbl., 65:238-254. Weiss, P. 1950 Perspectives in the field of mor- phogenesis. Quart. Rev. Biol., 25:177-198. Wettstein, F. v. 1940 Experimentelle Unter- GENE ACTION suchungen zum Artbildungsproblem, II. Zur Frage der Polyploidie als Artbildungsfaktor. Ber. Deutsch. Bot. Ges., 58:374-388. Whiting, A. 1934 Eye colors in the parasitic wasp Habrobracon and their behavior in multiple recessives and in mosaics. J. Genet., 29:99-107. Whiting, P. W. 1940 Multiple alleles in sex de- termination of Habrobracon. J. Morph., 66:323- 355. , Greb, J., and Speicher, B. R. 1934 A new type of sex intergrade. Biol. Bull., 66:152-165. Winge, O., and Laustsen, O. 1939 On 14 new yeast types, produced by hybridization. Compt. rend. Trav. Lab. Carlsberg, Serie Physiol., 22: 337-351. 169 Wright, S. 1916 An intensive study of the in- heritance of color and of other coat characters in guinea pigs with especial reference to graded variation. Carnegie Inst. Washington Publ. No. 241, pp. 59-160. 1934 Physiological and evolutionary the- ories of dominance. Amer. Nat., 68:24-53. 1941 The physiology of the gene. Phys- iol. Rev., 27:487-527. 1945a Physiological aspects of genetics. Ann. Rev. Physiol., 7:75-106. 1945b Genes as physiological agents: general considerations. Amer. Nat., 79:289-303. Section V EMBRYOGENESIS: PREPARATORY PHASES CHAPTER Gametogenesis, Fertilization and Parthenogenesis ALBERT TYLER GAMETOGENESIS ORIGIN OF THE GERM CELLS AN UNDERSTANDING of the factors that endow egg and spermatozoon with the ability to unite and produce a new individual may be expected to depend largely on knowledge of the manner of origin and formation of the gametes. While such knowledge is very far from complete there is a considerable body of descriptive and experimental information re- lating to it. To illustrate the nature of this information and some of the problems in- volved certain of the pertinent investigations will be briefly reviewed here. Questions con- cerning the determination of the gonad are discussed in another chapter of this book. Much of the stimulus for work on the origin of the germ cells was undoubtedly provided by Weismann’s (1883, 1893) views concerning the continuity and segregation of the germ plasm. His concept of a distinct germ plasm is sufficiently familiar, even to beginning biologists, so that it does not re- quire detailed presentation here. A critical review of the original basis of the concept has been published recently by Berrill and Liu (48). In many species of animals the primordial germ cells may be recognized at an early stage of development and, particularly among the invertebrates, there are cases in which a “germ-line” is manifest from the start of cleavage (see Wilson, ’25, and Boun- oure, 39, for references and brief accounts). In these the identification of the primordial germ cells is based on certain distinguishing 170 features of the nucleus or the cytoplasm. We shall consider some examples briefly here. Chromatin Diminution in Ascaris. The clas- sic case is that of Ascaris megalocephala (Parascaris equorum) in which Boveri (1887- 1910) described a process, termed chromatin diminution (Herla, 1895), occurring during cleavage in the cells that are to form somatic tissue but not in those that are to form the germ cells. In this process the mid-portion of each of the chromosomes (two in A. megalocephala univalens) breaks up into a number of small fragments (about 10 per chromosome according to Fogg, ’30), leaving two large terminal pieces. The smaller frag- ments reconstitute the daughter nuclei and continue regular mitotic division but the terminal pieces remain in the cytoplasm and slowly disappear. The ability of the small fragments to continue mitotic division is evi- dently due to the fact that each possesses a centromere, the original chromosomes being polycentric, as Schrader (’35) and White (36) have pointed out. It is now well known that chromosomal fragments devoid of a centromere fail to attach to the spindle and to divide. The diminution phenomenon starts at the second or the third division of the egg. In the former case it occurs in the dorsal, so- matic cell (Si) and not in the ventral, stem cell (Pi), as illustrated in Figure 45. In the latter case it occurs in three cells at once but not in the stem cell Pe. In both cases Pz divides into a somatic cell which under- goes diminution (at its next division) and another stem cell, P3. The latter again di- GAMETOGENESIS, FERTILIZATION AND PARTHENOGENESIS Lgl vides (without diminution) into a somatic cell (that undergoes diminution at its next division) and the definitive primordial germ cellisea veg (Boveri, 10a; Hogue, *10). The centrifugation experiments provided the more direct evi- dence. In many of the centrifuged eggs (as high as 36 per cent in some experiments) it Fig. 45. Cleavage of Ascaris megalocephala (Parascaris equorum), illustrating location of the stem cells (Py—-P4) which do not undergo chromatin diminution and from which the germ cells are derived. Note bend- ing of the criginal egg axis as a result of the shifting of the blastomeres in the 4-cell stage. A, 2-cell stage; B, C, D, 4-cell stage; E, 7-cell stage; F, 24-cell stage; an., animal pole; veg., vegetal pole. (From von Ubisch, °43, after Boveri, ’99.) Fig. 46. Second (A) and third (B and C) cleavages of Ascaris eggs that had been centrifuged before the first cleavage; showing (in B and C) chromatin diminution in the upper cells but not in the two lower cells (from Hogue, 710). Boveri surmised that differences in the cytoplasm of different regions of the egg were responsible for the occurrence or non- occurrence of the diminution process and, at the same time, for the determination of the germ cells. This view was supported by the results of studies on polyspermic eggs (Boveri, 04, °10b) and on centrifuged eggs was found that two stem cells formed instead of the single one that normally occurs. Fig- ure 46 illustrates such eggs at the second and third cleavages. Diminution is seen in the upper two blastomeres (adjacent to an ex- truded ball of cytoplasmic material) but not in the two lower cells. This effect is inter- preted as being due to a shift in direction iz of the first cleavage plane of the centrifuged egg so that the first two blastomeres have essentially the same cytoplasmic constitution with regard to factors inducing or prevent- ing diminution. The next division, then, is differential with regard to these factors in each blastomere. More recently King and Beams (’38) have subjected Ascaris eggs to high speed (150,000 g.) centrifugation whereby cytoplasmic di- vision was suppressed while nuclear division continued. They noted that, in such eggs, diminution usually occurs in all of the nuclei AB 8/16 (ECTODERM) (ECTODERM) (MESODERM) (ENTODERM) (PRIM. GERM CELL) (ECTODERM) 9/16 10/16 W/16 +12/16 13/16 14/16 15/16 + 16/16 (ECTODERM) Fig. 47. Diagram illustrating location, in the un- cleaved Ascaris egg, of the material for the primor- dial germ cell (P4) (after von Ubisch, *43). at the second or the third mitosis. This led them to interpret diminution on the basis of a “diminisher” substance D, that is produced from some cytoplasmic constituent, An, which is concentrated in the animal region of the egg and grades off to zero at the vegetal pole. Normally, D reaches sufficient concentration to cause diminution in the polar cell or cells at the second or third cleavage, and in the somatic cells derived from the stem cells at succeeding divisions. In the absence of cell boundaries, in the centrifuged egg, D is assumed to be free to diffuse and, upon reaching sufficient concen- tration, causes diminution to occur in all of the nuclei. This view has been criticized by von Ubisch (’43), who points out that the re- gion of the uncleaved egg that is to be in- corporated into the primordial germ cell (P4) does not comprise the most vegetal ma- terial but is located about halfway between the equator and the vegetal pole (see Fig. 47). This fact was evident in the original cell-lineage studies of Boveri (1899) but has been overlooked in most accounts of the EMBRYOGENESIS: PREPARATORY PHASES work on Ascaris. However, it seems to the present author that the essential feature of the interpretation of King and Beams re- mains valid. That feature is the localized production of some substance (either induc- ing or preventing diminution) during early cleavage. It seems simplest to regard this in terms of the production of materials essential for the continued reproduction of those parts (the ends and possibly the substance in the regions of fracture) of the chromosomes that are eliminated in the cells that undergo dim- inution. The site of production would be that region of the cytoplasm halfway be- tween the equator and the vegetal pole of the uncleaved egg, and the formation of cell boundaries would presumably prevent dif- fusion of the material to other cells. Chromosome Elimination in Sciara. Another remarkable example of differential behavior of the chromosomes in germ cells and somatic cells occurs in the fungus fly Sciara, studied extensively by C. W. Metz and his co-workers (see Metz, ’38, for review, and Berry, ’41, for some later details). The zygote starts devel- opment with three pairs of autosomes, three X-chromosomes and, in twelve of fourteen species examined, one, two or three large chromosomes called “limited” chromosomes (Fig. 48). At the sixth division (sometimes the fifth) of the zygote the “limited” chro- mosomes are eliminated from all of the so- matic nuclei. At this stage the nuclei have migrated from the middle of the egg, where the zygote nucleus is originally located, nearly to the periphery. One or two of the nuclei at the posterior pole of the egg form the primordial germ cells, which retain for some time the full complement of chromo- somes present in the zygote. Another elimination of chromosomes oc- curs in the somatic cells at the seventh or eighth division. At this time the somatic nuclei of the female-producing eggs elimi- nate one of the three X-chromosomes and the male-producing eggs eliminate two X-chromosomes. In these elimination divi- sions the “limited” chromosomes and X-chromosomes that are to be discarded fail to divide, or divide incompletely. They are, then, left on the middle of the spindle in anaphase and do not become incorporated in the daughter nuclei, but slowly break up and disappear in the later embryo. In Sciara an elimination also occurs later in the germ cells after they have migrated to the site of formation of the gonads, one of the X-chromosomes (one of the two pa- ternal X’s) being extruded from each nu- GAMETOGENESIS, FERTILIZATION AND PARTHENOGENESIS 173 cleus in both male and female larvae. This elimination is described (Berry, ’41) as oc- curring in a resting nucleus, the chromo- some being expelled through the nuclear membrane into the cytoplasm. Presumably there is also an elimination of one or more “limited” chromosomes from the germ cells ? Line ot et = az a” ! adii\\ 7 ni Egg Sad pe ULE [A] om Zz eeu Ist Elimination SL iar we o Zygote : x) (ex pee WZ Wy NV Se PN H i Mm Xx Sperm Somatic Ne NW AVL aa Fin ee a bud while the maternal chromosomes and the “limited” chromosomes all move to the pole and are incorporated in the single sec- ond spermatocyte. The latter forms a_ bi- polar spindle, and the sister chromatids of each chromosome except the X separate to opposite poles. Both chromatids of tne x Paternal dh iN : iy VI Te at WL eee ae mre Wh eh MW XX Ist Division d Xf Spermatogonia Selective meee \ Segregation eet YZ 2nd Elimination | “in Sperm if AV E aet@anossesc > = at x 3 Soma | Limited Chromosomes () Maternal x H Ave i Paternal x Tat Bale “4 a Autosomes xXx 2 Soma Fig. 48. Diagram of behavior of the chromosomes in Sciara coprophila, illustrating difference between so- matic and germ line cells in regard to elimination of chromosomes. The female germ line (not illustrated) would be the same as in diagram A except that the meiotic divisions do not involve the cytological pecu- larities exhibited in spermatogenesis. See text for further description. (From Metz, ’38.) in some cases, but this has not, as yet, been described. Since the germ cells of Sciara have the same chromosome set in the male as in the female the determination of whether they are to produce spermatozoa or eggs is evi- dently effected by the somatic cells which, as a result of the elimination of one or two X’s at the seventh (or eighth) division, have the typical XO constitution in the male and XX in the female. This, in turn, is evidently determined in the female parent which, as a rule, produces unisexual families. The genetic evidence accumulated by Metz and his co-workers shows that the two X-chromo- somes are alike (XX) in the male-producers but differ genetically (XX’) in the female producers. The results of fertilization in these two types are diagrammed in Figure 49. The unusual chromosome constitution of the zygote in Sciara results from peculiari- ties in spermatogenesis. In the first sperma- tocyte a monocentric spindle is formed and the chromosomes are separated without pair- ing, in such a way that all the paternal chromosomes except the “limited” chromo- somes are extruded in a small cytoplasmic go to one pole and are incorporated in the single spermatid that forms, while the chromosomes at the other pole are included in a small bud that later degenerates. All of the spermatozoa that form are, then, alike Female Male Female ( %-producing) (¢-producing) Germ line: X'X XX XX Gametes: Eggs Sperms Eggs ' x XX x Germ line: X' X x X X Female Male (¢-producing) Female ( %-producing) Fig. 49. Metz’s interpretation of the production of “unisexual” families in Sciara coprophila on the basis of genetically different (X and X’) sex chro- mosomes (from Metz, ’38). and contain a haploid set of autosomes, the “limited” chromosomes and two X-chromo- somes. Oogenesis, on the other hand, is regu- lar with random segregation of chromo- somes. Chromosome Elimination in Other Animals. Even more extensive elimination of chromo- somes from somatic cells has been described 174 in the gall midges (Cecidomyidae), which have been studied most recently by White (46, ’47a, b, 48). In Miastor, for example, there are 48 chromosomes in the zygote but 36 of these are eliminated from the so- matic nuclei of the female and of the pedo- genetic larvae at the third and fourth cleav- age divisions, while 42 are eliminated in the males. The full complement is retained by the germ cells, which are set aside at the posterior pole of the egg at this time. Chromosome- or chromatin-diminution also occurs in various animals without rela- iii ltt l il Mill il EMBRYOGENESIS: PREPARATORY PHASES of the germ cells, have been described in several species of animals among the scyph- ozoans (Equorea), chaetognaths (Sagitta), rotifers (Asplanchna), insects (Chironomus, Calliphora, Calligrapha, Lepintotarsa, Copi- dosoma, Trichograma, Apanteles), crustace- ans (Cyclops, Diaptomus, Polyphemus, Moina) and amphibians (Rana) (see Heg- ner, °14; Wilson, ’25; Bounoure, ’39). Ac- cording to the accounts the cells that receive these granules ultimately become germ cells. However, there is, as yet, no direct experi- mental evidence, such as might be derived Fig. 50. Geigy’s method for exposing the posterior pole of the egg of Drosophila to ultraviolet radiation. C, wax trough; U, ultraviolet beam; S, brass screen; E, egg; P, pole cells. (After Geigy, ’31.) tion to germ-cell determination (see Berry, 41, for references). For example, it occurs in the polar body divisions of several species of Lepidoptera. Here, however, the elimi- nated chromatin is Feulgen-negative whereas in Ascaris, Sciara, etc., it is Feulgen-positive (Bauer, 732, °33). One of the most striking examples has been described in the grass mite, Pediculopsis graminum (Cooper, ’39, 41). Feulgen-negative “chromatin” bodies that correspond in number and position to the metaphase chromosomes are left at the equator of the spindle at both polar body divisions and at each of the first nine cleav- age divisions, but not in the succeeding mitoses of the embryo or of the oogonia. Such cases do not, however, rule out the possibility that. where elimination occurs in connection with germ-cell determination, the chromatin that is eliminated from the somatic cells may contain genes that are of importance for the development of the gametes. Localized Cytoplasmic Factors in Germ-Cell Determination. Special cytoplasmic granules (termed germ-cell determinants), that are considered to be important for the formation from centrifugation experiments, to show that the particular granules are causative agents in germ-cell determinations. Apart from the questions of particular granules and of chromatin diminution, the importance of localized cytoplasmic factors in germ-cell determination is very well illustrated in the experiments of Geigy (31), Aboim (45), and Geigy and Aboim (44) on Drosophila. Geigy (731) succeeded in inhibiting the formation of germ cells by subjecting the posterior pole of the egg to ultraviolet irradiation. The method of local- ized irradiation is shown in Figure 50. After irradiation the cleavage nuclei that wander into the damaged pole plasm degenerate and the formation of pole cells is partially or completely suppressed. However, even in cases of complete suppression, adult flies are obtained which appear completely nor- mal and have gonads of normal structure although of reduced size. This work not only strikingly confirms the earlier indica- tions that factors essential for germ-cell formation are localized in the posterior cyto- plasm of the egg, but also furnishes further convincing evidence that the mesodermal GAMETOGENESIS, FERTILIZATION AND PARTHENOGENESIS 75 parts of the gonad can develop normally in the absence of the germ cells. Somewhat similar experiments have been performed in frogs by Bounoure (35a, b, 37a, b, c). He irradiated the vegetal pole of the egg and obtained a great reduction in the number of germ cells of the larvae that developed. While none of the larvae were completely agametic, at least two cases were obtained in which one gonad was en- tirely devoid of germ cells. Here, too, the \s \ il A upon the occurrence of a migration of primordial germ cells to the genital ridge from some other region of the embryo or from extra-embryonic areas, such as de- scribed by Allen (’07) in frogs and by Swift (14) in birds. For a recent example of such studies reference may be made to the work of Witschi (48) on human em- bryos, in which migration of the germ cells from the yolk sac to the genital folds is convincingly described. B Fig. 51. Maps showing distribution of germ cells in chick blastoderms. A, head-process stage; B, 3-somite stage. (From Willier, ’37.) somatic constituents of the agametic gonad appeared normal. Other Experiments Concerning the Origin of the Germ Cells in Vertebrates. The ques- tion of the origin of the germ cells in vertebrates has been the subject of much controversy centering primarily upon whether or not they may be formed from “differentiated” somatic cells in various stages of embryonic development and in the adult. The subject has been frequently and extensively reviewed (Willier, 39; Bounoure, 39; Dantschakoff, ’41; Everett, ’45; Nieuw- koop, ’46). The present discussion will re- strict itself mainly to some of the more recent experimental work. There is now fairly general agreement In many respects birds have provided the more favorable experimental material in attempts to deprive the embryo of germ cells, and several workers have reported success upon subjecting the area of Swift’s germinal crescent (see Fig. 51) to extirpation (Reagan, 16; Willier, ’33, ’37), ultraviolet irradiation (Benoit, °30), and cauterization (Dantscha- Koft yet sal’, 23), a Whlllier’s, (@37)) wexperi- ments on chorio-allantoic grafting of por- tions of chick blastoderms, it has been demonstrated that a gonad free of germ cells may develop and claims (e.g., Dantschakoff et al., 31) of the necessity of the presence of germ cells for differentiation of the gonad are refuted. Willier cautions against inter- preting the results of his experiments, and 176 those of others, as definitive proof of the ex- tragonadal origin of all of the germ cells in the chick. He points to three somewhat anomalous results: (1) Sterile gonads may be obtained from grafts that contain the germinal crescent as well as from those in which that area has been excluded. (2) Sterile gonads may also form from grafts of the gonad-forming area taken at a stage when it contains germ cells. (3) Germ cells Dorsal fs *, a PES Aho in * i EN oie EE, Ventral Fig. 52. Diagram illustrating the location of the presumptive primordial germ cells in urodeles at the early yolk plug stage according to experiments of Nieuwkoop (746). d.l., Dorsal lip of blastopore; P, area of presumptive lateral plate and nephrog- enous cord mesoderm containing the presumptive primordial germ cells. may appear in grafts of Hensen’s node taken at the head process stage. Willier suggests that results 1 and 2 may be at- tributed to lack of development of a germ- cell transporting mechanism (blood vascular system) in the graft and to lack of some condition in the graft that is essential for proper growth of the gonad. Result 3 may be interpreted on the basis of observations that show the presence of some primordial germ cells in regions posterior to the antero- lateral crescentic zone of the area pellucida that had previously been considered to be their locus. While the experiments are not presented as definitive proof of the extra- gonadal origin of the primordial germ cells in birds, they offer strong support for that view. Evidence that the mature sex cells are EMBRYOGENESIS: PREPARATORY PHASES derived from the primordial germ cells rather than secondary germ cells is provided by experiments on removal of the left ovary of young chicks. As is well known, the rudimentary right ovary then tends to develop into a testis. Normally the primor- dial germ cells of the rudimentary right gonad disappear after the third week. If removal of the left ovary is performed prior to this disappearance the right gonad may form a testis with mature spermatozoa, whereas upon later removal no spermatozoa are produced in the testis that develops (Domm, ’29). Anurans and urodeles apparently differ in regard to the site in which cells that have the appearance of primordial germ cells are first found. In the former it is the ento- derm of the gut wall; in the latter it is the lateral plate mesoderm. The results of experiments in these two groups seem, on the whole, to be consistent with the observed differences in location of such cells. Thus, in the anuran Discoglossus, Monroy (739) obtained germ cell-free embryos when the ventral part of the entoderm was removed from caudal halves of early neurulae, while removal of the dorsal entoderm did not alter the number and location of the germ cells. However, Monroy’s experiments do not ex- clude the interpretation, proposed by Nieuw- koop (746) for salamanders, that the ventral entoderm acts as an inductor on “predis- posed” primordial germ cells that originate elsewhere. In urodeles the most extensive experiments appear to be those of Humphrey (727, 728, 29, °33) and Nieuwkoop (746). By means of a large variety of extirpation and trans- plantation experiments Nieuwkoop has con- firmed Humphrey’s view that the germ cells originate in the presumptive lateral plate mesoderm which, in the gastrula, is repre- sented by the ventral and ventrolateral lips of the blastopore. In the uncleaved egg this material would be located equatorially op- posite the grey crescent rather than at the vegetal pole that Bounoure regards as its position in anurans. Nieuwkoop was able to distinguish germ cells of different species by differences in their content of pigment granules. So heteroplastic transplantation could be used to demonstrate that the pre- sumptive lateral plate mesoderm is the only source of germ cells (see Fig. 52). However, when the presumptive dorsocaudal entoderm is removed at the gastrula stage, or the whole entoderm at the neurula stage, there is a considerable reduction in the number of GAMETOGENESIS, FERTILIZATION AND PARTHENOGENESIS iF germ cells. These and related experiments are interpreted to mean that the dorsocaudal entoderm has a germ cell—inducing in- fluence on certain predisposed cells of the presumptive lateral plate mesoderm, with which it comes in contact at the end of gastrulation and during neurulation. On the basis of early determination and segregation this implies that the germ cells are at first simply endowed with the capacity (compe- tence) to react to such inductive stimulus. The inductive stimulus is apparently not pro- vided by other organs such as the notochord and Wolffian duct. That the reactive cells are not simply ordinary lateral plate cells is illustrated by the absence of germ cells in embryos in which the lateral plate was partially reformed (after removal of its presumptive material) by regulation from more dorsal mesoderm. Nieuwkoop also finds, as Humphrey had demonstrated ear- lier, that the germ cells do not cross over from one side of the embryo to the other. Concerning the formation of a second gen- eration of germ cells in the genital ridge, Nieuwkoop finds no indication of this in any of the experimental animals in which the presumptive germ-cell material was removed. This is very well shown, too, in his heteroplastic transplantations in which the host germ ridge may contain exclusively donor germ cells. So the possibility that the presence of primary primordial germ cells may induce genital ridge cells to form secondary germ cells is also ruled out. GROWTH OF THE OOCYTE Cytological Investigations. There has been a considerable amount of cytological investiga- tion into the manner of yolk formation, cen- tering primarily upon the possible role of various cytoplasmic and nuclear structures, such as the mitochondria, Golgi bodies, chromidia, nucleoli and hyaloplasm. Vari- ous workers have implicated one or another, or some combination, of these structures in the process and there is considerable differ- ence of opinion even when the same material is studied. (See, for example Harvey, ’29, for references to four investigations, and four different interpretations, of yolk forma- tion in the centipede Lithobius.) The earlier work has been reviewed by Wilson (25) and MacBride and Hewer (31). For some of the more recent work reference may be made to the studies of Subramaniam and Aiyar (36) on sea urchins, of Narain (737) and Singh and Boyle (738) on fish, of Beams and King (38) on the guinea pig, of Worley (’44, 46) on mussels and of Bretschneider (’46) on snails. No general conclusion concerning the mechanism of yolk formation appears to be, as yet, firmly established. The only point on which there seems to be fair agree- ment among various workers is that the fatty yolk is formed in association with Golgi bodies (see Worley, ’44; Bretschneider, ’46). Two techniques that hold promise of effec- tive use in further research in this field are centrifugation, employed by Beams and King (738) and Singh and Boyle (’38), and vital staining, employed by Worley (’44). Taste 9. Relative Radiophosphorus (P#*) Con- tent of the Phosphatide-Phosphorus Extracted from Various Organs of a Laying Hen 5 Hours after a Subcu- taneous Injection of 10 mg. of La- belled Sodium Phosphate (from Hevesy and Hahn, ’38) p®2 CONTENT PER MG. P RELATIVE ORGAN TO THE P® PER MG. INORGANIC P OF THE PLASMA TAKEN AS 100 Liver 54 Plasma 43 Ovary DoW) Yolk BS) Intestine 11 Spleen 0.1 Yolk of freshly 0 laid egg Chemical Investigations. In addition to the more strictly cytological investigations there have also been many contributions of a histochemical type on the changes that occur during oogenesis. Extensive accounts of these are given by Needham (31, °42), Marza (38), and Brachet (47), along with summaries of work done by direct chemical methods. In some of the recent work in this field, as in others, the use of radioactive tracers has provided valuable information. In particular this technique has contributed information concerning the question whether various constituents of the yolk are synthe- sized within the oocyte or in some other tissue of the body. A brief account of the pertinent experiments is presented in the following section. Tracer Experiments. The experiments re- lating to this have been done with radio- active phosphorus (P22) as a tracer for the formation of phosphatides (lecithin and cephalin) and phosphoprotein (vitellin) in eges of the hen by Hevesy and Hahn (’38), 178 Entenman et al. (’38), Chargaff (42) and Lorenz et al. (43) (see also Hevesy, °47). It is known, in the first place (see Needham, 31), that these compounds are formed by hens when the diet contains phosphorus in inorganic form. The organic phosphorus compounds are evidently synthesized by the hen. It is also known that the serum of lay- ing birds, as well as of reptiles and fish, con- tains larger amounts of phosphatides and phosphoproteins than are present in the serum of the males or nonlaying females (Roepke and Bushnell, *36; Laskowski, ’38; Landauer et al., 39; Zondek and Marx, ’39; Riddle, ’42; Chaikoff and Entenman, °46). EMBRYOGENESIS: PREPARATORY PHASES in the albumin and shell. Other experiments show that active phosphatides are only found in laid eggs that had been in the ovary at the time of injection, and tests up to 6% days after injection show a progressive in- crease in the activity of this fraction. Another control, soaking eggs for one day in a solution of labelled inorganic phosphate, showed no incorporation of P®? in the phosphatides extracted from the yolk, al- though active inorganic phosphate had pene- trated it. In the experiments of Entenman et al. (38) determinations were also made of the relative amounts of P®? incorporated in the TasxeE 10. Relative Radiophospholipid of Tissues of Laying and Nonlaying Birds at 6 and 12 Hours after Injection of 50 mg. of Phosphorus as NazHPO4 Containing 10° Ra- dioactive Units (1 Radioactive Unit = 2x10"? curie) (from Entenman et al., °38) LAYING NONLAYING 6 HRS. 12 HRs. 6 HRS. 12 uRs. Total radiophospholipid of bird as per cent ACLIMEMIS CONE CM lejos wee arenes eso ace Pave aiei stave ot orene elniav ars Per cent of total radiophospholipid found in: Pasthoimtestinaletnacteme reir ecko irekeieten tier: MACKS 5 > leeMe Sp lleeels ga nueadoccoonccnous00N reproductive system (ovary, oviducts, and yolks)... . This suggests that these substances might be synthesized in some other organ than the ovary. Hevesy and Hahn (738) determined the content of radioactive phosphorus in inorganic and organic form in the plasma, liver, intestinal mucosa, ovary, yolks and eggs of hens at various times after the sub- cutaneous injection of labelled sodium phos- phate. Table 9 gives the results of a set of analyses performed 5 hours after the injec- tion. The labelled phosphorus content of the phosphatide-phosporus is found to be relatively low in the ovary and the yolk of an ovarian egg at this time. It is much higher in the plasma and highest in the liver. Since the oocytes are growing rather rapidly, in contrast to the liver, the experiment gives strong indication that the phosphatides are synthesized in the liver and transported through the plasma to the ovary. Various controls were run in these experiments. No labelled phosphorus was found in the phos- phatides of the yolk of an egg laid at 5 hours after the injection, although active in- organic phosphate is found in it as well as =e OZ 4.55 3E25 4.57 560 ll) 10 23 15 wee 36 Dil 35 se cll 20 0.4 0.2 saa44 9) 47 44 phosphatides of various parts of the laying and nonlaying hen. The results of one of their experiments, given in Table 10, show that about half of the P®?-containing phos- phatide of the bird is present in the liver although this organ contains only about 5 per cent of the total phosphatides of the bird. In other experiments from Chaikoff’s laboratory the incorporation of P®? in phos- phatides and other compounds of the yolk was shown to be expressable as a function of the rate of growth of the yolk and the amount of P#?-phosphatide present in the plasma at the time (Lorenz et al., 43). These workers and others have also shown a high rate of labelled phosphatide forma- tion by liver slices of male birds, which could be increased significantly by injec- tion of estrogenic hormone (diethylstil- bestrol) (see Taurog et al., ’44). The incorporation of P#®? into the vitellin as well as in the free phosphatide and lipo- vitellin (a complex containing about half of the total phosphatide of the yolk) of the growing oocyte has heen investigated by GAMETOGENESIS, FERTILIZATION AND PARTHENOGENESIS 179 Chargaff (42). He finds equal rates of for- mation of free and bound _ phosphatide and somewhat higher initial formation of vitellin. The results of the various experiments strongly support the view that these com- pounds are formed elsewhere in the body and transported through the plasma to the growing oocyte. In the case of the phos- phatides the main site of synthesis appears to be the liver. To what extent this situation may hold for the various substances that the oocyte and estimate that the nuclei are 512-ploid when they have reached full size. On the basis of this and other considerations Painter (40, °45a,b) proposes that one of the chief functions of the nurse cells is to supply large quantities of nucleoprotein ma- terial to the egg so that it may be readily available for the formation of chromo- somes in the period of rapid division that fol- lows fertilization. For those species of ani- mals whose oocytes are not associated with nurse cells Painter suggests that the large germinal vesicle assumes this function. He Fig. 53. Oocyte and nurse cells in the leech Pisciola. Groups of four or five “oogonia” are set free into the lumen of the ovary and divide into a mass of about 50 cells (A) of which only one, as a rule, becomes an oocyte (B), while the remainder function as nurse cells, contributing to the growth of the oocyte (C) and finally de- generating. (From Jorgensen, 713.) accumulates is not known. However, it would not be too surprising if the oocyte were found to be incapable of synthesizing many or most of its stored materials. Pre- sumably the same enzymes that function in synthesis also operate in dissociation, so the lack or inactivation of various enzymes in the oocyte may permit it to perform its storage function more efficiently. Nurse Cells and Nucleoprotein Absorption. Another example of an important constitu- ent of the yolk that appears to be largely supplied to the oocyte in rather complete form is nucleoprotein. This is illustrated in oocytes of various species of animals that are associated during their growth with large nurse cells. The nurse cells at first increase in size and then, as the oocyte approaches full size, the contents of the nurse cells are absorbed by the egg (see Fig. 53). During the early stages the nuclei of the nurse cells increase considerably in volume. Painter and Reindorp (739) have described a series of endomitotic cycles oc- curring during the growth phase of the nurse cell nuclei of Drosophila melanogaster infers that endomitosis also occurs in the germinal vesicle and upon breakdown of the latter at the time of the first maturation division a considerable amount of nucleo- protein is set free in the cytoplasm. This view is opposed by Ris (°45), who interprets the “lampbrush” chromosomes of the ger- minal vesicle as typical diplotene chromo- somes in which there is a great longitudinal growth of the chromonemata and in which the apparent side branches do not represent additional chromosomal material but sim- ply major coils of laterally separated strands. It is, however, of interest to note that the oocyte nucleus remains relatively small in species that have nurse cells, whereas it forms the relatively large germinal vesicle in species that are not provided with such cells. Along with their general studies of the nucleic acid metabolism of various kinds of cells, Caspersson and Schultz (738, *40), Schultz (41), and Brachet (47) have con- tributed some interesting observations and experiments relating to the formation of nucleoproteins during oogenesis. Caspersson 180 and Schultz (’40) determined the ultravio- let absorption spectra of different regions within the nucleus and cytoplasm of a sea urchin oocyte. Their results are shown in Fig. 54. The curve for the nucleolus shows an absorption maximum at 2600 A typical of nucleic acid and a small rise at 2800 A indicative of proteins containing aromatic amino acids such as tyrosine and trypto- Cytoplasm near nucleus Peripheral, vesicle Germinal 4 | EXTINCTION COEFFICIENT ~ 2500 3000 A WAVE LENGTH Fig. 54. Ultraviolet absorption spectra of different parts of an ovarian egg of the sea urchin Psam- mechinus miliaris (from Caspersson and Schultz, 40). phan. The cytoplasm adjacent to the nu- cleus shows a similar type of curve with somewhat more protein indicated. The more peripheral cytoplasm gives a rather different curve indicative of considerably less nu- cleic acid. The nuclear sap appears to con- tain relatively little nucleic acid. This dis- tribution of nucleic acid, which is in accordance with cytochemical observations of Brachet (’47), suggests diffusion of nu- cleic acid formed in the nucleus out to the cytoplasm or diffusion from the nucleus of some agent active in this synthesis, or syn- thesis at the nuclear membrane. MATURATION DIVISIONS The details of meiosis are beyond the scope of the present work and have been presented in several texts (Wilson, ’25; Dar- lington, °37; Geitler, ’38; Sharp, ’43). A few words may, however, be said here con- EMBRYOGENESIS: PREPARATORY PHASES cerning possible causative factors. One strik- ing feature of the phenomenon is that it is essentially similar in cells that are so radi- cally different in appearance as spermato- cytes and oocytes. However, it is well known that preparations for the meiotic divisions are made at an early stage, shortly after the last gonial division, when the germ cells of the two sexes are more nearly alike in appearance. These preparations involve the intimate pairing (synapsis) of homologous chromosomes, a phenomenon that does not occur in the prophase of division of somatic cells except in the Diptera. Among various suggestions as to the fac- tors responsible for synapsis, the appeal to antigen-antibody type of interaction first made by Lindegren and Bridges (’38) seems to be particularly cogent. Friedrich-Freksa (40) has interpreted the gene by gene spec- ificity of pairing as due primarily to their arrangement in similar order in the homolo- gous chromosomes, which synapse by the operation of nonspecific forces. However, the preciseness with which the homologous genes adhere, even when translocated or inverted, would seem to require the inter- vention of forces having the high degree of specificity that is exhibited in antigen-anti- body reactions. Lindegren and Bridges sug- gest that the union is effected by the pres- ence of agglutinins that are specific for each chromomere of the chromosome. No suggestions are made concerning such im- portant details as to where and when these antibodies are formed, or what reverses the reaction in order to permit the chromosomes to separate. Another interesting hypothesis to account for homologus chromosome pairing has been offered by Fabergé (42). This assumes the operation of a hydrodynamical phenomenon known as the Guyot-Bjerknes effect, in which two spheres that pulsate in phase move towards one another as a result of the Bernoulli principle. The pairing units of the chromosomes are considered to have the properties of such pulsating spheres, each with a characteristic frequency so that it can attract its homologue but not non- homologous units having other frequencies. Fabergé points out that the force is a long range force such as seems to be required to achieve synapsis between homologous chro- mosomes that may be separated initially by a distance of several micra in the nucleus. Evidence for the operation of such long range forces has been presented recently by Hinton (’46) in a study of pairing of trans- GAMETOGENESIS, FERTILIZATION AND PARTHENOGENESIS 181 location-bearing chromosomes of Drosophila. However, Cooper (748) has effectively dis- puted the significance of this evidence and concludes that there is, as yet, no necessity for the assumption of long range forces. A view of meiosis that has gained both widespread acceptance and controversy is the “precocity theory” of Darlington (731, 37). According to this view the prophase begins relatively earlier in meiosis than in mitosis, so that the leptotene chromosomes are single rather than split. Synapsis then results from a “universal mitotic affinity of half-chromosomes for one another in pairs.” It is further assumed that this relationship is upset by the division of the chromosomes (at pachytene) but that chiasmata resulting from crossing over hold the original pairs together until metaphase, so that normal segregation may occur. It has, however, been shown by a number of workers (see Cooper 44, °49, for references and convincing evi- dence) that in many species of animals crossing over and the resultant chiasmata are not necessary for metaphase pairing and or- derly segregation. So, that part of Darling- ton’s hypothesis cannot be considered to have general validity. Also, there is considerable evidence (Schrader, 44; Mickey, ’46, °47) that the leptotene threads are already split, as are the prophase chromosomes in ordinary mitosis. A view that seems to be more con- sistent with the known features of meiosis is the “retardation theory” proposed by Sax and Sax (735). According to this view the spiral form of the chromonemata of mitotic chromosomes prevents (because a sufficient number of identical loci cannot come in con- tact) intimate association of homologous chromosomes. In the prolonged prophase of meiosis the chromonemata are greatly ex- tended and uncoiled, or have only very loose remnant coils, so that homologous chromo- somes can adhere intimately. Upon the oc- currence of a new split in each chromatid at late pachytene, coiling and separation of homologues begins. This view is consistent with the fact that the chromosomes are con- siderably longer in meiotic prophase than in ordinary mitotic prophase and with such apparently unusual cases as the dipteran giant salivary chromosomes in which the greatly extended polytene homologues re- main permanently synapsed. Undoubtedly the analysis of meiosis would be greatly helped by its experimental induc- tion in somatic tissue. It is interesting to note, then, that Huskins (’48) has reported the occurrence of chromosome segregation and reduction in onion root tips grown in a solution of sodium nucleate. However, as Huskins points out, further work will be re- quired to determine the relationship of this “somatic meiosis” to the gonocytic meiosis. FERTILIZATION One may define fertilization as the series of processes by which the spermatozoon initiates and participates in the development of the egg. As such it includes all steps from the approach of the spermatozoon to the fu- sion of the pronuclei within the egg. In this chapter we shall consider briefly some of the factors that may be operative in various phases of this series of processes. APPROACH OF THE SPERMATOZOON In most species of animals the meeting of egg and spermatozoon is facilitated by virtue of the fact that the latter is a motile cell. Also, the large numbers that are ordinarily available in both external and internal in- semination contribute to the likelihood of contact being made with the eggs. For ex- ample, Farris (49) reports that fertile men supply semen samples with a total of 83 million or more motile sperm, whereas a lower total active sperm count is correlated with infertility. However, the question has been raised by many investigators whether or not the spermatozoa are attracted in some way to the egg when their random move- ments have brought them within a certain distance of the latter. The early investiga- tions concerning chemotaxis has been re- viewed by Morgan (727), who concluded that there is no critical evidence demonstrat- ing its existence. Such evidence, as well as more recent work along this line (Corn- man, ’41; Hartmann, ’40; Vasseur and Hag- strom, ’46), rests on demonstrating a local accumulation of sperm within tubes or other devices containing eggs or certain materials derived from the eggs. However, it is not readily feasible to distinguish between an attractive influence and a trap action effect such as was described many years ago by Jennings (06). In ferns and mosses there is good evidence for chemotaxis but it has not, as yet, been adequately demonstrated for ani- mal spermatozoa (cf. Rothschild, ’51a, b, 52). On the other hand an increase in the mo- tility of spermatozoa under the influence of material emanating from the egg has been noted in some species of animals. This in- crease does not generally occur when the 182 spermatozoa are in a highly active condition (see Lillie, °19; Gray, 28). Along with in- creasing the activity of the spermatozoa the egg water may also increase the rate of oxygen uptake. For example, Gray (28) found increases ranging from 212 to 425 per cent in the sea urchin Echinus esculen- tus, and similar increases have been found by Tyler (48a) in the keyhole limpet Megathura crenulata. Hartmann et al. (739) reported that the agent responsible for the activation of the spermatozoa in Arbacia pustulosa is the pigment echinochrome which is present in the eggs of this species of sea urchin. However, attempts to confirm this were unsuccessful both in a species of sea urchin whose eggs do not contain echino- chrome and in one whose eggs are so pig- mented (Tyler, ’°39b; Cornman, ’41). Tyler and Fox (739, ’40) found the activating agent in the egg waters of Strongylocentrotus and of Megathura to remain associated, after dialysis and precipitation, with the large molecular substance that has agglutinating action (see below) on the sperm. While such association has also been reported by Corn- man (41), Kuhn and Wallenfels (’40), and Vasseur and Hagstrém (46), these workers find the activating agent to be at least par- tially dialyzable. Many years ago Clowes and Bachman (’21) found that distillates of Arbacia egg water would activate the sperm and this was confirmed more recently by Cornman (741). It would appear, then, that the activating agent in egg water is normally bound to the agglutinating agent, from which it may be split off as a relatively small molecular substance, but its exact chemical nature has not, as yet, been determined. LIFE SPAN OF THE GAMETES The conditions of insemination in most animals require that the gametes remain in a fertilizable state for a period of time fol- lowing release from the gonads. It is of in- terest, then, to inquire into the factors involved in the aging of the gametes. Senescence of Spermatozoa. In some species of animals, such as the bat, the honeybee and certain terrestrial isopods (Vandel, ’41) the sperm may survive for periods of several months to several years in the female genital tract. Apart from such special cases the usual life span of shed sperm is of the order of several hours to a few days (see Hart- man, ’°39). An important factor that is known to affect markedly the functional life span of sperm is dilution. Many workers have shown EMBRYOGENESIS: PREPARATORY PHASES (see Morgan, °27; Gray, ’28; Rothschild, 48a, b,c) that the duration of life varies directly with the concentration of the sus- pension. The relative or complete lack of motility of sperm in undiluted semen of sea urchins and other animals has frequently been attributed to a relatively high tension of carbon dioxide. Rothschild (’48b) has shown, however, that lowering the carbon dioxide tension does not induce motility, whereas increasing the oxygen tension ac- tivates the spermatozoa. By exposing the semen alternately to nitrogen and oxygen the spermatozoa could be made alternately inactive and active. Hartmann, Schartau and Wallenfels (40) had proposed that immo- bility in semen, as well as subsequent senes- cence of sea urchin sperm, was due to a substance, termed androgamone I, that is liberated by the spermatozoa. This view stood in contradiction to observations of Gray (28) and Hayashi (45) showing that sea urchin sperm were as motile when di- luted with seminal plasma as with sea water. Rothschild (48c) showed in addition that the supernatant of a dense two-hour sperm suspension, which is supposed to contain androgamone I, had no inhibitory effect on respiration. These and other experiments show that senescence is not attributable to substances, such as the hypothetical andro- gamone I, diffusing into the medium from the sperm and that the “dilution effect” cannot be explained on the basis of a dilu- tion of such inhibitor. More recently it has been shown (Tyler and Atkinson, ’50) that the life span of sea urchin sperm can be considerably extended by addition of any one of a number of amino acids to the suspension. Similar results have been obtained in birds and mammals (Lorenz and Tyler, 51; Tyler and Tanabe, 52). It was shown (Tyler and Rothschild, 51) that the amino acid is not utilized metabolically, although it has effects on the respiratory metabolism of the sperm and even enables sea urchin sperm to remain motile anaerobi- cally, whereas they ordinarily die promptly in absence of oxygen. The results suggested that the amino acids act by binding certain trace metals present in the sea water. Tests with other kinds of metal-chelating agents such as ethylenediaminetetraacetic acid (Versene*), diethyldithiocarbamic acid, 8-hy- droxyquinoline, and a-benzoinoxime have given similar prolongation of the life span of the sea urchin spermatozoa (Tyler, 53). * Trade name of Bersworth Chemical Co., Fram- ingham, Mass. GAMETOGENESIS, FERTILIZATION AND PARTHENOGENESIS These findings can account in large part for the “dilution effect.” The proteins of the seminal plasma can also bind heavy metals. Since such protein would be present in higher concentration the less the semen is diluted, the spermatozoa in denser suspen- sions would be protected from the toxic ac- tion of the trace metals. This is consistent with the favorable action of seminal plasma found by Gray (?28) and Hayashi (’45). Another interesting effect of the amino acids and other metal-chelating agents is 100 80 60 pl. O2 40 20 O 20 40 183 fluid. Motility is maintained under anaerobic conditions in the presence of glycolyzable substrates. Mann (46, ’49) has recently identified the normal substrate in seminal fluid as fructose. The work of Mann (45a, b, 49), Lardy and Philips (’41—45), Macleod (3946) and others has shown that the glycolysis in sperm is essentially the same as in other animal tissues and in yeast (phosphorylation of the sugar by adenosine triphosphate in the presence of hexokinase to hexosemonophosphate, then hexose di- O { { ' | ' ' 1 { 1 i} ' ' 4 60 80 100 TIME (MIN.) Fig. 55. Photo-reversible inhibition by carbon monoxide of the respiration of sperm of the sea urchin Echinus esculentus. Gas mixture, 90% CO and 10% Og; dark periods, 0 to 30 and 60 to 90 minutes; light periods (700 watt lamp), 30 to 60 and 90 to 120 minutes; temperature, 15:1° C.; sperm suspension, 4.4 108 per ml. Broken line shows average rate of O2 consumption per hour. (From Rothschild, ’48a.) their ability to enable the sperm to induce a good fertilization-reaction under condi- tions in which the sperm, diluted and aged in ordinary sea water, call forth poor mem- brane elevation (Tyler and Atkinson, 50; Tyler, 53). The results demonstrate that the type of response given by the egg can be determined by the vitality of the imseminat- ing sperm. The spermatozoon evidently does not act simply in “all or none” manner in the sense of operating a trigger mechanism in the egg. Investigation of the substrate that is normally utilized by the sperm and the enzy- matic pathways through which it is metabo- lized has been confined mainly to mam- malian spermatozoa. It is now well established that mammalian sperm have a_ predom- inantly glycolytic metabolism even un- der aerobic conditions and that the substrate is a reducing sugar present in the seminal phosphate, and breakdown through triose- phosphate, phosphoglyceric acid and pyruvic acid to lactic acid; with resynthesis of ATP by transfer of phosphate from phosphogly- ceric acid to adenylic acid or reaction be- tween inorganic phosphate and adenylic acid catalyzed by cozymase). Aerobically in the absence of external glycolyzable substrate the spermatozoa apparently oxidize phos- pholipid. The mammalian sperm have been found to contain all the enzymes involved in the Krebs’ cycle, as well as the cyto- chromes a, b, and c and cytochrome oxidase. However, under aerobic conditions motility is relatively quickly lost even in the presence of glycolyzable substrate. This has been at- tributed by MacLeod (46a) to the produc- tion of hydrogen peroxide, which he believes may reach sufficient concentration to kill the sperm. Addition of substances containing catalase was found to retard the aerobic 184 loss of motility. These experiments of Mac- Leod on human spermatozoa have been con- firmed and extended by Tosic and Walton (46) and Tosic (47) with bull spermatozoa. In contrast to mammalian sperm, that of sea urchins becomes rapidly immotile under anaerobic conditions, as Harvey (730) dem- onstrated. Rothschild (’48a, b,c) and Spikes (48, °49b) have shown that glycolysis does not occur to any very appreciable extent aerobically or anaerobically in presence or absence of glycolyzable sugar. Spikes pre- sents evidence to show that oxidation pro- ceeds through the usual fructose diphosphate, triosephosphate and Krebs cycle pathway. From the spectroscopic identification of cyto- chromes a, a3, b, c, and COaz (Ball and Meyerhof, ’40; Rothschild, ’48a) and the demonstration (Rothschild, ’48a,c) of photo- reversible CO-inhibition (Fig. 55), it is con- cluded that the respiration of the sea urchin sperm is under control of the cytochrome system. Rothschild (’48c) has also shown that oxidizable substrates and their dehydro- genases are still present in sea urchin sperm that have aged to the point of no motility and respiration. Recent evidence (Rothschild and Cleland, 52) points to phospholipid as the principal endogenous source of energy for motility. From the experiments on pro- longing the duration of motility and fertiliz- ing capacity by means of amino acids and other metal-chelating agents (Tyler, °53) it is also clear that death of the spermatozoa upon dilution in ordinary sea water is not due to exhaustion of their food supply. These agents, evidently by binding toxic trace met- als, enable the sperm to utilize their endog- enous substrate more fully. Senescence of Eggs. The unfertilized egg likewise has a relatively limited life-span under ordinary conditions, both in animals with external fertilization and in those with internal fertilization. In certain cases, as in many fish and Amphibia, fertilizability is lost within a few seconds or minutes after deposition in water. In these, marked visible changes, involving elevation of a membrane, are generally noted upon contact with the new medium and these are very likely in- volved in the loss of fertilizability. A rapid loss of fertilizability may, however. occur without such visible changes, as in the case of Platynereis described by Just (715b,c). In most animals loss of fertilizability and cytolysis of the ege occur in a period of sev- eral hours to one or two days under ordinary conditions. This relatively rapid senescence cannot be attributed to depletion of endog- EMBRYOGENESIS: PREPARATORY PHASES enous nutrient since, if fertilized, the eggs can survive considerably longer without any added nutrient. For example, fertilized sea urchin eggs will survive for about two weeks without any external source of food, while respiring at a rate that is at least ten times that of the unfertilized egg. The latter might, then, be expected to survive over twenty weeks instead of the two days obtained under ordinary conditions. Experiments by Whitaker (’37), Tyler, Ricci, and Horowitz (’38), Tyler and Dessel (739), Schechter (737, ’41) and others have shown that various agents such as weak alcohol, slight acidity, and low calcium con- tent of the medium can prolong the fer- tilizable life of eggs of marine animals to some extent. It has also been shown (Tyler et al., °38) that sterile conditions extend the survival time of sea urchin eggs by five-fold or more. The deleterious effect of bacteria does not manifest itself until after a certain immune period, since freshly shed eggs can survive in dense bacterial suspensions almost as long as in ordinary sea water. Corre- sponding with the onset of the susceptible period the surface of the egg undergoes some disintegrative changes which are manifest by the formation of a tight membrane or no membrane upon fertilization. In the absence of bacteria the eggs remain viable and fer- tilizable for a considerable time after these changes have begun. The agents mentioned above that prolong the functional life of the egg in nonsterile conditions evidently operate by delaying the onset of these auto- lytic changes. Runnstrém (’49) has described in some detail the changes that unfertilized eggs of sea urchins undergo upon aging. Alterations in the surface are detectable by various sorts of tests, such as the hypertonicity test in which, as the eggs shrink, the formation of wrinkles is followed. Ripe unfertilized eggs, shrinking in hypertonic solution, form nu- merous wrinkles on the surface (indicative of a semisolid state), which later smooth out. Eggs which have aged, so that they form low membranes or no membranes upon fertilization. shrink with a smooth surface, indicative of a liquefied state of the surface. It seems reasonable to assume that energy - is required to prevent the breakdown of the surface that occurs upon aging. It is of considerable interest, then, that adenosine triphosphate, at low concentrations (0.002 to 0.001 M) has been found (Wicklund, °49, reported by Runnstrém, °49) to be effective in prolonging fertilizable life and even capa- Ke r\ Vv) : : » ' ble of restoring normal membrane-forming capacity to aged eggs. This may mean that the rapid senescence of the unfertilized eggs is due to loss (failure of resynthesis or un- availability) of some such energy-transport- ing agent. ATTACHMENT AND PENETRATION The specificity of fertilization is mani- fested in the initial steps of the union of the gametes. In general spermatozoa will not adhere to, or penetrate, eggs of other spe- cies or other tissues of the same species. Some exceptions to this are known. Thus sperm of the nemertean worm Cerebratulus penetrate eggs of the sand dollar Echinarach- nius (Chambers, ’33). The sperm remain in the cortex without activating the egg, which can subsequently develop normally upon fer- tilization with the species sperm. Instances of penetration of sperm into tissue cells have been less frequently reported (see Terni and Maleci, *37). In view of the marked specificity and the superficial resemblance to phagocytosis it is not surprising that the first detailed modern theory of fertilization, namely that of F. R. Lillie (19), should be based on immunological analogies. This was undoubtedly favored, too, by the discovery (Lillie, 12, °13a,b) that spermatozoa of certain marine animals can be made to ag- glutinate by the addition of some egg water (supernatant of a suspension of eggs) of the same species. The agglutinating sub- stance in the egg water was termed fer- tilizin and Lillie assigned it a central role in his theory of fertilization. The early work in this field has been reviewed by Lillie (19), Lillie and Just (24), Morgan (727), Dalcq (28b) and Just (730). The present discussion will be based mainly on the more recent work, which has been reviewed in some detail by Tyler (’48a, 49), Bielig and Medem (749), and Runnstrém (’49). We shall consider four groups of inter- acting substances that have been obtained from eggs and sperm and which are termed fertilizins, antifertilizins from sperm, anti- fertilizins from eggs, and lytic agents from sperm. The terms gynogamone and andro- gamone were introduced by Hartmann (40) to designate the substances derived from eggs and sperm, but this terminology does not seem to be any less prejudicial than that of Lillie, nor sufficiently advantageous to warrant its adoption. Fertilizins. Agglutination of sperm by ho- mologous egg water has been reported in J’ GAMETOGENESIS, FERTILIZATION AND PARTHENOGENESIS 185 many species of animals among the echino- derms, annelids, mollusks, tunicates and ver- tebrates (see Tyler, “48a, and Bielig and Medem, *49, for references). The reaction in strong egg water is usually visible within a few seconds. In sea urchins the clumping is ordinarily head to head (Lillie, 19; Elster, 35). Tail to tail and head to tail, along with head to head unions have been re- ported in certain mollusks and may occur also in aged sea urchin sperm (Sampson, 29; Tyler, ’40a). Figures 56 to 58 illus- trate the agglutination reaction of sperm of the keyhole limpet Megathura crenulata. The reaction is similar in appearance to that exhibited by sperm that are aggluti- nated by an immune serum (compare figures of Henle et al., °38). The reaction also ex- hibits such serological features as the zone phenomenon (Tyler, ’40a; Spikes, 49a). Lil- lie (19) used the term fertilizin to designate the agent or agents in egg water that were responsible for agglutinating, activating and possibly chemotactic and other effects. At present it seems best to restrict the use of this term to the agglutinin. A unique feature of the reaction in sea urchins is its spontaneous reversal after a period of time ranging from a few seconds in dilute egg water to many minutes in con- centrated egg water, whereas in other groups of animals the agglutination is essentially permanent or long-lasting. Following spon- taneous reversal, in the sea urchin, the spermatozoa are incapable of being re- agglutinated by the addition of fresh egg water and have lost their fertilizing capacity, although they are fully motile (Lillie, 719; Tyler, 41). An interesting analogy to this is found in the Hirst (’42) reaction of ag- glutination of vertebrate red blood cells by influenza, and other viruses. This reaction also reverses spontaneously and the red cells are incapable of re-agglutination. In the case of the sea urchin the reversal has been in- terpreted (Tyler, ’41) as due to a splitting of the individual fertilizin molecules that bind the spermatozoa to one another. This interpretation presumes that the now gen- erally accepted mutual multivalence theory of antigen-antibody reactions (Heidelberger, 38, °39; Marrack, ’38; Pauling, ’40) holds for the agglutination of sperm by fertilizin. According to this theory both antigen and antibody must be multivalent with respect to their combining groups in order for their interaction to result in precipitation or ag- glutination. In other words, an agglutinin molecule must possess two or more combin- 186 EMBRYOGENESIS: PREPARATORY PHASES Fig. 56. Macroscopic appearance of agglutination reaction in the keyhole limpet Megathura crenulata. Photographed in Syracuse dishes, x 14. a, Untreated sperm suspension (ca. 2 per cent); b, c, and d, 15 sec- onds, 30 seconds and 10 minutes, respectively, after addition of solution of fertilizin. (After Tyler, ’40a.) C Fig. 57. Photomicrographs of agglutinated sperm of the keyhole limpet Megathura crenulata. a, Three agglutinates formed in a moderately strong solution of fertilizin, showing spherical shell of sperm heads sur- rounding central mass of sperm; < 50. b, An agglutinate formed in a strong solution of fertilizin, showing incomplete shells of sperm heads attached to main mass by the ends of the tails; X 170. c and d, Fusion of two agglutinates; d was photographed 15 seconds after c; X 85, (From Tyler, ’40a.) GAMETOGENESIS, FERTILIZATION AND PARTHENOGENESIS 187 ing groups available to unite separate cells. Univalent molecules would combine with the separate cells and be incapable of tying them together. Some support for the above interpretation of the spontaneous reversal of agglutination in sea urchins is obtained from experiments apparent absence of fertilizin in many spe- cies of animals. It is known that the egg waters of many species of animals, such as the abalone and Cumingia (Sampson, ’22), Urechis (Tyler, 41), some starfish (Metz, °45) and even certain sea urchins (Vasseur and Hagstrém, °46) fail to agglutinate the Fig. 58. Photomicrographs of sperm of Megathura crenulata, X 260. a, b, and c, Show union by heads and by end pieces of tails in strong, moderate and weak fertilizin solutions, respectively. d and e, Show head ag- gregates originally present in sperm suspension. (From Tyler, ’40a.) in which the ordinary agglutinin is con- verted into a nonagglutinating (univalent) form by the action of heat, proteolytic en- zymes, ultraviolet or x-radiation (Tyler, *41, 42; Metz, ’42a,b). Treatment of the sperm with such univalent fertilizin inhibits their agglutination by ordinary fertilizin and also renders them nonfertilizing, without im- pairing their motility. The concept of univalent antigen or anti- body also provides an interpretation for the homologous spermatozoa. Such failure to demonstrate fertilizin in all species of ani- mals has been used as an argument against considering fertilizin to have any general significance for fertilization. However, that the egg water of such nonagglutinating spe- cies might contain a fertilizin-like substance was indicated in experiments of Tyler (41) and clearly established by Metz (745). The latter worker was able to cause specific ag- glutination of sperm, in species of starfish 188 that normally do not give an agglutination reaction, by the addition of a nonspecific adjuvant (from hen’s egg white and other sources) along with the egg water. The ad- juvant alone has no agglutinating action on the sperm. Thus the failure of agglutina- tion to occur ordinarily in certain species of animals is not necessarily indicative of the absence of fertilizin in the egg water but can be interpreted on the basis of a uni- Taste 11. Method of Preparation of Sea Urchin Fertilizin (from Tyler, 48a) 1. Extract 20% suspension of washed eggs in sea water at pH 3.5. 2. Decant supernatant (agglutination 1000), centrifuge or filter. 3. Add 40 ml. N/1 NaOH per liter of supernatant (fertilizin precipitates with the Ca and Mg salts of sea water). 4. Suspend precipitate in 3.3% NaCl, neutralize, dialyze. 5. Remove insoluble particles and precipitate with 1144 to 14% volumes 95% alcohol. 6. Wash with alcohol and dry (re-precipitate with alcohol or saturated (NH4)2SO,4). (Yield = ca. 250 mg. /liter) titer = ca. valent condition of either the fertilizin or the sperm. A parallel to univalent fertilizin can be found in the field of immunology in recent studies on Rh antibodies. In most cases these antibodies are found to occur in a nonag- glutinating, univalent form (Wiener, 44; Race, ’44; Fisk and Morrow, ’45; Levine and Walker, ’46). Here, too, the addition of an adjuvant (such as serum albumin) enables the antibodies to agglutinate the Rh-positive cells (Wiener, *45a,b; Wiener and Gordon, "48; Diamond and Denton, °45; de Burgh et al., 46). Corresponding to the experimen- tal conversion of fertilizin into a univalent form, immune antibodies have been simi- larly altered by treatment with various agents, such as heat, diazo compounds, for- maldehyde and photo-oxidation (see Tyler, 45a, for references). The alteration appears to involve not simply a splitting of the antibody molecule but also a reassociation of the univalent fragments with fragmented non-antibody protein of the antiserum. Spec- ificity is retained by such univalent anti- bodies and, in the case of an antitoxin, protective properties persist. The treated antisera are found to have lowered anti- genicity and may, therefore, offer a means of avoiding serum sickness (Tyler, ’45a, b; EMBRYOGENESIS: PREPARATORY PHASES Tyler and Swingle, ’45). In connection with fertilization, further use has been made of univalent antibodies in experiments with sea urchin sperm (see below). A fair amount of information is now avail- able concerning the chemical nature of fertilizin. Lillie (19) showed that the gelat- inous coat of the sea urchin egg contained large amounts of fertilizin but believed that it was being continuously secreted by the ripe unfertilized egg throughout its functional life. However, more recent ex- periments (Tyler and Fox, ’39, ’40; Tyler, 40a, 41) have demonstrated that it is iden- tical with material of the gelatinous coat itself and not obtainable from denuded eggs. This has been corroborated by several other workers (Evans et al., 41; Hartmann, 40; Vasseur and Hagstrom, 46; Runnstrém and Lindvall, ’46). The presence of fertilizin in ordinary egg water results, then, from the slow dissolution of the coat. For chemical purposes sea urchin fertilizin can be ob- tained in high titer by dissolving the coat TaBe 12. Analysis of Electrophoretically Ho- mogeneous Preparations of Fertiliz- in of the Sea Urchin Strongylocen- trotus purpuratus (from Tyler, ’49, and unpublished) Nitrogen 5.6—58 % Carbon ses) Hydrogen 5G Sulfate 23 % Phosphate 0.06% Reducing sugar S25. YG Amino acids >20 1% Galactose at Methylpentose (fucose) a Glucuronic acid — Probable amino acids (by paper chroma- tography and micro- biological assay) Glycine, alanine, serine, threonine, valine, leucine, isoleucine, as- partic, glutamic, argin- ine, lysine, phenylal- anine, tyrosine, trypto- phan, proline Molecular weight ca. 300,000 with dilute acid, without damage to the rest of the egg. With such preparations Tyler and Fox (739, ’40) obtained evidence for the protein nature of fertilizin of the sea urchin and of the keyhole limpet, on the basis of such properties as non-dialyzabil- ity, salting out, common color tests and inactivation with proteolytic enzymes. How- ever, the low values (ca. 5 per cent) ob- tained for the nitrogen content indicated that the material was not a simple protein. GAMETOGENESIS, FERTILIZATION AND PARTHENOGENESIS 189 Carbohydrate was later found in purified fer- tilizin (Tyler, *42) and in electrophoretically homogeneous gelatinous coat (Runnstrém et al., °42) of sea urchins. From this and later work (Tyler, ’48a,b, "49; Krauss, °49; Vasseur, ’48a,b, 49) it has become clear that fertilizin is a complex of amino acids and sugars that may be termed a glycoprotein. A procedure for the preparation of elec- trophoretically homogeneous sea urchin fer- tilizin is outlined in Table 11. Some of the analytical data obtained (Tyler, ’48a,b, ’49) on such preparations of fertilizin of the sea da Upon electrophoresis the fertilizins of E. cordatum (Runnstrom et al., ’42) and of S. purpuratus (Tyler, ’48a, ’49) migrate towards the anode at pH values as low as pH 2. This highly acidic character is prob- ably correlated with the high content (ca. 25 per cent) of sulfate found by Vasseur (C47) in E. esculentus and confirmed by Ty- ler (48b) in S. purpuratus. Upon ultracen- trifugation of a jelly coat solution of Psammechinus, Runnstrém et al. (42) found a main component with a_ sedimentation constant of 2.9 x 10°13 varying with con- Fig. 59. Eggs of the sea urchin Strongylocentrotus purpuratus, photographed in Syracuse dishes, < 749. a, Un- treated egg-suspension. b, 15 minutes after addition of a solution of antifertilizin. (From Tyler, ’40b.) urchin Strongylocentrotus purpuratus are given in Table 12. The values given there for amino acids and reducing sugars are probably low, since some of these very likely contribute to the humin residue that forms upon acid hydrolysis. By means of paper chromatography Vasseur and Immers (’49) find differences in the sugar components of the hydrolyzed jelly coat of different species of sea urchins; namely galactose in Echinus esculentus, fucose in Echinocardium corda- tum, fucose and galactose in Strongylocen- trotus droebachiensis, fucose and glucose in Paracentrotus lividus. The amino acids have also been investigated by microbiological methods, but there is as yet insufficient evi- dence to indicate whether or not different species differ in this respect too. It is prema- ture to conclude that specificity of the fer- tilizins is dependent upon differences in par- ticular sugar or amino acid constituents. Even when the constituents are the same, differences in configuration of the molecule may determine specificity of action, as ap- pears to be the situation with immune anti- bodies. centration in the manner indicative of non- spherical molecules. With active fertilizin of S. purpuratus a sedimentation constant of 6.3 X 107% has been obtained (Tyler, °49). The molecular weight is evidently, then, greater than 82,000, which would be the value for spherical shape. Antifertilizins from Sperm. The substance on the surface of the sperm with which fertilizin reacts has been termed antifer- tilizin. Frank (’39) and Tyler (’39a) were able to extract such a substance from the sperm of sea urchins and the keyhole limpet, by means of brief heating or freezing and thawing. It is also extractable (Tyler and O’Melveny, *41) by slight acidification of the suspension. It can be assayed by its ability to neutralize the agglutinating ac- tion of fertilizin on sperm. Another manifes- tation of its activity is its ability to agglutinate a suspension of eggs (Fig. 59). In so doing it causes a precipitation mem- brane to form on the surface of the gelati- nous coat of the egg (Fig. 60). With strong solutions this membrane thickens and con- tracts within a short time to the surface 190 of the egg and is then no longer readily visible. It has been claimed (Hartmann et al., 40) that the jelly coat dissolves under the influence of such extracts. However, d | f EMBRYOGENESIS: PREPARATORY PHASES shown it to be an acidic protein, isoelectric at pH 3, and containing about 16 per cent nitrogen. Hultin (’47a,b) considers the active principle to be a basic protein. This oe : . cou o Mate : Fig. 60. Eggs of the sea urchin Lytechinus pictus. a, b, c, and d, Successive pictures of the same egg at about 14 minute intervals after addition of a solution of antifertilizin, showing formation of the precipitation membrane and its contraction to the surface of the egg; x 400. e and f, Successive pictures of the same egg and an adjacent isolated jelly-hull at 1 minute and 20 minutes, respectively, after addition of antifertilizin, showing the persistence of the material of the isolated jelly-hull at the time when the precipitation membrane has contracted to the egg’s surface and has become indistinguishable from it; X 350. g and h, Successive pic- tures of the same fertilized egg at 1 minute and 3 minutes after addition of antifertilizin. i and 7, Successive pictures of the same fertilized egg in 2-cell stage at 1 minute and 4 minutes after addition of antifertilizin; < 350. (From Tyler, ’49.) tests with isolated jelly hulls (see Fig. 60e, f) show that the precipitation mem- brane persists long after it has reached the surface in the control eggs. Chemical studies on antifertilizin (Tyler, 39a, ’40b, ’48a; Runnstrém et al., 42) have contention is refuted by Metz (749) who had earlier (42a) found that basic proteins from sea urchin sperm have antifertilizin-like ef- fects, but, in contrast to the above anti- fertilizin preparations, they also agglutinate sperm. GAMETOGENESIS, FERTILIZATION AND PARTHENOGENESIS 191 Upon extraction of antifertilizin by acidi- fied sea water the head of the spermatozoon is found to round up and the surface to become diffuse in the region between the acrosome and midpiece, while the latter c of this agent are similar to those described for the antifertilizin from sperm. In other words, antifertilizin from within the egg can react with fertilizin on the surface of the same egg. The active agent appears to d Fig. 61. Electron micrographs of sperm of the sea urchin Lytecyinus pictus, showing changes upon ex- traction of antifertilizi; x 30,000. a, Control, unextracted sperm. b, Extracted at pH 3.5. c, Extracted at pH 3.0. d, Extracted at pH 2.8. (From Tyler, ’49.) structures and tail are not visibly affected (Fig. 61). Antifertilizin is, then, considered to be located on the lateral surface of the head between the acrosome and midpiece. Antifertilizin from Eggs. Lillie (14) had postulated the presence of an antifertilizin within the egg and Tyler (40b) was able to obtain such an agent by extraction of sea urchin eggs that had been previously di- vested of their gelatinous coat. The effects be a protein (Tyler, *40b, ’48a). Injection of the extracts into rabbits induces the pro- duction of antibodies that not only pre- cipitate the homologous antigen but also agglutinate sperm of the same species, indi- cating antigenic similarity to the antifer- tilizin of sperm. Consideration of the finding that sub- stances capable of interacting in antigen- antibody—like manner can be obtained from 192 the same cell, along with other evidence from the literature of immunology, has sug- gested an auto-antibody concept of cell struc- ture, growth and differentiation, the details of which are presented elsewhere (Tyler, ’40b, °42, °46b, °47). Lytic Agents. Sperm extracts of various species of vertebrates and invertebrates have been found to possess the property of break- EMBRYOGENESIS: PREPARATORY PHASES (Tyler ’39a, ’48a; Krauss, ’49). Evidence has been obtained that its activity is dependent upon the presence of sulfhydryl groups in the molecule. Considerable attention has been paid in recent years to the lytic agent in the sperm extracts of mammals (see reviews by Duran- Reynals, °42; Meyer, ’47; Meyer and Rap- port, °52). This agent causes dispersal of sia se ie Fig. 62. Photomicrograph of an egg of Megathura crenulata at: a, 1 minute; b, 134 minutes; c, 214 minutes; d, 34%, minutes after addition of a sperm extract containing the egg-membrane lysin. X 200. (From Tyler, ”39a.) ing down the membrane or viscous coat that normally surrounds the unfertilized egg in these species (see Hibbard, ’28; Wintrebert, (2905335) vamane,, 30.) 35: Pimcus., 30, 2360: Tyler, ’39a; Medem, *42). Figure 62 illus- trates the action of such an agent on eggs of the keyhole limpet. The dissolution of the membrane (which can withstand hours of treatment in-concentrated acid) can occur in less than a minute in concentrated sperm extracts containing the lytic agent. The ac- tive agent has protein properties, is highly heat-labile (about 1 minute half-life at 50° C.), and separable from the antifertilizin present in keyhole limpet sperm extracts the follicle cells that normally surround the unfertilized tubal egg, by dissolving the cementing material between them. McClean and Rowlands (’42) and Fekete and Duran- Reynals (’43) showed that this was evi- dently the same agent that was earlier shown to have a dissolving action on the intercel- lular cementing material of skin and that had been designated hyaluronidase because of its ability to break down the muco- polysaccharide known as hyaluronic acid. Hyaluronidase is also a heat-labile protein and it has, as yet, been only partially puri- fied (see Meyer, 47; Meyer and Rapport, 52). The occurrence of this agent in vari- / GAMETOGENESIS, FERTILIZATION AND PARTHENOGENESIS ous invasive bacteria has served to emphasize the relation of the processes of fertilization to those of infection. Another type of lytic agent has been ob- tained by Runnstroém et al. (42~46) in methanol extracts of sea urchin sperm. This agent has a liquefying action on the vitel- line membrane of the unfertilized egg, an effect which can be duplicated by certain detergents. Present evidence (Runnstrém, 49) indicates that it is an unsaturated, eighteen-carbon, fatty acid. 193 motility) is also impaired when they are treated with specific antisera that are pre- pared by injection of rabbits with antifertil- izin and that have been rendered non- agglutinating (univalent) so as to avoid the complications that would be introduced by agglutination (Tyler, ’46a). Studies of the specificity of the reaction have provided some further information concerning the role of these substances. Lillie’s (719) demonstration that fertilizin was not obtained from other tissues than the Tasie 13. Comparison of Cross-fertilization with Cross-agglutination among Echinoids (from Tyler, ’49)* EGGS OR FERTILIZIN OF S. purp S. fran. S. purpuratus 3000 52 S. franciscanus 1 600 0 512 L. pictus 2 64 32. D. excentricus 1% 40 4 SPERMATOZOA OF L. pictus D. excent. 1 5 64 4 1 2% 4 0 850 2% 64 8 2 4400 1 128 * The upper figures of each pair of rows represent the number of times the sperm suspension is diluted in giving the end point value (2 per cent) of fertilization under certain standard conditions. The lower figures are the agglutination titers in terms of the highest dilution of fertilizin solution that gives visible agglutination. Role in Fertilization. Various experiments have been performed to attempt to elucidate the function of fertilizin and antifertilizin in fertilization (see Tyler, 48a; Runnstrom, ’49). When sea urchin eggs are deprived of ‘their gelatinous coat they are still fertilizable, but require a higher concentration of sperm /\. to effect fertilization. Whether or not, in a addition to facilitating fertilization, the presence of fertilizin is essential for the process cannot, as yet, be readily decided since upon removal of the gelatinous coat the surface of the egg still evidently possesses a thin layer of fertilizin that cannot be re- moved without damage to the egg. The presence of excess fertilizin in solution around the eggs tends to inhibit rather than promote fertilization. The sperm that have reacted with fertilizin at some distance from the egg are apparently incapable of adhering to the egg, presumably because their com- bining sites are already occupied by fertilizin. When sperm are partially depleted of anti- fertilizin by treatment that does not markedly impair their respiratory activity there is considerable lowering of their fertilizing capacity (Tyler and O’Melveny, ’41). The fertilizing capacity of sperm (but not their eggs has been amply confirmed and impli- cates this agent as furnishing the basis for the tissue-specificity of fertilization. That the action of fertilizin is predominantly species-specific has also been demonstrated by many workers (see Lillie, °19; Just, ’30; Elster, °35; Tyler, ’48a, 49), and in several investigations comparison has been made between the degree of cross-agglutination and that of cross-fertilization of various species of animals. The data of Table 13 il- lustrate some of the results that have been obtained in experiments of this type per- formed with four different species of echi- noids. Other species of animals among the asteroids, annelids, and mollusks (not in- cluded in the table) that do not cross- agglutinate with the echinoids, do not give cross-fertilization. In general the results show that the degree of cross-agglutination is greater than that of cross-fertilization among the various species. On the other hand, in combinations where cross-reaction of fertilizin and sperm is lack- ing fertilization does not generally occur. There are apparent exceptions to this, such as may be noted in Table 13, where S. franciscanus fertilizin fails to agglutinate 194 S. purpuratus and D. excentricus sperm. In these, interaction can nevertheless be shown by the fact that the sperm of the latter species can absorb S. franciscanus fertilizin. It may be concluded, then, that the specific- ity of fertilization is based partly on the specificity of the fertilizin-antifertilizin re- action. Since a number of other interactions are undoubtedly involved in the various steps in fertilization it is not surprising to find that specificity is not entirely determined by one of these. On the basis of the present evi- dence the fertilizin-antifertilizin reaction is concerned in the initial attachment of the spermatozoon to the egg. A scheme for the mechanism of such attachment that has been recently proposed (Tyler, ’48a) assumes the same type of interaction that is manifest in antigen-antibody reactions and relates it to the general problem of the mutual adherence of the cells of the various tissues and organs of a multicellular organism. Lillie (19) proposed that fertilizin was also involved in other steps including es- tablishment of the block to polyspermy and activation of the egg through interaction with antifertilizin within the egg. Although the presence of an antifertilizin within the egg has been demonstrated, experimental evi- dence concerning its function is lacking. The role of the lytic agents, such as the egg membrane lysin and hyaluronidase, is manifestly to enable the sperm to penetrate the membrane barriers that surround the unfertilized egg. In this action the hyaluron- idase of mammals shows a rather broad species-specificity while that of the egg- membrane lysin of mollusks is relatively nar- row. No evidence is as yet available to in- dicate whether or not these agents are involved in the further penetration of the surface of the egg proper. For the egg-surface lysin Runnstrém (’49) suggests a role in the establishment of the block to polyspermy. While this is in harmony with its fertiliza- tion inhibiting properties, the availability of this agent under physiological conditions needs to be demonstrated. The work on hyaluronidase stimulated many attempts to apply this agent clinically to overcome sterility in humans. Although there were some early claims of success, recent controlled experiments (Chang, °47; Leonard et al., °47) with rabbits and rats have shown that the addition of hyaluron- idase to inseminates does not enhance fer- tilization and that fertilization can be effected without visible dispersal of the fol- licle cells. These results do not refute the EMBRYOGENESIS: PREPARATORY PHASES above assigned role but rather show that the individual spermatozoon carries sufficient enzyme to make a path for itself through the follicle coat of the egg. REACTION OF THE EGG Cortical Change and Block to Polyspermy. The visible changes that occur at the surface of the egg have been investigated extensively, especially in sea urchins, and most of the recent work is discussed in the review by Runnstrém (’49). Mention may be made here of newer work by Rothschild and Swann (49), who have examined the question of whether or not the rate of propagation of the visible cortical change is sufficient to account for the block to polyspermy. These workers followed cinematographically the darkfield brightening of the surface of the egg that spreads out from the point of contact of the spermatozoon and constitutes the first visible cortical change. They find that the time required for this change to progress to the opposite side of the egg averages 20 seconds at 18° C. in Echinus esculentus. To decide whether or not this is rapid enough to account for the block to polyspermy, estimates are made of the chance of a second sperm striking an un- altered part of the surface. For these calcula- tions measurements are made of the trans- latory speed of swimming of the spermatozoa. The frequency of collision with an egg is then calculated from the kinetic equation, N=ra?nc, in which a is the radius of the egg (50 microns), 7 is the density of sperm sus- pension and ¢ is the mean translatory speed of the spermatozoa (200 microns per second). For sperm densities of 10°, 106 and 107 per milliliter the number of collisions per second would be 0.16, 1.6 and 16, respectively. Since a density of 107 per milliliter does not give any appreciable polyspermy in sea urchin eggs that are in good condition it is concluded either that the block to polyspermy is es- tablished much more rapidly than the ob- served surface change or that only a small fraction (about %6o) of collisions result in fertilization. Further experiments with oocytes, that respond to insemination by for- mation of papillae, each of which is associ- ated with a spermatozoon, show the number of papillae to be considerably less than the number of sperm-oocyte collisions. On the basis of this and other considerations, Roths- child and Swann favor the view that the ob- served cortical change may represent the block to polyspermy and that-many of the GAMETOGENESIS, FERTILIZATION AND PARTHENOGENESIS 195 collisions are ineffective owing to such factors as “muzzling” of the spermatozoa by fertili- zin in solution, necessity for particular orien- tation of sperm head to egg surface, etc. While this analysis involves several assump- tions, of which the authors are well aware, it represents an excellent approach to the solution of this perplexing problem. The present author, too, is inclined to favor the view that the block to polyspermy may be established much more slowly than is gen- erally assumed. Experiments on the so-called enters, in others a hyaline process persists for some time—for example, as long as 15 minutes in Urechis (Tyler, ’31). Of special interest is the formation (Fig. 63), by star- fish eggs, of a long filament originally de- scribed by Fol (1877, 1879) and more re- cently by Chambers (730) and Horstadius (39). The filaments extend from the surface of the egg to spermatozoa in the outer part of the gelatinous coat and their contraction brings the sperm to the surface. According to Horstadius (739) the filament has the Fig. 63. Reproduction of some of Fol’s (’79) figures of fertilization in Asterias glacialis, showing filament connecting the spermatozoon with the surface of the egg. Three successive stages for one egg are illustrated in a, b, and c, and for another egg in d, e, and f. reversal of fertilization in Urechis (Tyler and Schultz, ’32), in which treatment of the eggs with weak acid within as much as two or three minutes after insemination stops development and permits subsequent refertilization, accord with this view. Fertilization Cones. It is well known that the fertilizing spermatozoon becomes motion- less upon attachment to the surface of the egg. Penetration is, then, not effected by mechanical activity of the sperm. In many species of animals the egg elevates, at the point of sperm contact, a cone-shaped process that engulfs the sperm. Examples of various types of entrance cone formation may be found in studies by Chambers (30, ’33) and these, perhaps, suffice to emphasize that the process differs so markedly in different spe- cies that the common element in each is difficult to discern. While in some species the elevation may disappear as the sperm form of a hollow cylinder, and Tyler (48a) interprets it as a precipitate resulting from fertilizin-antifertilizin interaction. Membrane Elevation. While the elevation of a membrane upon fertilization is not characteristic of fertilization in all groups of animals, its occurrence is a manifestation of important surface changes that occur upon fertilization. Recent studies of this process have been largely confined to echinoderms and are reviewed by Runnstrém (49) who, with his co-workers, has contributed exten- sively to the subject. According to these workers the fertilization membrane of the sea urchin egg forms as a result of the cortical granules of the unfertilized egg merging with the vitelline membrane (Fig. 64). Its elevation is attributed to the osmotic pressure of colloids below it. Since the fertilization membrane can be removed shortly after fertilization without 196 affecting subsequent development it evidently has no special significance for the later events. Even when its elevation is inhibited, normal development can ensue (Tyler, ’37; Tyler and Scheer, ’37). If the unfertilized eggs are treated with such agents as isotonic urea or trypsin, that presumably dissolve the vitelline membrane, fertilization and cleavage can occur without membrane forma- tion (Moore, ’30a,b, ’32, 49). Of particular interest in this connection is the report by Hultin (’48a,b) that such treatment renders EMBRYOGENESIS: PREPARATORY PHASES changes in volume of the egg upon fertiliza- tion. There are, however, conspicuous changes in shape in many species. Where such changes occur it is invariably in the direction of greater sphericity. Thus many species of echinoids have ellipsoidal, some- what irregular, eggs that become spherical upon fertilization. Such changes undoubtedly reflect changes in the rigidity of the surface and the viscosity of the egg contents (see Tyler, ’32 and Runnstrém, ’49, for further discussion). cg » VOLE B ue cg 9 TILT Fig. 64. Diagrams of part of surface of the sea urchin egg, illustrating (A) the different layers of the un- fertilized egg and (B to E) steps in the formation of the fertilization membrane, according to the views of Runnstrém (’49). j, Gelatinous coat; vm, vitelline membrane; cg, cortical granules; c, cortex; e, endoplasm; ps, plasma surface; m, fertilization membrane; hl, hyaline layer. (From Runnstrém, ’49.) the eggs more susceptible to cross-fertiliza- tion. The result implies that specificity is, at least, partly controlled by the vitelline membrane. In another species of animal, namely the annelid Nereis, in which experiments on removal of the vitelline membrane have been performed, the denuded eggs do not fertilize (Costello, ’45, ’49). In this species the vitel- line membrane is a thick conspicuous struc- ture through which a large amount of gelatinous material is exuded after fertiliza- tion. When the unfertilized eggs are placed in an alkaline saline solution the jelly swells but does not pass through the membrane, which thereupon stretches and bursts. If the treatment is delayed until shortly after fertilization the membrane can be removed without preventing the attached sperm from entering the egg and without interfering with normal development. Changes in Shape and Volume. In most species of animals there are no marked Changes in Viscosity and Protein Solubility. Along with work on viscosity changes in various kinds of cells, Heilbrunn (728, ’43) and his students have supplied some data concerning fertilization. In general fertiliza- tion results in a marked increase, which Heilbrunn interprets on the basis of his calcium-release, protoplasmic gelation, the- ory of activation. Mirsky (’36) reported that the amount of protein that is soluble in 1 M potassium chloride decreases upon fertiliza- tion in sea urchins. This potassium chloride— soluble protein forms highly viscous solutions which show double refraction of flow and has, therefore, been considered a structural pro- tein like myosin. Connors and Scheer (47) find it to be electrophoretically homogeneous, but, unlike myosin preparations, it possesses no adenosine triphosphatase activity. It is not, as yet, clear how the change in solubility of this material may be correlated with viscosity changes of the intact egg or with the decrease in viscosity upon fertilization re- \ GAMETOGENESIS, FERTILIZATION AND PARTHENOGENESIS 197 ported by Ruffo and Monroy (745) for egg brei. Changes in Permeability. In addition to the early demonstration that fertilization results in an increased permeability to water (R. S. Lillie, 716) and such nonelectrolytes as ethylene glycol (Stewart and Jacobs, 32), recent tracer experiments (Abelson, 747; Brooks and Chambers, ’48; Lindberg, 48) show a great increase in permeability to phosphate. The fate of the phosphate that is incorporated into the fertilized egg will be discussed further below. Changes in Electrical Properties. Activation of eggs has often been considered analogous to stimulation of nerves. However, attempts (Peterfi and Rothschild, ’35; Rothschild, ’38, 46) to detect the propagation of an action potential over the egg surface upon fertiliza- tion in sea urchins and frogs have been un- successful. In fact, no potential difference was found across the surface of the sea urchin egg. While this might be attributable to short-circuiting between the electrodes, the fact that eggs can be fertilized after insertion of the electrode indicates that the fertiliza- tion reaction does not depend upon there be- ing a potential difference across the egg surface. Also, the membrane resistance is ap- parently unchanged upon fertilization in sea urchins and frogs (Cole, ’35; Cole and Spen- cer, 38; Cole and Guttman, ’42). Membrane capacitance is approximately doubled upon fertilization in sea urchins, indicative of change in dielectric constant, but no change is obtained in eggs of Cumingia, Chaetopterus or the frog (Cole and Curtis, ’38; Cole and Guttman, 742). Changes in Metabolism. Detailed surveys of the metabolic changes that occur upon fertilization have been presented by Need- ham (’42) and Brachet (47). The newer outlook in this field is largely due to the work of Whitaker (’31a,b,c, ’33a,b), who refuted the idea that fertilization was in- variably accompanied by a rapid rise in rate of oxidation. In fact a decrease was found to occur in some species, such as the annelid Chaetopterus and the mollusk Cumingia, while in other animals such as the starfish and the annelid Nereis there is no very marked change. In a single species of animal, too, there may be an increase, decrease or no change depending upon the condition of the unfertilized egg, as Tyler and Humason (37) have shown in Urechis. Even in sea urchins, the classic rise, first described by Warburg (’08), does not occur if the eggs are fertilized very soon after removal from the ovary, since, as Borei (748, 749) has shown, the freshly shed unfertilized egg has a respiratory rate as high as that of the fertilized egg. Theories of activation that are dependent upon an increased rate of oxidation are evidently no longer tenable. It has also been held that the oxidative processes differ qualitatively in the fertilized and unfertilized egg. For example, Ruben- stein and Gerard (’34) and Korr (37) have reported large differences in the temperature coefficient of the rate of oxygen uptake of fertilized and unfertilized eggs of sea urchins. On the other hand, Tyler and Huma- son (’37) found no significant differences in sea urchins and other species of animals, and Borei and Lybing (’49) confirm this with sea urchins. This is consistent, too, with the newer findings concerning sensitivity to cyanide and carbon monoxide. Runnstrém (28, ’30a) and Korr (’37) had reported that these agents do not inhibit the respira- tion of unfertilized sea urchin eggs. Along with the initial failure of many investigators to detect cytochrome spectroscopically in sea urchin eggs, this led to the view (Korr, ’37) that before fertilization the respiration is mediated by a nonferrous autoxidizable fer- ment, such as a flavoprotein, while upon fer- tilization the previously unavailable cyto- chrome is thrown into circulation. However, Robbie (746) has shown that if precautions are taken to avoid absorption of the hydrogen cyanide by the alkali well of the respiration vessel, the oxygen uptake of the unfertilized egg can be almost completely inhibited. Also, photo-reversible inhibition of oxygen uptake by carbon monoxide has now been demonstrated by Rothschild (’49a), who ac- counts for the earlier misinterpretations on the basis of a simultaneously occurring light- inhibition of respiration and stimulation of oxygen uptake by carbon monoxide (prob- ably by oxidation of the carbon monoxide). On the basis of these experiments, the demon- stration of the presence of a cytochrome oxidase (Krahl et al., ’41) and the spectro- scopic identification of cytochromes (Roths- child, 49a), it may now be concluded that respiration is mediated by the cytochrome system before as well as after fertilization. In regard to the dehydrogenase part of the respiratory system Runnstrém (30a) and Orstrom (’32) found no change in rate of methylene blue reduction upon fertilization in sea urchins, whereas Ballentine (740), using ferricyanide as the electron acceptor, found an increase. This contradiction still needs to be resolved. Concerning the path- 198 way of carbohydrate breakdown in the sea urchin egg, it has been suggested (Lindberg and Ernster, ’48) that this is through the mechanism known as the hexose monophos- phate shunt (Dickens, ’38), but more re- ent evidence (Ycas, 50; Cleland and Roths- child, *52) shows the presence and opera- tion of the conventional glycolytic mecha- nisms. Sea urchin eggs are known to contain diphosphopyridine nucleotide (Runnstrém, 33; Jandorf and Krahl, ’42) but no change MG. POLYSACCHARIDE PER |00 MG. EGG NITROGEN eo) 1K) 60 EMBRYOGENESIS: PREPARATORY PHASES is inhibited by cyanide. This might operate by interference with the resynthesis of ATP. While there is evidently no over-all change in content of ATP (or other acid-soluble phosphorus compounds) upon fertilization in sea urchins (Runnstrém, ’33; Orstrém and Lindberg, ’40; Lindberg, ’43; Whiteley, ’49), this does not, of course exclude it as an im- portant agent in the activation of the egg. There is evidence (Harvey, 30; Runnstrém, ’°30b; Barron, ’32; Kitching and Moser, 40) that sea urchin eggs can be fertilized or Unfertilized Fertilized 120 180 MINUTES AFTER FERTILIZATION Fig. 65. Change in content of a glycogen-like polysaccharide (estimated as glucose) after fertilization in eggs of the sea urchin Paracentrotus lividus (from data of Orstr6m and Lindberg, ’40). in the content of this coenzyme is found upon fertilization. Sea urchin eggs also con- tain an adenosinetriphosphatase (Runnstrém, 33) and its activity has been found (Connors and Scheer, ’47) to be over twice as high in homogenates prepared from fertilized eggs as in those from unfertilized eggs. Since the ATPase activity is increased by calcium this difference might be correlated with the re- lease of bound calcium (Heilbrunn, °43) upon fertilization. The tracer experiments to which reference was made above have also shown that, as the labelled phosphate is taken up by the fertilized egg, increasing radioactivity is found in the adenosinetri- phosphate (ATP) that is prepared from the eggs. Studies of the distribution of the labelled phosphate (Abelson, ’48; Chambers et al., °48; Whiteley, 49) have shown that 96 to 97 per cent is in the trichloroacetic acid— soluble fraction (about a third of which is probably ATP and ADP). It is of further interest that the uptake of labelled phosphate artificially activated under anaerobic condi- tions. It would be of interest to learn whether or not ATP and other possible energy-yield- ing agents decrease under such conditions. Fertilization in sea urchins also results in the temporary (15 to 20 minutes) production of an as yet unknown acid (Runnstrém, ’30b, 33; Orstrém, °35; Borei, °33; Laser and Rothschild, *39) and in a temporary (10 minutes) increased production of ammonia (Orstrém, 41). The exact significance of this for fertilization is unknown. Orstrém has also reported that fertilized eggs can produce glutamine when ammonia and glumatic acid are added, whereas unfertilized eggs lack this ability. This is interpreted as reflecting a general lack of synthetic ability on the part of the unfertilized, in contrast to the fertilized, egg. From the work of Orstrém and Lindberg (40) there appears to be a considerable breakdown of a polysaccharide (that they designate as, but which may not be, glyco- GAMETOGENESIS, FERTILIZATION AND PARTHENOGENESIS 199 gen) upon fertilization in the sea urchin Paracentrotus lividus. Expressed as glucose the breakdown in the first 10 minutes after fertilization averages 4.7 mg. per 100 mg. of ers muimozen. (— 2.8 cc. of egzss— 5 x 108 eggs = 0.7 gm. dry weight of eggs), whereas during the next 3 hours it is less than 3 per cent of that value (see Fig. 65). If completely oxidized this breakdown would be more than ample to account for the oxygen uptake fol- lowing fertilization. The available data on FUSION OF PRONUCLEI Very little is known concerning the factors responsible for the union of the pronuclei within the egg. Here again attractive in- fluences have been proposed. However, it is known that the egg nucleus in artificially activated eggs can migrate in the same manner as if it were fertilized. Similarly, upon fertilization of non-nucleated eggs the sperm nucleus attains a normal position for Fig. 66. Behavior of pronuclei in centrifuged eggs of the sand dollar Dendraster. A, Two successive stages (at interval of 144 hour) of an egg in which a sperm entered at centripetal end, fused with egg pronucleus and the fusion nucleus moved into large centrifugal part of the egg. B, Two successive stages (at interval of 2 hours) of an egg in which a sperm entered at the centrifugal end, failed to unite with egg pronucleus and proceeded to segment the centrifugal end. e, Egg pronucleus; s, sperm pronucleus; es, fusion nucleus; o, oil cap. (After Moore, 737.) the respiratory quotients, however, show a decrease from about unity in the unfertilized sea urchin egg (Borei, ’33) to values vari- ously reported as ranging from 0.64 to 0.85 (Ephrussi, *33; Borei, ’33; Laser and Roths- child, *39; Ohman, *40) after fertilization. From analyses of lipids Ohman (44) sug- gests that these represent the chief energy source, whereas from analysis of ammonia liberation Hutchens et al. (42) suggest oxi- dation of protein. These various investiga- tions have evidently all been carefully per- formed, but it is evidently still necesary to determine to what extent the divergent re- sults are due to technical difficulties and to species differences. Most investigators, how- ever, are in accord with the view that differ- ent substrates are utilized before and after fertilization (see Brachet, °47, for further details). cleavage of the cell. The fact that egg and sperm pronuclei may meet at some distance from the ultimate position of the fusion nucleus does not imply attraction, since it has not been shown that they are diverted from the path that they would take inde- pendently. Wilson (25) and Morgan (’27) have reviewed the early work on this subject and little has been added since that time. Moore (737) has studied the movements of the pronuclei in echinoid eggs that had been drawn out into flask-shaped form as a result of high speed centrifugation. He found that when the sperm entered the end of the egg containing the egg nucleus it fused with the latter and the fusion nucleus moved through the “neck” into the larger mass of cytoplasm (Fig. 66A). When the sperm entered the opposite end (Fig. 66B) it re- mained there and proceeded to segment that 200 end of the egg while the part with the egg nucleus remained undivided (as is to be expected from the early experiments of Ziegler and others showing that fragments of fertilized eggs containing the egg nucleus O1ON(25 76) 3:30 EMBRYOGENESIS: PREPARATORY PHASES and that cannot operate normally when the pronuclei are separated by too great a dis- tance within the egg. However, in the ab- sence of demonstration of such contractile processes many alternative interpretations 1:30 2:30 4:00 Fig. 67. A, Diagrams illustrating behavior of the pronuclei in physiological polyspermy in the salamander Triton. Sequence is indicated by time (hours and minutes) after insemination. All four sperm that enter develop: normally until 2:30, at which time the accessory asters attain their maximum size, and one sperm nucleus makes contact with the egg nucleus. At 3:00, when sperm and egg nuclei fuse, the accessory sperm nucleus nearest the fusion nucleus shows signs of degeneration. At 3:30 to 4:00 all of the acces- sory sperm nuclei progressively degenerate and the one in the animal hemisphere is pushed out of that region by the cleavage amphiaster. (From Fankhauser, *48.) B, Diagrams illustrating constriction of a salamander egg by means of a hair loop. The unpigmented area represents the location of the second polar spindle. The three darkly pigmented spots mark the entrance points of the spermatozoa. In the constricted egg both fragments can develop into embryos, although nor- mally the spermatozoon in the left fragment, upon failure to fuse with the egg nucleus, would degenerate. (From Fankhauser, 34.) fail to divide while those with the sperm nucleus do). Moore interprets the results of these and similar experiments on the basis of contractile processes that pull the separate and the fused nuclei to their proper positions may be suggested. For example, the develop- ing sperm aster could well be responsible for failure of the sperm pronucleus to traverse the narrow neck of the centrifuged egg. Studies of physiological polyspermy in GAMETOGENESIS, FERTILIZATION AND PARTHENOGENESIS 201 urodeles, that have been undertaken prin- cipally by Fankhauser (see ’48 for references and review), have contributed important information concerning the factors that in- fluence pronuclear behavior. In urodeles all of the sperm that enter undergo the normal pronuclear changes during the first few hours. However, after one of the sperm pro- nuclei has fused with the egg pronucleus the remainder begin to degenerate, those nearest the fusion nucleus regressing first (Fig. 67A). If the egg is constricted after fertilization so that one fragment contains a sperm pronucleus (see Fig. 678), this can proceed to form a spindle, and the fragment can undergo cleavage and normal develop- ment, as earlier work had also shown. How- ever, if the constriction leaves a rather broad connecting neck between the two parts, the development of the accessory sperm pro- nucleus is inhibited. These and other ob- servations establish that some inhibitory influence emanates from the fusion nucleus. This could involve either the production of an inhibitory agent or the removal of some agent essential for the division of the acces- sory cytaster and the nuclear transformations. It could also involve mechanical factors, since the enlarging amphiaster displaces the accessory sperm pronuclei towards regions of the egg where conditions may not be proper for their further development. ARTIFICIAL PARTHENOGENESIS Relatively little work in this field has appeared since the last general review of the subject (Tyler, ’41). Detailed accounts of earlier work are available in texts by Loeb (13), Lillie (19), Morgan (’27) and Dalcq (28b). Summaries of methods employed in the artificial activation of eggs of various ani- mals have been given by Harvey (710) and Runnstr6m (728). In this section a_ brief summary is presented of some of the features of artificial parthenogenesis previously dis- cussed (Tyler, 41), along with a short ac- count of recent views on the problem of activation. SUMMARY OF SOME GENERAL FEATURES OF ARTIFICIAL PARTHENOGENESIS Advanced Stages. It has been established that artificially activated eggs of many ani- mals (sea urchins, starfish, moths, fish, frogs, rabbits) can be reared to the adult condition and sexual maturity. Normal early develop- ment is obtainable in practically all the major groups of animals and it is clear that when normal embryos are produced, obtain- ing the adults is mainly a matter of appro- priate culture methods. Frequency of Normal Development. There have been only occasional reports of experi- ments in which all of the eggs respond with normal development to a particular activat- ing treatment. In general the percentage of normal development is quite low even when a particular treatment initiates development in all of the eggs in a manner indistinguish- able from that induced by the sperm. This is attributable to various tactors, such as irregularities in distribution of chromosomes, haploidy, lack of a proper division mecha- nism (normally supplied by the central body of the sperm) and, in some instances, to fail- ure to establish a plane of bilateral symmetry (determined in some species by the entrance point of the sperm). Origin of Cleavage Amphiaster. In sea urchins cleavage amphiasters are evidently derived from cytasters that are induced by the artificial treatment. In various annelids and mollusks cytasters are not in general formed. Here the amphiaster arises as a result of suppressed polar body divisions and con- tinued cleavage involves a renewal of the capacity for division on the part of the central bodies of the egg and polar bodies. In polar body suppression, (a) the presump- tive polar body spindle may be converted into a cleavage spindle, (6) submerged polar divisions (without cytoplasmic division) may occur followed by fusion of nuclei and as- sociation of central bodies. In the latter type of suppression an amphiaster, triaster or tetraster may form depending upon whether only the first, the first and one of the second, or the first and both second polar divisions are submerged, and the first cleavage then correspondingly results in two, three or four cells (see Fig. 68). Activation of Non-nucleate Egg Fragments. The formation of cytasters capable of multi- plication can be induced in non-nucleate fragments of sea urchin eggs. Such fragments may undergo repeated cleavages. This im- portant work of E. B. Harvey acquires further interest in that it may indicate the possibility of de novo formation of self-duplicating bodies in the cytoplasm. Chromosome Numbers. Artificially acti- vated eggs do not invariably develop as haploids. In fact those that have been reared to advanced stages have the diploid chromo- some number, whereas the haploids generally 202 die in an embryonic stage. Regulation to diploidy is found to occur in a number of ways depending on the species of animal, the stage of maturation division of the egg at the time of treatment, and the method of treatment. Known methods of regulation in- clude fusion of haploid nuclei at the first or a later cleavage, fusion of egg nucleus with a polar body nucleus, and utilization of a polar spindle for cleavage. EMBRYOGENESIS: PREPARATORY PHASES by different methods it is not known to what extent these represent differences in effective- ness rather than failure to find the proper exposure. Eggs of different groups of animals frequently differ in respect to the kind of agent to which they will best respond. As yet no single generally effective method of treatment has been described, nor have any correlations been found between the effective treatments for eggs of various species and ~My eS: cde Ng . OO { | (1 ~N Fig. 68. Diagram of the various types of behavior of the polar divisions and cleavage in Urechis eggs acti- vated with ammoniacal sea water. Upper figure represents the unfertilized egg. A, B, C1, C2 and C3 represent the types obtained with increasing time of treatment (e.g., with 0.01 N NHs3 at 20° C., 4-1 min. > type A; 1-10 min. > type B; 10-20 min. > types C1, C2 and C3). Row J shows condition of egg at time when nor- mally fertilized egg has extruded first polar body; row JI, after time of second polar division; row J//, after first cleavage time. The small broken circles represent nuclei within the egg; the other circles represent ex- truded polar bodies. Eggs IJ/C1, I1IC2 and IIIC3 are figured in polar view, the others in side view. The type B egg, in which polar body formation is normal, fails to undergo cleavage. The other types con- tinue to cleave and can develop into embryos of which many are normal. (After Tyler, ’41.) Sex. In parthenogenetic frogs, silkworms and rabbits the sex is found to accord, in general, with the expectations on the basis of whether the species is male or female digametic and on the method of regulation to diploidy. Methods of Activation. It is now well known that a large variety of chemical and physical agents may be used to activate eggs of vari- ous animals. Even in a single species of animals (e.g., the sea urchin Arbacia) the eggs can respond to hypertonicity, hypoton- icity, acids, bases, neutral salt solutions, certain alkaloids, fat solvents, heat, cold, puncture with a fine needle, ultraviolet radi- ation and radium emanations. Although there are differences in the results obtained properties or characteristics of the eggs. As a result of the early discovery of the diversity of methods that may be employed to activate the egg the problem of activation has come to be regarded as part of the general problem of cell stimulation. Theories of activation have, then, centered about attempts to find the common factor in the action of various parthenogenetic agents. CURRENT THEORIES OF ACTIVATION The basis for Loeb’s (713) concept, that activation resulted from the operation of a cytolytic and a “corrective” factor, was largely removed by Just’s (’22, ’30) demon- stration that the “corrective” hypertonic GAMETOGENESIS, FERTILIZATION AND PARTHENOGENESIS 203 solution alone is effective. Also, Whitaker’s (733) demonstration that activation could occur with a decrease in respiratory rate re- moved any general validity from that aspect of Loeb’s theory that involved the stimula- tion of oxidative processes. No newer investi- gations offer support for Loeb’s views or any recognizable modification thereof. R. S. Lillie (41) proposed what may be termed a bimolecular theory of activation in which it is assumed that an activating substance (A) is formed from the union of two substances (B and S) that may be in- itially present in low concentration in the egg. Substance B is considered to be a prod- uct of hydrolytic processes in the egg and its formation to be stimulated by the action of such agents as acids and heat, which can act under anaerobic conditions. Substance S is considered to be formed from synthetic processes that may be stimulated by agents such as hypertonicity acting only in aerobic conditions. Since some of each is initially present the threshold concentration of A can be reached by increase in either B or S. This theory provides a consistent formal interpre- tation of a great mass of data that Lillie has accumulated, mainly on starfish eggs. Thus it accounts for additive and for sensitizing effects when such agents as heat, acids, hypertonicity and anaerobiosis are employed in various combinations. In its general form it is consistent, too, with other views, such as the proposal that release of calcium is an important factor in activation. The significance of calcium in activation has been emphasized in two rather inde- pendently developed theories, namely the “sensitization to calcium” theory of Dalcq, Pasteels and Brachet (’36; see also Brachet, ’47; Pasteels, ’38a,b, ’41; Pasteels and Fautrez, ’47) and the “calcium release—protoplasmic clotting” theory of Heilbrunn (’30-43). These are based on experiments showing that isotonic calcium chloride solutions or cal- cium-rich solutions are effective activators in many species of animals, that activation by other agents is dependent upon the pres- ence of calcium, and that depriving eggs of calcium will sensitize them to the subse- quent action of calcium solutions. Further support for the general view may be found in experiments by Scheer and Scheer (’47) and by Runnstrém and Monné (45a, b). Heilbrunn’s theory was developed for the phenomenon of stimulation of cells in gen- eral. In somewhat specific form it proposes that the stimulating agent releases calcium, from a calcium proteinate in the cell cortex, and the free calcium initiates a protoplasmic clotting somewhat like blood clotting. It is of interest in this connection that the fertili- zin-antifertilizin interaction has been shown to be largely dependent upon the presence of calcium (Loeb, 713; Vasseur, ’49) and that fertilizin possesses some heparin-like activity (Immers, ’49). However, there is, as yet, no direct evidence indicating that fertilizin is concerned in the activation steps in the processes of fertilization. There seems to be no question that calcium is intimately in- volved in the activation process and one might speculate on its role on the basis of its action on various enzymes, such as ATPase. However, the details of the picture are still vague. There are very likely a large number of substances concerned in the initial changes of activation and it is not known which constitute the primary reactants or whether activation might be initiated by various pathways. While the discovery of artificial partheno- genesis did not bring the realization of the early hopes that problems of fertilization would be readily solved, it has provided im- portant information concerning various as- pects of the problem. It has, also, greatly enlarged the scope of the attack on the prob- lem of activation by substituting relatively simple chemical and physical agents for the spermatozoon. Further, since activation of the egg involves changes in the activity of various enzyme systems, artificial partheno- genesis may come to be a useful tool for the general problem of the factors responsible for regulation of the enzymatic activities of cells. REFERENCES Abelson, P. H. 1947 Permeability of eggs of Arbacia punctulata to radioactive phosphorus. Biol. Bull., 93:203. 1948 Studies of the chemical form of P3? after entry into the Arbacia egg. Biol. Bull., 95: 262. Aboim, A. N. 1945 Développement embryon- naire et post-embryonnaire des gonades normales et agamétiques de Drosophila melanogaster. Rev. Suisse Zool., 52:53-154. Allen, B. M. 1907 An important period in the history of the sex-cells of Rana pipiens. Anat. Anz., 31:339-347. Ball, E. G., and Meyerhof, B. 1940 On the occur- rence of iron-porphyrin compounds and succinic dehydrogenase in marine organisms possessing the copper blood pigment hemocyanin. J. Biol. Chem., 134:483-493. Ballentine, R. 1940 Analysis of the changes in respiratory activity accompanying the fertiliza- 204 tion of marine eggs. J. Cell. Comp. Physiol.; 75: 217-232. ° Barron, E. S.G. 1932 The effect of anaérobiosis on the eggs and sperm of sea urchin, starfish and Nereis and fertilization under anaérobic condi- tions. Biol. Bull., 62:46—53. Bauer, H. 1932 Die Feulgensche Nuklealfar- bung in ihrer Anwendung auf cytologische Un- tersuchungen. Z. Zellforsch. u. mikroskop. Anat., 15:225-247, 1933. Die wachsenden Oocytenkerne einiger Insekten in ihrem Verhalten zur Nukleal- farbung. Z. Zellforsch. u. mikroskop. Anat., 18: 254-298. Beams, H. W., and King, R. L. 1938 A study of the cytoplasmic components and inclusions of the developing guinea pig egg. Cytologia, 8:353- 367. Benoit, J. 1930 Contribution a l'étude de la lig- née germinale chez le Poulet. Destruction pré- coce des gonocytes primaires par les rayons ultraviolets. Compt. rend. soc. biol., 704:1329- 1331. Berrill, N. J., and Liu, C. K. 1948 Germplasm, Weismann and hydrozoa. Quart. Rev. Biol., 23: 124-132. Berry, R. O. 1941 Chromosome behavior in the germ cells and development of the gonads in Sciara ocellaris. J. Morphol., 68:547-583. Bielig, H. J., and Medem, F. 1949 Wirkstoffe der tierischen Befruchtung. Experientia, 5:11-30. Borei, H. 1933 Beitrage zur Kenntnis der Vor- gange bei der Befruchtung des Echinodermeneies. Z. vergleich. Physiol., 20:258-266. 1948 Respiration of oocytes, unfertilized eggs and fertilized eggs from Psammechinus and Asterias. Biol. Bull., 95:124-150. 1949 Independence of post-fertilization respiration in the sea urchin egg from the level of respiration before fertilization. Biol. Bull., 96: 117-122. , and Lybing, S. 1949 Temperature co- efficients of respiration in Psammechinus eggs. Biol. Bull., 96:107-116. Bounoure, L. 1935a Une preuve expérimentale du réle du déterminant germinal chez la Gre- nouille rousse. Compt. rend. acad. sci., 207:1223- 1295" 1935b Sur la possibilité de réaliser une castration dans ]’oeuf chez la Grenouille rousse; résultats anatomiques (avec projections). Compt. rend. soc. biol., 720:1316-1319. 1937a Le sort de la lignée germinale chez la Grenouille rousse aprés l’action des rayons ultraviolets sur le péle inférieur de l’oeuf. Compt. rend. acad. sci., 204:1837-1839. 1937b La constitution des glandes géni- tales chez la Grenouille rousse aprés destruction de la lignée germinale par l’action des rayons ul- traviolets sur l’oeuf. Compt. rend. acad. sci., 204: 1957-1959. 1937c Le déterminant germinal est-il bien en cause dans |’atrophie des gonades conséc- utive a l’action des rayons ultraviolets sur le pdle inférieur de l’oeuf de Grenouille? Compt. rend. soc. biol., 725:895-897. ' Clowes, G. H. A., and Bachman, E. 1921 EMBRYOGENESIS: PREPARATORY PHASES 1939 L’Origine des Cellules Reproduc- trices et le Probleme de la Lignée Germinale. Gauthier-Villars, Paris. Boveri, T. 1887 Ueber Differenzierung der Zell- kerne wahrend der Furchung des Eies von As- caris megalocephala. Anat. Anz., 2:688-693. 1899 Die Entwickelung von Ascaris megalocephala mit besonderer Riicksicht auf die Kernverhaltnisse. Gustav Fischer, Jena. 1904 Protoplasmadifferenzierung als auslésender Faktor fiir Kernverschiedenheit. Sit- zungs-Berichte der Physikalisch-Medicinischen Gesellschaft zu Wiirzburg, pp. 1-5. 1910a Ueber die Teilung centrifugierter Eier von Ascaris megalocephala. Roux’ Arch. Entw.-mech. 30:101-125. 1910b Die Potenzen der Ascaris-Blasto- meren bei abgeanderter Furchung, zugleich ein Beitrag zur Frage qualitativ ungleicher Chromo- somen-Teilung. Festschrift zum sechzigsten Ge- burgstag Richard Hertwigs, 3:131-214. Brachet, J. 1947 Embryologie Chimique. Mas- son et Cie., Paris. Bretschneider, L. H. 1946 Odgenesis; in Experi- mental Embryology in the Netherlands, 1940- 1945 (Woerdeman and Raven, editors). Elsevier, Amsterdam. Brooks, S. C., and Chambers, E.L. 1948 Penetra- tion of radioactive phosphate into the eggs of Strongylocentrotus purpuratus, S. franciscanus and Urechis caupo. Biol. Bull., 95:262-263. Caspersson, T., and Schultz, J. 1938 Nucleic acid metabolism of the chromosomes in relation to gene reproduction. Nature, 742:294-295. , and Schultz, J. 1940 Ribonucleic acids in both nucleus and cytoplasm, and the function of the nucleolus. Proc. Nat. Acad. Sci., 20:507- 515. Chaikoff, I. L., and Entenman, C. 1946 The lipids of blood, liver and egg yolk of the turtle. J. Biol. Chem., 166:683-689. Chambers, E. L., Whiteley, A., Chambers, R., and Brooks, S. C. 1948 Distribution of radioactive phosphate in the eggs of the sea urchin, Lytechi- nus pictus. Biol. Bull., 95:263. Chambers, R. 1930 The manner of sperm entry in the starfish egg. Biol. Bull., 58:344-369. 1933. The manner of sperm entry in various marine ova. J. Exp. Biol., 10:130-141. Chang,M.C. 1947 Effects of testis hyaluronidase and seminal fluids on the fertilizing capacity of rabbit spermatozoa. Proc. Soc. Exp. Biol. & Med., 66:51-54. Chargaff, E. 1942 The formation of the phos- phorus compounds in egg yolk. J. Biol. Chem., 142:505-512. Cleland, K. W., and Rothschild, Lord 1952 The metabolism of the sea-urchin egg. J. Exp. Biol.. 29:285-294. A vola- tile sperm-stimulating substance derived from marine eggs. Proc. Soc. Exp. Biol. & Med., 78: 120-121. Cole, K.S. 1935 Electric impedance of Hipponoé eggs. J. Gen. Physiol., 18:877-887. . and Curtis, H. J. 1938 Electric imped- GAMETOGENESIS, FERTILIZATION AND PARTHENOGENESIS 205 ance of single marine eggs. J. Gen. Physiol., 27: 591-599. Cole, K. S., and Guttman, R. M. 1942 Electric im- pedance of the frog egg. J. Gen. Physiol., 25:765- 115. , and Spencer, J. M. 1938 Electric imped- ance of fertilized Arbacia egg suspensions. J. Gen. Physiol., 27:583-590. Connors, W. M., and Scheer, B. T. 1947 Adeno- sinetriphosphatase in the sea urchin egg. J. Cell. & Comp. Physiol., 30:271-284. Cooper, K. W. 1939 The nuclear cytology of the grass mite, Pediculopsis graminum (Reut.), with special reference to karyomerokinesis. Chromo- soma, 7:51-103. 1941 Visibility of the primary spindle fibres and the course of mitosis in the living blastomeres of the mite, Pediculopsis graminum (Reut.). Proc. Nat. Acad. Sci., 27:480-483. 1944. Analysis of meiotic pairing in Ol- fersia and consideration of the reciprocal chias- mata hypothesis of sex chromosome conjunction in male Drosophila. Genetics, 29:537-568. 1948 The evidence for long range specific attractive forces during the somatic pairing of dipteran chromosomes. J. Exp. Zool., 108:327- 336. 1949 The cytogenetics of meiosis in Dro- sophila: mitotic and meiotic autosomal chiasmata without crossing over in the male. J. Morph., 84: 81-122. Cornman, I. 1941 Sperm activation by Arbacia egg extracts, with special relation to echino- chrome. Biol. Bull., 80:202-207. Costello, D. P. 1945 Experimental studies of germinal localization in Nereis. I. The develop- ment of isolated blastomeres. J. Exp. Zool., 100: 19-66. 1949 The relations of the plasma mem- brane, vitelline membrane and jelly in the egg of Nereis limbata. J. Gen. Physiol., 32:351-366. Dalcq, A. 1928a Le réle du calcium et du potas- sium dans l’entrée en maturation de loeuf de pholade (Barnea candida). Protoplasma, 4:18-44. 1928b Les Bases Physiologiques de la Fécondation et de la Parthénogénése. Presses Uni- versitaires, Paris. , Pasteels, J., and Brachet, J. 1936 Don- nées nouvelles (Asterias glacialis, Phascolion strombi, Rana fusca), et considérations théoriques sur l’inertie de Voeuf vierge. Mém. Mus. Hist. nat. Belg., 2 Série, 3:881—912. Dantschakoff, V. 1941 Der Aufbau des Ge- schlechts beim Hoheren Wirbeltier. Gustav Fischer, Jena. Dantschakoff, W., Dantschakoff, W., Jr., and Bere- skina, L. 1931 Keimzelle und Gonade. I A. Identitat der Urkeimzellen und der entodermalen Wanderzellen; experimentelle Beweise. Z. Zell- forsch. u. mikroskop. Anat., 74:323-375. Darlington, C. D. 1931 Meiosis. Biol. Rev., 6: 221-264: 1937 Recent Advances in Cytology. 2d ed. P. Blakiston’s Son and Co., Inc., Philadelphia. de Burgh, P. M., Sanger, R. A., and Walsh, R. J. 1946 Some aspects of the immunology of the Rh factor. Australian J. Exp. Biol. & Med., 24:293- 300. Diamond, L. K., and Denton, R. L. 1945 Rh ag- glutination in various media with particular ref- erence to value of albumin. J. Lab. & Clin. Med., 30:821-830. Dickens, F. 1938a Oxidation of phosphohexonate and pentose phosphoric acids by yeast enzymes. Biochem. J., 32:1626-44. 1938b Yeast fermentation of pentose phosphoric acids. Biochem. J., 32:1645-53. Domm, L. V. 1929 Spermatogenesis following early ovariotomy in the brown leghorn fowl. Roux’ Arch. Entw.-mech., 779:171-187. Duran-Reynals, F. 1942 Tissue permeability and the spreading factors in injection. Bact. Rev., 6: 197-252. Elster, H. J. 1935 Experimentelle Beitrage zur Kenntnis der Physiologie der Befruchtung bei Echinoideen. Roux’ Arch. Entw.-mech., 133:1-87. Entenman, C., Ruben, S., Perlman, I., Lorenz, F. W., and Chaikoff, I. L. 1938 Radioactive phos- phorus as an indicator of phospholipid metab- olism. III. The conversion of phosphate to lipoid phosphorus by the tissues of the laying and non- laying bird. J. Biol. Chem., 724:795-802. Ephrussi, B. 1933 Contribution a l’analyse des premiers stades du développement de loeuf: ac- tion de la température. Arch. biol., 44:1-147. Evans, T. C., Beams, H. W., and Smith, M.E. 1941 Effects of roentgen radiation on the jelly of the Arbacia egg. Biol. Bull., 80:363-370. Everett, N. B. 1945 The present status of the germ-cell problem in vertebrates. Biol. Rev. 20: 45-55. Fabergé, A. C. 1942 Homologous chromosome pairing: the physical problem. J. Genetics, 43: 121-144. Fankhauser, G. 1934 Cytological studies on the egg fragments of the salamander Triton. III. The early development of the sperm nuclei in egg fragments without the egg nucleus. J. Exp. Zool., 67:159-215. 1948 The organization of the amphibian egg during fertilization and cleavage. Ann. New York Acad. Sci., 49:684-708. Farris, E.J. 1949 The number of motile sperma- tozoa as an index of fertility in man: a study of 406 semen specimens. J. Urology, 67:1099-1104. Fekete, E., and Duran-Reynals, F. 1943 Hyalu- ronidase in the fertilization of mammalian ova. Proc. Soc. Exp. Biol. & Med., 52:119-121. Fisk, R. T., and Morrow, P. 1945 The occurrence of anti-Rh, blocking antibodies in anti-Rh’ serums. Proc. Soc. Exp. Biol. & Med., 58:72-73. Fogg, L. C. 1930 A study of chromatin diminu- tion in Ascaris and Ephestia. J. Morph., 50:413- 451. Fol, H. 1877 Sur le premier développement d’une étoile de mer. Compt. rend. acad. sci., 84: 357-360. 1879 Recherches sur la fécondation et la commencement de V’hénogénie chez divers ani- maux. Mém. Soc. Phys. et d’hist. nat. Genéve, 26: 89-397. Frank, J. A. 1939 Some properties of sperm ex- 206 tracts and their relationship to the fertilization reaction in Arbacia punctulata. Biol. Bull., 76: 190-216. Friedrich-Freksa, H. 1940 Bei der Chromoso- menkonjugation wirksame Krafte und ihre Be- deutung fiir die identische Verdoppelung von Nukleoproteinen. Naturwiss., 28:376-379. Geigy, R. 1931 Action de l’ultraviolet sur le péle germinal dans l’oeuf de Drosophila melanogaster (Castration et Mutabilité). Rev. suisse zool., 38: 187-288. , and Aboim, A. N. 1944 Gonadenent- wicklung bei Drosophila nach friihembryonaler Ausschaltung der Geschlechtszellen. Rev. suisse zool., 51:410—-417. Geitler, L. 1938 Chromosomenbau. (Protoplas- ma-Monographien 14.) Borntraeger, Berlin. Gray, J. 1928 The effect of egg secretions on spermatozoa. Brit. J. Exp. Biol., 5:362-365. Hartman, C.G. 1939 Ovulation, fertilization and the transport and viability of eggs and sperma- ~tozoa; in Sex and Internal Secretions. 2d ed. Wil- liams & Wilkins Co., Baltimore, pp. 630-719. Hartmann, M. 1940 Die stofflichen Grundlagen der Befruchtung und Sexualitat im Pflanzen- und Tierreich. I. Die Befruchtungsstoffe (Gamone) der Seeigel. Naturwiss., 28:807-813. , Kuhn, R., Schartau, O., and Wallenfels, K. 1939 Uber die Sexualstoffe der Seeigel. Natur- wiss., 27:433. , Kuhn, R., Schartau, O., and Wallenfels, K. 1940 Uber die Wechselwirkung von Gyno- und Andro-Gamonen bei der Befruchtung der Eier des Seeigels. Naturwiss., 28:144. , Schartau, O., and Wallenfels, K. 1940 Un- tersuchungen iiber die Befruchtungsstoffe der Seeigel, II. Biol. Zentr., 60:398-423. Harvey, E. B. 1930 The effect of lack of oxygen on the sperm and unfertilized eggs of Arbacia ..punctulata, and on fertilization. Biol. Bull., 58: 288-292. Harvey, E. N. 1910 Methods of artificial par- thenogenesis. Biol. Bull., 78:269-280. Harvey, L. A. 1929 The odgenesis of Carcinus moenas Penn with special reference to yolk formation. Trans. Roy. Soc. Edinburgh, 56:157- 174. Hayashi, T. 1945 Dilution medium and survival of the spermatozoa of Arbacia punctulata. I. Ef- fect of the medium on fertilizing power. Biol. Bull., 89:162-179. Hegner, R. W. 1914 The Germ Cell Cycle in Animals. The Macmillan Co., New York. Heidelberger, M. 1938 The Chemistry of the Amino Acids and Proteins. Charles C Thomas, Springfield, Illinois, pp. 953-974. 1939 Chemical aspects of precipitin and agglutinin reactions. Chem. Rev., 24:323-343. Heilbrunn, L. V. 1928 The Colloid Chemistry of Protoplasm. (Protoplasma-Monographien, Vol. I.) Borntraeger, Berlin. 1943 Outline of General Physiology. 2d ed. W. B. Saunders Co., Philadelphia. , and Wilbur, K. M. 1937 Stimulation and nuclear breakdown in the Nereis egg. Biol. Bull., 73:557-564. EMBRYOGENESIS: PREPARATORY PHASES , and Young, R. A. 1930 The action of ultra-violet rays on Arbacia egg protoplasm. Physiol. Zool., 3:330-341. Henle, W., Henle, G., and Chambers, L. A. 1938 Studies on the antigenic structure of some mam- malian spermatozoa. J. Exp. Med., 68:335-352. Herla, V. 1895 Etude des variations de la mitose chez l’Ascaride mégalocéphale. Arch. Biol., 73: 423-520. Hevesy,G. 1947 Some applications of radioactive indicators in turnover studies. Advances in En- zymol., 7:111-214. ,and Hahn, L. 1938 Origin of phosphorus compounds in hens’ eggs. Kgl. Danske Videnska- bernes Selskab. Biol. Medd., 74(2) :3-39. Hibbard, H. 1928 Contribution a Vétude de Vovogénése, de la fécondation et de l’histogénése chez Discoglossus pictus Otth. Arch. biol., 38:251- 326. Hinton, T. 1946 The physical forces involved in somatic pairing in the Diptera. J. Exp. Zool., 102: 237-251. Hirst, G. K. 1942 The quantitative determina- tion of influenza virus and antibodies by means of red cell agglutination. J. Exp. Med., 75:49-64. Hogue, M. J. 1910 Uber die Wirkung der Cen- trifugalkraft auf die Eier von Ascaris megalo- cephala. Roux’ Arch. Entw.-mech., 29:109-145. Horstadius, S. 1939 Uber die Entwicklung von Astropecten araneiacus 1.. Pubbl. Staz. Zool. Napoli, 17:221-312. Hultin, T. 1947a Some physiological effects of basic sperm proteins. Ark. Kemi, Mineral., Geol., 24B, No. 12. 1947b On the question of sperm antifer- tilizin. Estratto Pubbl. Staz. Zool. Napoli, 27:153- 163. 1948a Species specificity in fertilization reaction. I. The role of the vitelline membrane of sea urchin eggs in species specificity. Ark. Zool., 40A, No. 12. 1948b Species specificity in fertilization reaction. II. Influence of certain factors on the cross-fertilization capacity of Arbacia lixula (L.). Ark. Zool., 40, No. 20. Humphrey, R. R. 1927 Extirpation of the pri- mordial germ cells of Amblystoma: its effect upon the development of the gonad. J. Exp. Zool., 49: 363-399. 1928 The developmental potencies of the intermediate mesoderm of Amblystoma when transplanted into ventrolateral sites in other em- bryos: the primordial germ cells of such grafts and their role in the development of a gonad. Anat. Rec., 40:67-90. 1929 The early history of the primordial germ cells in the Urodeles: evidence from experi- mental studies. Anat. Rec., 42:301-303. 1933 The development and sex differen- tiation of the gonad in the wood frog (Rana sylvatica) following extirpation or orthotopic im- plantation of the intermediate segment and ad- jacent mesoderm. J. Exp. Zool., 65:243-264. Huskins, C. L. 1948 Segregation and reduction in somatic tissues: initial observations on Allium cepa. J. Heredity, 39:310-325. GAMETOGENESIS, FERTILIZATION AND PARTHENOGENESIS 207 Hutchens, J. O., Keltch, A. K., Krahl, M. E., and Clowes, G. H. A. 1942 Studies on cell metab- olism and cell division. VI. J. Gen. Physiol., 25: 717-731. Immers, J. 1949 On the heparin-like effect of fertilizin from sea urchin eggs on blood coagula- tion and the modifying action of basic proteins and cephalin on the effect. Ark. Zool., 42A, No. 6. Jandorf, B. J., and Krahl, M. E. 1942 Studies on cell metabolism and cell division. VIII. The diphosphopyridine nucleotide (cozymase) content of eggs of Arbacia punctulata. J. Gen. Physiol., 25:749-754. Jennings, H. S. 1906 Behavior of the Lower Or- ganisms. Columbia University Press, New York. Jorgensen, M. 1913 Zellenstudien I. Morpholo- gische Beitrage zum Problem des Eiwachstums. Arch. Zellforsch., 70:1-126. Just, E. E. 1915a Initiation of development in Nereis. Biol. Bull., 28:1-17. 1915b The experimental analysis of fer- tilization in Platynereis megalops. Biol. Bull., 28: 93-114. 1915c The morphology of normal fertil- ization in Platynereis megalops. J. Morph., 26: 217-232. 1922 Initiation of development in the egg of Arbacia. I. Effect of hypertonic sea water in producing membrane separation, cleavage and top-swimming plutei. Biol. Bull., 43:383-400. 1930 The present status of the fertilizin theory of fertilization. Protoplasma, 70:300-342. King, R. L., and Beams, H. W. 1938 An experi- mental study of chromatin diminution in Ascaris. J. Exp. Zool., 77:425-443. Kitching, J. A., and Moser, F. 1940 Studies on a cortical layer response to stimulating agents in the Arbacia egg. Biol. Bull., 75:80-91. Korr, I. M. 1937 Respiratory mechanisms in the unfertilized and fertilized sea urchin egg: a tem- perature analysis. J. Cell. Comp. Physiol., 720: 461-485. Krahl, M. E., Keltch, A. K., Neubeck, C. E., and Clowes, G. H. A. 1941 Studies on cell metab- olism and cell division. V. Cytochrome oxidase activity in the eggs of Arbacia punctulata. J. Gen. Physiol., 24:597-617. Krauss, M. 1949 A mucin clot reaction with sea urchin fertilizin. Biol. Bull., 96:74-89. Kuhn, R., and Wallenfels, K. 1940 Echino- chrome als prosthetische Gruppen hochmoleku- larer Symplexe in den Eiern von Arbacia pustu- losa. Ber. deut. chem. Ges., 73:458—464. Landauer, W., Pfeiffer, C. A., Gardner, W. U., and Man, E. B. 1939 Hypercalcification, calcemia and lipemia in chickens following administration of estrogens. Proc. Soc. Exp. Biol. & Med., 41:80- 82. Lardy, H. A., and Philips, P.H. 1941a The inter- relation of oxidative and glycolytic processes as sources of energy for bull spermatozoa. Am. J. Physiol., 733:602-609. , and Philips, P. H. 1941b Phospholipids as a source of energy for motility of bull sperma- tozoa. Am. J. Physiol., 134:542-548. , and Philips, P. H. 1941a The effect of certain inhibitors and activators on sperm metab- olism. J. Biol. Chem., 738:195-202. , and Philips, P. H. 1943a Effect of pH and certain electrolytes on the metabolism of ejac- ulated spermatozoa. Am. J. Physiol., 138:741-746. , and Philips, P. H. 1943b Inhibition of sperm respiration and reversibility of the effects of metabolic inhibitors. J. Biol. Chem., 148:333- 344. , and Philips, P. H. 1943c Inhibition of sperm glycolysis and reversibilty of the effects of metabolic inhibitors. J. Biol. Chem., 148:343- 347. , and Philips, P. H. 1943d The effect of thyroxine and dinitrophenol on sperm metab- olism. J. Biol. Chem., 749:177-182. , and Philips, P. H. 1944 Acetate utiliza- tion for maintenance of motility of bull sperma- tozoa. Nature, 753:168-169. , and Philips, P. H. 1945 Studies of fat and carbohydrate oxidation in mammalian sper- matozoa. Arch. Biochem., 6:53-61. Laser, H., and Rothschild, Lord 1939 The metab- olism of the eggs of Psarmmmechinus miliaris dur- ing the fertilization reaction. Proc. Roy. Soc. London, 126:539-557. Laskowski, M. 1938 The gonadotropic hormone and the level of blood phosphorus in the hen. Bio- chem. J., 32:1176-1180. Leonard, S. L., Perlman, P. L., and Kurzrok, R. 1947 Relation between time of fertilization and follicle cell dispersal in rat ova. Proc. Soc. Exp. Biol. & Med., 66:517-518. Levine, P., and Walker, R. K. 1946 On blocking antibody and zone phenomenon in human anti- Rh sera. Science, 703:389-391. Lillie, F. R. 1912 The production of sperm iso- agglutinins by ova. Science, 36:527-530. 1913a Studies of fertilization. V. The behavior of the spermatozoa of Nereis and Ar- bacia with special reference to egg extractives. J. Exp. Zool., 14:515-574. 1913b The mechanism of fertilization. Science, 38:524-528. 1914 Studies of fertilization. VI. The mechanism of fertilization in Arbacia. J. Exp. Zool., 16:523-590. — 1919 Problems of Fertilization. Uni- versity of Chicago Press, Chicago. , and Just, E. E. 1924 Cowdry’s General Cytology. University of Chicago Press, Chicago, pp. 481-536. Lilie, R. S. 1916 Increase of permeability to water following normal and artificial activation in sea urchin eggs. Am. J. Physiol., 40:249-266. 1941 Further experiments on artificial parthenogenesis in starfish eggs, with a review. Physiol. Zool., 14:239-267. Lindberg, O. 1943 Studien iiber das Problem des Kohlehydratabbaus und der Saurebildung bei der Befruchtung des Seeigeleis. Ark. Kemi, Min- eral., Geol., 76A, No. 15. 1948 On the turnover of adenosine triphosphate in the sea urchin egg. Ark. Kemi, Mineral., Geol., 26B, No. 13. , and Ernster, L. 1948 On carbohydrate 208 metabolism in homogenized sea-urchin eggs. Biochim. Biophys. Acta, 2:471-7. Lindegren, C. C., and Bridges, C. B. 1938 Is ag- glutination an explanation for the occurrence and for the chromomere-to-chromomere specificity of synapsis? Science, 87:510-511. Loeb, J. 1913 Artificial Parthenogenesis and Fertilization. University of Chicago Press, Chi- cago. Lorenz, F. W., Perlman, I., and Chaikoff, I. L. 1943 Phosphorus deposition in the egg as meas- ured with radioactive phosphorus. Am. J. Phys- iol., 138:318-327. , and Tyler, A. 1951 Extension of motile life span of spermatozoa of the domestic fowl by amino acids and proteins. Proc. Soc. Exp. Biol. & Med., 78:57-62. MacBride, E. W., and Hewer, H. R. 1931 Zool- ogy; in Recent Advances in Microscopy (Piney, ed.). The Blakiston Co., Philadelphia, pp. 88-123, 140-154. MacLeod, J. 1939 The metabolism of human spermatozoa. Proc. Soc. Exp. Biol. & Med., 42: 153-155. 1941 The metabolism of human sperma- tozoa. Am. J. Physiol., 732:193-201. 1943a The role of oxygen in the metabo- lism and motility of human spermatozoa. Am. J. Physiol., 138:512-518. 1943b The physiology of mammalian semen. Ann. Rev. Physiol., 5:399-412. 1946a Metabolism and motility of hu- man spermatozoa; in The Problem of Fertility (E. T. Engle, ed.). Princeton University Press, Princeton, N. J., pp. 154-168. 1946b The semen specimen: laboratory examination; in Diagnosis in Sterility (E. T. Engle, ed.). Charles C Thomas, Springfield, Il- linois, p. 3. Mann, T. 1945a Studies on the metabolism of semen. Biochem. J., 39:451-458. 1945b Anaerobic metabolism of sperma- tozoa. Nature, 756:80-81. 1946 Fructose, a constituent of semen. Nature, 157:79. 1949 Metabolism of semen. Advances in Enzymol., 9:329-390. Marrack, J. R. 1938 The Chemistry of Antigens and Antibodies. Special Report, Series No. 230, Medical Research Council, London. Marza, V. D. 1938 Histophysiologie de l’Ovo- genese. Actualités Scientifiques et Industrielles, No. 751. Hermann et Cie., Paris. McClean, D., and Rowlands, I. W. 1942 Role of hyaluronidase in fertilization. Nature, 150:627- 628. Medem, F. G. von 1942 Beitrage zur Frage der Befruchtungsstoffe bei marinen Mollusken. Biol. Zentr., 62:431-446. Metz, C.B. 1942a Egg and sperm agglutination in invertebrates. Doctorate thesis, California In- stitute of Technology. 1942b The inactivation by fertilizin and its conversion to the “univalent” form by x-rays and ultraviolet light. Biol. Bull., 82:446-454. EMBRYOGENESIS: PREPARATORY PHASES 1945 The agglutination of starfish sperm by fertilizin. Biol. Bull., 89:84—-94. 1949 Agglutination of sea urchin eggs and sperm by basic proteins. Proc. Soc. Exp. Biol. & Med., 70:422-424. Metz, C. W. 1931 Chromosomal differences be- tween germ cells and soma in Sciara. Biol. Zentr., 51:119-124., 1938 Chromosome behavior, inheritance and sex determination in Sciara. Amer. Nat., 72: 485-520. Meyer, K. 1947 The biological significance of hyaluronic acid and hyaluronidase. Physiol. Rev., 27:335-359. , and Rapport, M. M. 1952 MHyaluroni- dases. Advances in Enzymol., 73:199-236. Mickey, G. H. 1946 The presence of multiple strands in chromosomes of Romalea (Orthoptera). Amer. Nat., 80:446-452. 1947 Division cycle in grasshopper chro- mosomes. Proc. Louisiana Acad. Sci., 10:49-66. Mirsky, A. E. 1936 Protein coagulation as a re- sult of fertilization. Science, 84:333-334. Monroy, A. 1939 Sulla localizzazione delle cell- ule genitali primordiali in fasi precoci di sviluppo. Ricerche sperimentali in anfibi anuri. Arch. ital. Anat. Embriol., 47:368-389. Moore, A. R. 1930a Fertilization and develop- ment without membrane formation in the egg of the sea urchin, Strongylocentrotus purpuratus. Protoplasma, 9:9-17. 1930b Fertilization and development without the fertilization membrane in the egg of Dendraster excentricus. Protoplasma, 9:18-24. 1932 The dependence of cytoplasmic structures in the egg of the sea urchin on the ionic balance of the environment. J. Cell. Comp. Phys- iol., 2:44-61. 1937 On the centering of the nuclei in centrifuged eggs as a result of fertilization and artificial membrane formation. Protoplasma, 27: 544-551. 1949 The relation of ions to the appear- ance and persistence of fertilization and hyaline membranes in eggs of the sea urchin. Amer. Nat., 83:233-247. Morgan, T. H. 1927 Experimental Embryology. Columbia University Press, New York. Narain, D. 1937 Cytoplasmic inclusions in the oogenesis of Sacchobranchus fossiles, Calarias batrachus and Anabas scandens. 7.. Zellforsch., 26:625-640. Needham, J. 1931 Chemical Embryology. Cam- bridge University Press, Cambridge, England, Vol. III, App. II, pp. 1679-1684. 1942 Biochemistry and Morphogenesis. Cambridge University Press, Cambridge, Eng- land. Nieuwkoop, P. D. 1946 Experimental investiga- tions on the origin and determination of the germ cells, and on the development of the lateral plates and germ ridges in urodeles. Arch. Néer- land. Zool., 8:1—205. Ohman, L. O. 1940 Uber die Veranderung des respiratorischen Quotienten wahrend der Friih- GAMETOGENESIS, FERTILIZATION AND PARTHENOGENESIS 209 entwicklung des Seeigeleies. Ark. Zool., 32A, No. 15. Ohman, L.O. 1944 On the lipids of the sea urchin egg. Ark. Zool., 36A, No. 7. Orstr6m, A. 1932 Zur Analyse der Atmungs- steigerung bei der Befruchtung des Seeigeleis. Auf der Grundlage von Versuchen tiber Oxyda- tion und Reduktion von Dimethylparaphenylen- diamin in der Eizelle. Protoplasma, 15:566-589. 1935 Uber Ammoniakbildung bei der Entwicklungserregung des Seeigeleies. Ark. Zool., 28B, No. 6. 1941 Wher die chemischen Vorgange ins- besondere den Ammoniakstoffwechsel, bei der Entwicklungserregung des Seeigeleies. Z. Phys- iol. Chem., 277:1-176. ,and Lindberg, O. 1940 Uhber den Kohlen- hydratstoffwechsel bei der Befruchtung des Seeigeleies. Enzymologia, 8:367—384. Painter, T. S. 1940 On the synthesis of cleavage chromosomes. Proc. Nat. Acad. Sci., 26:95-100. 1945a Nuclear phenomena associated with secretion in certain gland cells with especial reference to the origin of cytoplasmic nucleic acid. J. Exp. Zool., 700:523-541. 1945b Chromatin diminution. Conn. Acad. Arts Sci., 36:443-448. , and Reindorp, E. 1939 Endomitosis in the nurse cells of the ovary of Drosophila melano- gaster. Chromosoma, 7:276-283. Pasteels, J. 1938a Le réle du calcium dans I’ac- tivation de l’oeuf de Pholade. Tray. Stat. Zool. Wimereux, 73:515-530. 1938b Sensibilisateurs et réalisateurs dans l’activation de Voeuf de Barnea candida. Bull. Acad. Belg. Cl. Sci., 24:721-731. 1941 Sur quelques particularités de l’ac- tivation de Voeuf d’oursin (Psammechinus milt- aris). Bull. Cl. Sci. Acad. Roy. Belg., 27:123- 129. Trans. , and Fautrez, J. 1947 Remarques sur les effets de la rupture de la balance des sels sur le cytoplasme ovulaire. (A propos d’observations faites sur Hydroides uncinatus et Ascidiella as- persa). Arch. Portugaies Sci. Biol., 9:42-58. Pauling, L. 1940 The theory of the structure and process of formation of antibodies. J. Am. Chem. Soc., 62:2643-2657. Peterfi, T., and Rothschild, V. 1935 Bio-electric transients during fertilization. Nature, 735:874— 875. Pincus, G. 1930 Observations on the living eggs of the rabbit. Proc. Royal Soc., 707:132-167. 1936 The Eggs of Mammals. The Mac- millan Co., New York. Race, R. R. 1944 “Incomplete” antibody in hu- man serum. Nature, 153:771-772. Reagan, F. P. 1916 Some results and possibilities of early embryonic castration. Anat. Rec., 17: 489-491. Riddle, O. 1942 Cyclic changes in the blood cal- cium, phosphorus and fat in relation to egg lay- ing and estrogen production. Endocrinology, 37: 498-506. Ris, H. 1945 The structure of meiotic chromo- somes in the grasshopper and its bearing on the nature of “chromomeres” and “Jamp-brush chro- mosomes.” Biol. Bull., 89:242-257. Robbie, W. A. 1946 The effect of cyanide on the oxygen consumption and cleavage of the sea urchin egg. J. Cell. Comp. Physiol., 28:305-324. Roepke, R. R., and Bushnell, L. D. 1936 A sero- logical comparison of the phosphoprotein of the serum of the laying hen and the vitellin of the egg yolk. J. Immunol., 30:109-113. Rothschild, Lord 1938 The biophysics of the egg surface of Echinus esculentus during fertilization and cytolysis. J. Exp. Biol., 75:209-216. 1946 Physiology of fertilization. Nature, 157:720-722. 1948a The physiology of Echinus es- culentus spermatozoa. J. Exp. Biol., 25:15-21. 1948b The physiology of sea urchin spermatozoa: lack of movement in semen. J. Exp. Biol., 25:344-352. 1948c The physiology of sea urchin spermatozoa: senescence and the dilution effect. J. Exp. Biol., 25:353-368. 1949a The metabolism of fertilized and unfertilized sea urchin eggs: the action of light and carbon monoxide. J. Exp. Biol., 26:100-111. 1949b The fertilization reaction in the sea urchin egg: a note on diffusion considerations. J. Exp. Biol., 26:177-181. 1951a Sea-urchin Rev., 26:1-27. 1951b Sperm-egg interacting substances and metabolic changes associated with fertiliza- tion. Biochem. Soc. Symposia, 7:40-51. 1952 Spermatozoa. Sci. Progress, 60: spermatozoa. Biol. 1-10. , and Cleland, K. W. 1952 The physiol- ogy of sea-urchin spermatozoa: the nature and location of the endogenous substrate. J. Exp. Biol., 29:66-71. , and Swann, M. M. 1949 The fertiliza- tion reaction in the sea urchin egg: a propagated response to sperm attachment. J. Exp. Biol., 26: 164-176. Rubenstein, B. B., and Gerard, R. W. 1934 Fer- tilization and the temperature coefficients of oxy- gen consumption in eggs of Arbacia punctulata. J. Gen. Physiol., 77:677-686. Ruffo, A., and Monroy, A. 1945 Variazioni di viscosita delle uova di riccio di mare durante la fecondazione. Boll. soc. ital. biol. sper., 20:6-7. Runnstrém, J. 1928 Struktur und Atmung bei der Entwicklungserregung des Seeigeleies. Acta Zool., 9:445-499. 1930a Atmungsmechanismus und Ent- wicklungserregung bei dem Seeigelei. Proto- plasma, 70:106-173. 1930b Spaltung und Atmung bei der Entwicklungserregung des Seeigeleies. Ark. Zool., 21B, No. 8. 1933 Zur Kenntnis der Stoffwechsel- vorgange bei der Entwicklungserregung des Seeigeleies. Biochem. Z.., 258:257-279. 1944 Notes on the formation of the fer- tilization membrane and some other features of 210 the early development of the Asterias egg. Acta Zool., 25:159-167. Runnstrém, J. 1949 The mechanism of fertiliza- tion in metazoa. Advances in Enzymol., 9:241-327. , and Lindvall, S. 1946 The effect of some agents upon the reaction of Echinocardium spermatozoa towards egg water. Ark. Zool., 38A, No. 10. , Lindvall, S., and Tiselius, A. 1944 Ga- mones from the sperm of sea urchin and salmon. Nature, 753:285-286. , and Monné, L. 1945a On some proper- ties of the surface layers of immature and mature sea urchin eggs, especially the changes accom- panying nuclear and cytoplasmic maturation. Ark. Zool., 36A, No. 18. ,and Monné, L. 1945b On changes in the properties of the surface layers of the sea urchin ege due to varying external conditions. Ark. Zool., 36A, No. 20. Monné, L., and Broman, L. 1943 On some properties of the surface layers in the sea urchin egg and their changes upon activation. Ark. Zool., 35A, No. 3. , Tiselius, A., and Lindvall, S. 1945 The action of Androgamone III on the sea urchin egg. Ark. Zoologi, 36A, No. 22. , Tiselius, A., and Vasseur, E. 1942 Zur Kenntnis der Gamonwirkungen bei Psammechi- nus miliaris und Echinocardium cordatum. Ark. Kemi, Mineral., Geol., 15A, No. 16. Sampson, M. M. 1922 Iso-agglutination and hetero-agglutination of spermatozoa. Biol. Bull., 43:267-284. Sax, H. J., and Sax, K. 1935 Chromosome struc- ture and behavior in mitosis and meiosis. J. Arnold Arboretum, 16:423-439. , Scheer, B. T., and Scheer, M. A. R. 1947 Some interrelations of drug and ion actions in the arti- ficial activation of marine eggs. Physiol. Zool., 20:15-32. Schechter, V. 1937 Calcium reduction and the prolongation of life in the egg cells of Arbacia punctulata. Biol. Bull., 72:366-376. 1941 Experimental studies upon the egg cells of the clam, Mactra solidissima, with special reference to longevity. J. Exp. Zool., 86:461-478. Schrader, F. 1935 Notes on the mitotic behay- iour of long chromosomes. Cytologia, 6:422-430. 1944. Mitosis: The Movements of Chro- mosomes in Cell Division. Columbia University Press, New York. Schultz, J. 1941 The evidence of the nucleopro- tein nature of the gene. Cold Spring Harbor Sym- posia on Quant. Biol., 9:55-65. Sharp, L. W. 1943 Fundamentals of Cytology. McGraw-Hill Book Co., New York. Singh, B. N., and Boyle, W. 1938 The vitello- genesis of Gasterosteus aculeatus (the stickle- back) investigated by the ultra-centrifuge. Quart. J. Micr. Sci., 87:81-106. Spikes, J. D. 1948 Experiments on fertilization and cleavage. Doctorate thesis for California In- stitute of Technology. 1949a The prezone phenomenon in sperm agglutination. Biol. Bull., 97:95-99. EMBRYOGENESIS: PREPARATORY PHASES 1949b Metabolism of sea urchin sperm. Amer. Nat., 83:285-301. Stewart, D., and Jacobs, M. H. 1932 The effect of fertilization on the permeability of the eggs of Arbacia and Asterias to ethylene glycol. J. Cell. Comp. Physiol., 7:83-92. Subramaniam, M. K., and Aiyar,G. 1936 Secre- tion of fatty and albuminous yolk by Golgi bodies in Stomopneustes variolaris Lamarck. Z. Zell- forsch. u. mikroskop. Anat., 24:576-584. Swift, C.H. 1914 Origin and early history of the primordial germ cells in the chick. Am. J. Anat., 15:483-516. Taurog, A., Lorenz, F. W., Entenman, C., and Chaikoff, I. L. 1944 The effect of diethylstil- bestrol on the in vitro formation of phospholipids in the liver as measured with radioactive phos- phorus. Endocrinology, 35:483-487. Terni, T., and Maleci, O. 1937 Sulla penetrazi- one di spermatozoi dentro cellule somatiche col- tivate in vitro. Monitore Zool. Ital., 47 (suppl.): 72-79. Tosic, J. 1947 Mechanism of hydrogen peroxide formation by spermatozoa and the role of amino acids in sperm motility. Nature, 159:544. Tosic, J., and Walton, A. 1946 Formation of hydrogen peroxide by spermatozoa and its in- hibitory effect on respiration. Nature, 158:485. Tyler, A. 1931 The production of normal em- bryos by artificial parthenogenesis in the echi- uroid, Urechis. Biol. Bull., 60:187-211. 1932 Changes in volume and surface of Urechis eggs upon fertilization. J. Exp. Zool., 63: 155-173. 1937 On the energetics of differentiation. V. Comparison of the rates of development and of oxygen consumption of tight membrane and normal echinoderm eggs. J. Exp. Zool., 76:395- 406. 1939a Extraction of an egg membrane- lysin from sperm of the giant keyhole limpet (Megathura crenulata). Proc. Nat. Acad. Sci., 25:317-323. 1939b Crystalline echinochrome and spinochrome: their failure to stimulate the respi- ration of eggs and of sperm of Strongylocentrotus. Proc. Nat. Acad. Sci., 25:523-528. 1940a Sperm agglutination in the key- hole limpet, Megathura crenulata. Biol. Bull., 78:159-178. 1940b Agglutination of sea urchin eggs by means of a substance extracted from the eggs. Proc. Nat. Acad. Sci., 26:249-256. 1941 The role of fertilizin in the fertil- ization of eggs of the sea urchin and other ani- mals. Biol. Bull., 87:190-204. 1942 Specific interacting substances of eggs and sperm. Western J. Surg., Obstet. & Gynecol., 50:126-138. 1945a Conversion of agglutinins and precipitins into “univalent” (non-agglutinating or non-precipitating) antibodies by photo-dynam- ic irradiation of rabbit-antisera vs. pneumococci, sheep-red-cells and sea urchin sperm. J. Im- munol., 57:157-172. 1945b Anaphylactic properties of photo- GAMETOGENESIS, FERTILIZATION AND PARTHENOGENESIS 2AA oxidized rabbit-antisera (vs. sheep-erythrocytes and pneumococci) and horse-antiserum (vs. diph- therial toxin) containing “univalent” antibodies. J. Immunol., 57:329-338. Tyler, A. 1946a Loss of fertilizing power of sea urchin and Urechis sperm treated with “unival- ent” antibodies vs. antifertilizin. Proc. Soc. Exp. Biol. & Med., 62:197-199. 1946b On natural auto-antibodies as evi- denced by anti-venin in serum and liver extract of the Gila monster. Proc. Nat. Acad. Sci., 32: 195-201. 1947 An auto-antibody concept of cell structure, growth and differentiation. Growth, 10(suppl.) :7-19. 1948a_ Fertilization and immunity. Phys- iol. Rev., 28:180-219. 1948b The chemistry of the fertilizin of the sea urchin Strongylocentrotus purpuratus. Anat. Rec., 101:8-9. 1949 Properties of fertilizin and related substances of eggs and sperm of marine animals. Amer. Nat., 83:195-219. 1950 Extension of the functional life span of spermatozoa by amino acids and peptides. Biol. Bull., 99:224. 1953 Prolongation of life span of sea- urchin spermatozoa, and improvement of the fer- tilization-reaction, by treatment of spermatozoa and eggs with metal-chelating agents (amino acids, versene, DEDTC, oxine, cupron). Biol. Bull., 104:224-239., , and Atkinson, E. 1950 Prolongation of the fertilizing capacity of sea-urchin spermatozoa by amino acids. Science, 772:783-785. , and Dessel, F. W. 1939 Increasing the life span of unfertilized Urechis eggs by acid. J. Exp. Zool., 82:459-472. , and Fox, S. W. 1939 Sperm agglutina- tion in the keyhole limpet and the sea urchin. Science, 90:516-517. , and Fox, S. W. 1940 Evidence for the protein nature of the sperm agglutinins of the keyhole limpet and the sea urchin. Biol. Bull., 79:153-165. , and Humason, W. D. 1937 On the ener- getics of differentiation, VI. Comparison of the temperature coefficients of the respiratory rates of unfertilized and of fertilized eggs. Biol. Bull., 78:261-279. , and O’Melveny, K. 1941 The role of antifertilizin in the fertilization of sea urchin eggs. Biol. Bull., 87:364-374. , Ricci, N., and Horowitz, N.H. 1938 The respiration and fertilizable life of Arbacia eggs under sterile and non-sterile conditions. J. Exp. Zool., 79:129-143. , and Rothschild, Lord 1951 Metabolism of sea-urchin spermatozoa and induced anaerobic motility in solutions of amino acids. Proc. Soc. Exp. Biol. & Med., 76:52-58. , and Scheer, B. T. 1937 Inhibition of fer- tilization in eggs of marine animals by means of acid. J. Exp. Zool., 75:179-197. , and Schultz, J. 1932 Inhibition and re- versal of fertilization in the eggs of the echiuroid worm, Urechis caupo. J. Exp. Zool., 63:509-532. , and Swingle, S. W. 1945 Protective value of “univalent” antibodies produced by photo-oxidation of antipneumococcal rabbit-se- rum and antidiphtheric horse-serum. J. Im- munol., 57:339-347. , and Tanabe, T. Y. 1952 Motile life of bovine spermatozoa in glycine and yolk-citrate diluents at high and low temperatures. Proc. Soc. Exp. Biol. & Med., 817:367-371. Ubisch, L., von 1943 Uber die Bedeutung der Diminution von Ascaris megalocephala. Acta Biotheoretica, 7:163-182. Vandel, A. 1941 Recherches sur la génétique et la sexualité des isopodes terrestres. VII. Sur la longévité des spermatozoides 4 Vintérieur de Vovaire d’ “Armadillidium vulgare.” Bull. Biol. France et Belgique, 75:364—368. Vasseur, E. 1947 The sulphuric acid content of the egg coat of the sea urchin, Strongylocentrotus droebachiensis Mill. Ark. Kemi, Mineral., Geol., 25B, No. 6. 1948a Chemical studies on the jelly coat of the sea urchin egg. Acta chem. Scandinav., 2: 900-913. 1948b Entod. ~ - ephrotome. ~~~. \ “Mandib. fo: °. Dee ca (* mesod..- ¢ 2 Rehome) ‘ = A (Prechord. plate)’ Notochord a: G Ee a SS AT Pa aC wks ( aan MTA iperrZ ta PF ith Fig. 85. Urodele neurula; the mesoderm mantle is exposed on the right side of the embryo. (Orig- inal.) pronephros areas (Muchmore, 51), has the same effect. Moreover, typical somites are dif- ferentiated after the complete extirpation of the notochord in the early neurula (Kitchin, 49; Horstadius, ’44; Nieuwkoop, ’46; Much- more, 51). In normal development, muscle differentiation is probably brought about by the synergistic action of various neighboring tissues. In addition to pronephric tubules and musculature, notochord and even neural tis- sue can be differentiated from prospective somites (Muchmore, ’51; Lopashov, *35b). Pronephros Area. Explants of this region when taken from Triturus neurulae can differentiate into typical pronephric tubules (Yamada, °40), but when isolated from Amblystoma punctatum they give rise to few and poorly developed tubules (Much- more, *51). This area is capable also of form- ing musculature when isolated in combina- tion with notochord or when grafted to the somite level (Yamada, ’37, 40) but it forms ventral mesoderm and blood islands when grafted to ventral regions (Yamada, ’37; see Fig. 86). Lateral plate and blood islands differentiate in explants largely according to their pros- 248 EMBRYOGENESIS: PROGRESSIVE DIFFERENTIATION pective significance, but when grafted to the somite and pronephros level, they differenti- ate “neighborwise,’ probably as a result of assimilative induction (Yamada, °37, ’40). However, when ventral mesoderm was iso- lated in combination with notochord, it fre- quently gave rise to pronephric tubules and the blood cell formation was suppressed (Yamada, 40; Fernald, ’43; Muchmore, 51; see Fig. 86). Blood isl. Prospective significance This situation has been interpreted in different ways: Modifying the ideas of Dalcq and Pasteels (’37, ’38), Yamada (’40) as- sumes that the qualitative differences in the differentiation of the mesoderm originate from quantitative differentials of the morpho- genetic activity within the mesoderm mantle. Each specific differentiation would be the manifestation of a certain “morphogenetic potential,” and (arbitrarily assigned) thres- Cultured with notochord Fig. 86. Schematic representation of the differentiations of prospective somite, pronephros, and blocd island regions, respectively, of the early neurula. First vertical row: normal fate; second vertical row: differentiation when isolated in ectodermal vesicle; third vertical row: differentiation when combined with notochord, in ectodermal vesicle. (After Yamada, 40.) All these data show that some parts of the mesoderm of the early neurula are endowed with stronger differentiation tendencies to- wards their prospective fate than they were in the early gastrula stage. This applies es- pecially to the prospective notochord. The pronephric and ventrolateral areas, while capable of differentiating selfwise and in- dependently of each other, can still be shifted into other routes of differentiation when ex- posed to the influence of prospective noto- chord or somites. The prospective somite region is particularly dependent on environ- mental factors. Although its tendency to form notochord or neural tissue is clearly reduced as compared to the behavior of explants or transplants from the blastoporal lip of the early gastrula, its transformation into muscle tissue is definitely enhanced by neighboring tissues, among which the notochord plays a predominant role. hold values would determine the lines of segregation of the primordia. The “potential” of the mid-dorsal area is presumably high enough for notochord differentiation, but that of the prospective somite area is too low for muscle differentiation and just high enough for the differentiation of pronephric tubules. Combination with notochord results in a raise of each level by one step: prospec- tive somite and pronephros are raised to muscle, lateral and ventral mesoderm to pronephros differentiation. Muchmore (751) envisages the organiza- tion of the mesoderm mantle in terms of several overlapping morphogenetic fields which are characterized by qualitative rather than quantitative differences. Tissues other than the notochord can also direct meso- dermal primordia into new channels of differentiation, and there is no evidence that these effects are due to quantitative grada- AMPHIBIANS tions of a single agent. The overlap of the pronephros over the somite field would ac- count for the formation of tubules in isolates of the somite area, and a mediad extension of the pronephros field would explain the tubule formation from ventrolateral meso- derm. The somite field seems to require extrinsic agents for its materialization to a higher degree than the other mesodermal primordia. POSTERIOR TRUNK AND TAIL MESODERM Vogt (’26b, ’29) has shown that in urodeles, at the stage of the closed blastopore, only the Ua ale SOMir, && (a) d 249 tion and elongation of the tail bud begins (for details see particularly Chuang, °47). According to the studies of Pasteels (739b, 43), the different structures of the posterior trunk and tail originate from localized tis- sue primordia which undergo morphogenetic movements considered to be a continuation of the gastrulation movements. They are not the result of differential growth processes, since there was no indication of a high or differential mitotic activity. Therefore, am- phibian development does not conform to the concept of Holmdahl (since ’25; see ’39), who ascribes the organogenesis of the caudal end to an outgrowth from an indifferent growth center, and not to the segregation of Fig. 87. Maps and morphogenetic movements of posterior trunk and tail in urodeles. a, Morphogenetic movements during neurulation; b, vital staining of three areas in the posterior medullary plate; c, the fate of the three marks. d, Topography of prospective tail structures in the neural groove stage; e, same in neurula with rising folds. The numbers in d and e indicate somite numbers. (After Chuang, ’47.) first six to eight somites, the adjacent lateral mesoderm, and the greater part of the noto- chord are invaginated. The posterior trunk and tail somites are still on the surface oc- cupying the posterior fifth of the medullary plate. Detailed maps of these regions were constructed on the basis of vital staining and transplantation experiments (Vogt, ’26b, ’29; Bijtel, 31; Nakamura, ’38; Pasteels, ’39b, ’42; Spofford, ’45; Chuang, ’47; see Fig. 87). Dur- ing neurulation, the tail material moves in- ward by way of complicated invagination and folding processes, whereupon the forma- germ layers characteristic of the more ante- rior regions. A distinction between “pri- mary” and “secondary” body formation is not warranted in amphibians. Holmdahl’s notion finds still less support in the results from defect, transplantation and isolation experiments. They have shown that, from early neurula stages on, the different primordia of the tail region are even more rigidly determined and less capable of regu- lation than the more anterior parts of the mesoderm mantle (Bytinski-Salz, °31, °36; Mangold, ’32, ’33b; Bijtel, 36; Von Aufsess, 250 EMBRYOGENESIS: PROGRESSIVE DIFFERENTIATION 41; Chuang, ’47). This restriction of poten- tialities of the caudal primordia is the more remarkable as in larval stages it is the tail and not the trunk which is outstanding for its capacity of regeneration. Hence, one can no longer subscribe to the notion that the determination of the axial organs proceeds along a cephalocaudal gradient. The different tissue primordia of the tail, namely notochord, spinal cord, somites and fin, are each independently capable of axial stretching. But it seems that the cooperation of several, or all, of them is necessary for the formation of a complete and straight tail (Vogt, ’26b; Holtfreter, ’33c; Bytinski-Salz, 36; Chuang, *47; Kitchin, ’49).* REGIONAL INDUCTION CAPACITIES OF THE MESODERM OF THE NEURULA We saw that even in the young gastrula a distinction can be made between head and trunk-tail organizers, although their borders are ill-defined. At the end of gastrulation, a more distinct localization of the various mesodermal inductors along the cephalo- caudal axis is manifested. The axial meso- derm has meanwhile separated into a poste- rior portion (notochord and somites) and an anterior one (prechordal mesoderm). The latter begins at the level of the midbrain, between the first and second visceral arch. It is subdivided into a narrow median strip, the “prechordal plate,” and the lateral ““man- dibular mesoderm” whose paraxial position corresponds to that of the trunk somites (Vogt, ’29; Adelmann, ’32; and others; see Fig. 85). Lehmann (’42b, ’45, 48) has pointed out that this mesodermal structuration is re- flected in differences of inductive capacity. He considers the prechordal mesoderm as a specific inductor for tel- and diencephalon which are designated as “archencephalic” structures, whereas the chordal-parachordal head mesoderm would induce posterior or “deuterencephalic” brain structures. Dalcq (46) prefers the terms “acrencephalic” and “chordencephalic” inductors. No structural difference can be detected between the latter and the spinal cord—inducing trunk portion of the archenteron roof. In order to examine possible regional dif- * This balance between independence and cooper- ation of the components of the tail is nicely demon- strated in L. S. Stone’s time-lapse film of the de- velopment of A. punctatum. The speeded-up pic- tures show an up-and-down wiggling of the elongat- ing tail bud, which seems to reflect alternate spurts of stretching in the ventral and dorsal tail portions, respectively. ferences in the inductive capacity of the archenteron roof, Mangold (’33b) divided the latter into four transverse strips and implanted each under the ectoderm of a gastrula. Grafts of the rostral strip were relatively inactive, possibly because they consisted largely of cephalic entoderm. The second strip induced predominantly anterior head structures (forebrain, eyes, nose, bal- ancers), the third strip mainly posterior head structures (rhombencephalon, otocysts), while the caudal strip induced regularly spinal cord and, frequently, pronephros and tail, but no brain. However, some structures, such as eye or otocysts, could be induced not only by that part of the mesoderm which under- les these structures normally, but by adja- cent regions as well. Similar results were obtained by Ter Horst (’48) in explants, and by Okada and Takaya (42), Okada and Hama (743, ’45) and Hama (’49), who tested the specificity of the cephalocaudal sections of the archenteron roof, before and after their invagination, by implanting them into whole embryos or ectoderm vesicles. These authors confirmed the rather puzzling observation of Mangold (’33b) that the anteriormost part of the archenteron was _ practically inactive. Unfortunately, neither Mangold nor the Japanese authors worked heteroplastically, and they disregarded possible effects of the host levels on the regional specificity of the inductions. However, in view of the consis- tency of their results, the conclusion seems justified that the different levels of the archenteron roof have to a certain degree regionally specific induction capacities. Ob- viously, we are dealing with overlapping and not sharply delimited “induction fields” (Organisationsfeld, Spemann, °21a; Deter- minationsfeld, Weiss, ’26). Indirect evidence for a regional inductive specificity of the archenteron roof may be derived from the experiments of F. E. Leh- mann (738) in which successive gastrula stages of Triturus taeniatus were exposed to solutions of lithium chloride for short peri- ods. The treatment resulted in localized defects in the anterior or posterior head or trunk mesoderm, respectively, depending on the stage subjected to lithium chloride. Corresponding deficiencies appeared in the overlying neural system. The results were in- terpreted in terms of stage-specific (‘‘phase- specific’) and localized susceptibilities of the mesodermal regions to lithium, and a rather strict regional correspondence between meso- dermal inductors and induced neural struc- AMPHIBIANS tures was assumed. However, Pasteels (’45), who repeated these experiments on urodeles and anurans, obtaind a series of continuous rather than discontinuous deficiencies both in the mesoderm and in its inductions, which he interpreted in terms of a cephalocaudal eradient of susceptibility. The experiments of Dalcq (46, ’47) give an indication of qualitative differences be- tween the inductivity of the prechordal and the chordal-parachordal mesoderm. He bi- sected the young gastrula of the anuran, Dis- coglossus, by a horizontal cut, rotated the upper half 180 degrees and healed the halves together. Two axial systems developed, one at the original dorsal side and one on the ventral side. Variations in the distance of the plane of cutting from the blastopore re- sulted in a series of incomplete axial sys- tems. Isolated forebrains occurred in the ventral systems, and they were invariably correlated with prechordal mesoderm. Whereas the autonomy of an archencephalic inductor region was thus established, no evi- dence was found for an independent deuter- encephalic inductor. Hindbrain and spinal cord were always induced together. The author considers the hindbrain induction merely as the result of a particularly strong inductive capacity of the cranial end of the chordal-parachordal mesoderm. Nieuwkoop (47, °50) transplanted upper blastoporal lips from different gastrula stages of Triturus into the ventral side of another gastrula and likewise obtained secondary embryos show- ing successive steps of brain deficiencies. The boundary between the induced archencepha- lon and deuterencephalon coincided invari- ably with that of prechordal and chordal- parachordal mesoderm. The detailed study of the deficient brains suggested to the author that the other subdivisions of the brain are not induced by a mosaic of qualitatively dif- ferent regions of the anterior archenteron roof but that they reflect threshold values of inductive potency within a _ cephalocaudal gradient of neural inductivity in the archen- teron roof. In the extensive experiments with adult inductive tissues or their fractions it has been found frequently that some of them induce predominantly head structures, others trunk- tail structures, or both (Chuang, 739, ’40; Toivonen, *40; and others, see p. 269). Iso- lated ectoderm, when exposed to neuralizing aqueous solutions formed exclusively anterior head structures (Barth, ’41; Shen, ’42; Holt- freter, "44b; Yamada, °50a). It may be doubted whether these inductors have any 2a similarity with the normal ones, but the findings indicate again that the factors which bring about an archencephalon differ from those which induce the more caudal parts of the nervous system. However, in all these experiments, no specific inductor for the deuterencephalon was found; the latter oc- curred always combined with archencephalic or with spinocaudal structures. It was pointed out above that the poste- rior fifth of the medullary plate does not form neural but mesodermal structures of the posterior trunk and tail (see Fig. 87e). Hence one might expect that the caudal part of the archenteron roof would have meso- derm-inducing capacities. Spofford (’48) has shown that this is, indeed, the case. He sub- stituted vitally stained pieces of early gas- trula ectoderm for caudal medullary plate and found that the implants formed trunk and tail somites and other mesodermal tis- sues. Altogether, there is good evidence for a distinction between a large anterior section of the archenteron roof which induces the various parts of the neural system and the mesectoderm, and a short caudal section which induces mesodermal tissues. The pre- chordal mesoderm operates (in cooperation with the tips of notochord and somites?) as an archencephalic inductor, but no sharp borderline seems to exist between the in- ductors for hindbrain and for spinal cord. The question of whether the inductive con- ditions which specify the different brain sections represent a discontinuous series of qualitative differences or a continuous series of merely quantitative differences of a single agent is undecided. This problem has given rise to extensive discussions (see for instance, F. E. Lehmann, 738, °45, °48; Dalcq, ’46; Nieuwkoop, °47, 50; Waddington and Yao, °50). We shall return to it later on. Such variables as intensity of inductive effect and period of contact between archenteron roof and overlying ectoderm have to be taken into account (Dalcq, ’46). Whereas in normal development only the median part of the mesoderm mantle sub- jacent to the medullary plate actually exerts an inductive influence, the more lateral meso- derm also contains latent inductive capacities (Holtfreter, ’33c, ’38a; Raven, ’35; Wadding- ton, ’36a). When pieces of gastrula ectoderm were placed over the dorsolateral mesoderm of a neurula, at different levels along the cephalocaudal axis, they were induced to form a great diversity of accessory ecto- and mesodermal structures, such as fragmentary Z52 EMBRYOGENESIS: PROGRESSIVE DIFFERENTIATION brains, sense organs, somites, pronephros, and others. The different structures occurred at approximately the same sites as the cor- responding host structures, their frequencies diminishing with the distance from these sites. This indicates that the dorsolateral mesoderm, like the more median archenteron roof, consists of a cephalocaudal series of specific, yet overlapping, induction fields. Some of the differentiations, such as bal- ancers, gills, or ear vesicles, revealed the presence of induction fields outside the med- ullary plate which operate in normal devel- opment (see p. 255). However, the occur- rence of supernumerary brains or tails in a lateral position seems to show that the neural plate inductors actually extend farther into the lateral mesoderm, beyond the boundary of the neural plate. It is probable that these peripheral induction capacities are normally not expressed because the ectoderm of the neurula—in contrast to the grafted gastrula ectoderm—is no longer competent to respond to them (p. 257). Corresponding experiments on still older hosts have shown that these induction fields remain potentially active far beyond the stage at which they are normally engaged in organ determination. But a significant dif- ference exists between these normally in- active inductors and the early organizer: Whereas the latter tends to assimilate the induced mesodermal tissues into its own field, the various tissues induced by the aged in- ductors are not incorporated but tend to es- tablish axial systems of their own. For discussion of this problem see p. 279. ORGANIZATION AND EARLY DIFFERENTIATION OF THE MEDULLARY PLATE After it had become evident that the dif- ferent regions of the archenteron roof are instrumental in blocking out the major sub- divisions of the medullary plate, the question arose: To what extent is the subjacent meso- derm responsible for the more detailed pat- terning of the central nervous system? Does it operate merely as a short-term “activator” or does it exert an influence over a longer period participating also in the determina- tion of the tissue organization of the induced areas? An examination of this question revealed that the capacity for neural differentiation and organization does not arise abruptly but develops progressively during the protracted period of gastrulation. This was demonstrated by transplantations and explantations of cir- cumscribed regions of the prospective or visible medullary plate of Triturus alpestris (Mangold, ’29a; Mangold and von Woell- warth, 50). A merely epidermal differentia- tion occurred when the material was re- moved before it had contact with invagi- nated mesoderm. When taken from middle or late gastrulae which already possess an arch- enteron roof, the pieces developed partly into neural structures, but the percentage of iden- tifiable brain parts and eyes was low. From early neurula stages on, the majority of the cases formed typical brain parts and single or synophthalmic eyes. Gallera (47, °48) and Damas (747) obtained similar results with Pleurodeles. However, they stressed the point that the grafts from the younger stages differentiate into pigment cells rather than into neural tissue and they assumed that this stage-linked differentiation reflects quan- titative dosage effects of the inductive sub- stratum. After the medullary plate has become visible, its different levels are capable of differentiating into typical fore- or hindbrain, eyes, or spinal cord, respectively, when iso- lated from the underlying mesoderm (Man- gold, ’33b; Nakamura, ’38; Ter Horst, 48; and others). However, these regions represent but indistinctly outlined morphogenetic fields whose parts are not yet rigidly de- termined. A variety of experiments has shown that this applies to urodeles as well as to anurans (for references see Mangold, *31a; Adelmann, ’36). Extirpations of large parts of the anterior medullary plate, or a rotation through 90 or 180 degrees of the median portion of this region, without the underlying mesoderm (Alderman, ’35) failed to produce marked abnormalities in brain and eye development. Small pieces of the prospective spinal cord region implanted in the brain region were assimilated by the latter (Umanski, ’35). Hence, the anterior part of the medullary plate does not repre- sent a mosaic pattern of cephalic primordia but a general and labile eye-forebrain field. This field seems to possess a mediolateral gradient because heterotopic transplants of its median strip formed an eye six times as frequently as did grafts of lateral strips (Adelmann, ’30). Its anteroposterior polarity is fixed as early as the neural groove stage, as was shown by rotation experiments of the entire anterior prospective medullary plate or of its lateral half (Roach, ’45). (In the above-mentioned rotation experiments of Alderman, the transplants were much smaller AMPHIBIANS and their inherent polarity was apparently changed by the surrounding tissue.) How- ever, the mediolateral axis was found to be still reversible in the medullary plate stage (Roach, ’45). The same holds for the hind- brain field; the extirpation of its lateral half is followed by restitution from the intact half (Harrison, *47). Systematic studies of Mangold (31a, ’36) and Adelmann (’37) lead to the conclusion that the normal segregation of the field into brain and bilaterally arranged eyes is con- 259 Instead of using surgical methods, one can prevent the development of the prechordal plate by exposing early gastrulae to certain concentrations of lithium chloride. In this case, a continuous layer of mandibular head mesoderm forms across the midline, and the induced structures exhibit again the syn- drome of a cyclopean head (Lehmann, ’33; Adelmann, ’34). These results indicate that the bilateral disposition of the material for forebrain and eyes is decisively influenced by the corresponding mass distribution of the Fig. 88. Semidiagrammatic illustration of the formative effects of somites, notochord and mesenchyme upon the shaping of the neural tube; see text. (Original.) trolled by formative effects of the underly- ing parts of the mesoderm. Following ex- tirpation of the whole prechordal mesoderm from a neurula, 75 to 100 per cent of the embryos developed all degrees of synophthal- mic to cyclopean abnormalities. The percent- age of defects was approximately the same when only the median “prechordal plate” (Fig. 85) was extirpated, but it dropped to 44 per cent when only the paraxial ““man- dibular’” portions of the mesoderm had been removed. Conversely, grafts containing the median strip of the anterior neural plate without the subjacent prechordal plate pro- duced only a single eye, but when grafted together with this substratum, they formed two complete eyes connected by brain tissue, in 70 per cent of the cases. subjacent mesoderm, especially the pre- chordal plate, and that this influence is exerted even after the brain-eye field has been induced. It is doubtful whether this continued for- mative effect of the mesodermal substratum should be classified as an “induction.” Adult tissues having no axial organization can like- wise induce brains with bilaterally arranged sense organs (Holtfreter, ’34b; Chuang, ’39; Toivonen, ’40). This makes it likely that the prechordal mesoderm merely provides the unspecific mechanical conditions for the bilateral segregation of the induced field. As such this influence would belong in the same category as that of the (non-inductive) mesenchyme which is necessary as a sup- porting matrix for the transformation of the 254 EMBRYOGENESIS: PROGRESSIVE DIFFERENTIATION optic vesicle into a double-layered optic cup (Holtfreter, ’39c). The mesoderm adjacent to the spinal cord plays apparently the same role in the bi- lateralization of this structure as does the head mesoderm in the bilateral distribution of the eyes. If the prospective notochord is extirpated (Lehmann, ’26, ’28; Horstadius, 44) or its differentiation is suppressed by exposure of early gastrulae to lithium chlo- ride (Lehmann, ’37; Pasteels, ’45) or to cen- trifugal force (Pasteels, ’40), the somites may fuse across the midline. Under these condi- tions, the wall of the adjacent neural tube, instead of forming the characteristic thin floor plate, forms a thick mass (Fig. 88a). This occurs likewise when an additional neural tube is induced on the outer surface of older somites (Holtfreter, ’33c). On the other hand, when neural tissue differentiates combined with notochord alone, its contact- ing side flattens out into a thin layer (Fig. 88c), while the presence of notochord to- gether with unilaterally located somites pro- duces the asymmetrical configuration of Figure 88d. But when the spinal cord de- velops surrounded merely by mesenchyme, its walls are uniformly thick (Holtfreter, ’34c, °39c; Fig. 88e). Thus notochord and somites not only represent the inductor sys- tem for the spinal cord but they continue to influence the cell distribution within the tube. Although their formative effects are antago- nistic, notochord and somites are normally arranged in such a way that their effects supplement each other, producing the typical thinning of the floor plate and the thickening of the lateral wails of the spinal cord (Fig. 88f). It should be mentioned that the presence of an epidermal covering is equally import- ant for the shaping of the neural tube, since without it the medullary plate fails to close and spina bifida (rachischisis) results. A multitude of externally applied chemical or physical agents can prevent the closure of the plate into a tube. Many of them are identical with the agents which inhibit the movements of gastrulation (p. 237), eg., hypertonicity, extreme temperatures, deter- gents, 10do-acetamide. Undoubtedly, the archenteron roof has an important share in the initial regional speci- fication of the central nervous system. Yet the experimental data give convincing evi- dence that it is only partly responsible for the subsequent morphological and histologi- cal patterning of the morphogenetic fields which it has projected onto the dorsal ecto- derm. At best, the inductive substratum facilitates the subsequent processes in a non-specific way. The elaboration of histo- logical details is left to the self-organizing capacity of the neural tissue which was re- ferred to above. If adult tissue can induce the ectoderm to form a typical bilaterally sym- metrical brain flanked by any of the sense organs (Fig. 99b), then it is plausible to assume that similar autonomous processes oc- cur within the different gradient fields of the normal medullary plate. These self-organiz- ing processes supplement the activities of the archenteron roof and carry on when induc- tions have ceased to operate. THE INDUCTION OF NEURAL CREST DERIVATIVES AND OF OTHER ECTODERMAL STRUCTURES We have followed the progressive differen- tiation up to the neurula. In subsequent stages, a great variety of new structures ap- pears. Their morphogenesis is analyzed in other chapters of this book. The present chapter is limited to some general aspects of the determination of ectodermal derivatives not belonging to the central nervous system. Prominent among them are the derivatives of the neural crest which, in turn, originate from the neural folds. We mention briefly the following crest derivatives: the majority of the chromatophores, large amounts of mesen- chyme of the head and trunk region, the principal components of the visceral carti- laginous skeleton and of the cranial ganglia, dental papillae, Schwann’s sheath cells, at least part of the corium and of the membran- ous coverings of the brain, the chromaffin bodies and the medulla of the adrenal glands (see the reviews by Harrison, ’38; Du Shane, °43; Rawles, ’48; Horstadius, ’50). Ectodermal placodes outside of the medullary plate pro- duce the olfactory epithelium and its nerve processes, the lens, the otocyst and its acous- tic ganglion, the lateral line sense organs and some components of the cranial ganglia. NEURAL CREST DERIVATIVES Since the neural crest originates in the neural folds, one might expect that it is induced in conjunction with the adjacent medullary plate, by the more lateral parts of the archenteron roof. Although some of its derivatives, such as mesenchyme cells, chro- matophores and Rohon-Beard neurons, can be induced independently of medullary plate tissues (Holtfreter, ’°33c; Raven, ’35), these AMPHIBIANS elements have occurred in most instances in combination with parts of the central nervous system. In fact, there is hardly any case known in which induced neural tissue has not been accompanied by neural crest deriva- tives. Dalcq (41a, *46) and Raven and Kloos (45) have interpreted these results in terms of a concentration gradient of a hypothetical inductive substance, “organisine,” present in the archenteron roof (see p. 278). The median strip of the archenteron roof, sup- posedly rich in organisine, would induce neural structures while the more lateral parts which elaborate it in smaller quantities would induce neural crest. There is little evidence to support this gradient concept. Whereas the experiments of Bautzmann (’28, ’29) seemed to indicate that the prospective notochord has a stronger neuralizing effect than the prospective so- mites, the experiments mentioned above and those of Bytinski-Salz (731) did not confirm this contention. They have shown that the frequency and amount of neural tissue in- duced by prospective somites is not inferior to that obtained by prospective notochord. The embryos, in which the formation of large sections of the notochord was sup- pressed by chemical treatment without any sign of reduction of the neural tube (F. E. Lehmann, °35, ’37, ’38; Lehmann and Ris, 38), gave further evidence that the prospec- tive somites and the notochord have equal potencies as inductors. The studies of Raven and Kloos (’45) indicate that neural crest derivatives are induced equally well by median and lateral pieces of the archenteron roof, whereas neural tissue was induced more frequently by the former than the latter. However, the small number of the experi- ments does not permit a statistical evalua- tion. Furthermore, the absence of neural inductions by lateral mesoderm can be cor- related with a very poor differentiation of the transplants themselves. The available evidence thus indicates merely that some neural crest derivatives can be induced in the absence of a neural plate and that, if this happens, paraxial mesoderm was responsible for it. As a matter of principle, the question of whether the emergence of different struc- tures is due to qualitative or quantitative differences in inductive substances can only be answered by experiments in which such substances are isolated and applied in graded dosages. Experiments with normal inductive tissues can never answer this question satis- factorily. J5)) At the stage of rising neural folds, a regional pattern of limited differentiation capacities is already laid down in the neural crest: The precursors of melanophores origi- nate mainly from the trunk part of the crest (Niu, ’47, for Triturus) and those of visceral cartilage exclusively from the head crest (Harrison, ’°25; Raven, ’31, °33; Hoérstadius and Sellman, °45). Little is known concern- ing the determination of these patterns. How- ever, final differentiation of both types of precursor cells depends on additional factors residing outside the archenteron roof and the neural crest itself. The chromatophore pat- tern will be discussed elsewhere. The forma- tion of visceral cartilages requires an activa- tion of their precursor cells by pharyngeal entoderm or, under experimental conditions, by notochord or intestinal wall of the trunk (Horstadius and Sellman, ’45; and others). In a similar way, final differentiation and functioning of the other derivatives of the neural crest seem to depend on additional stimuli which are ordinarily furnished at the sites where the migratory crest cells settle down. Thus, the formation of myelin depends on the establishment of contact be- tween Schwann’s sheath cells and nerve fibers, and the crest cells migrating into the mouth region produce dentine only when in contact with oral ecto- or entoderm. One may hesitate to give these additional environmental factors the same rank as the initial ones, since it is possible that at least some of them are merely necessary to realize the potencies evoked already by the initial inductive factors. STRUCTURES OUTSIDE THE MEDULLARY PLATE The situation is similar with respect to the placode derivatives: nose, lens and oto- cyst, and some other structures. In these instances, several inductive tissues are in- volved which operate either simultaneously or in succession, supplementing and rein- forcing each other. These inductions thus serve as new illustrations of an important general principle; namely, of the existence of synergistically active “inductor systems” (Holtfreter, ’35a,b). We have already en- countered this principle in the case of the in- duction of spinal cord by the chorda-somite system and in the case of the brain-eye induction by a multiplicity of primordia of the anterior archenteron roof. In otocyst in- duction, a conditioning of the ectoderm by lateral head mesoderm is followed by an 256 EMBRYOGENESIS: PROGRESSIVE DIFFERENTIATION inductive action on the part of the lateral wall of the rhombencephalon. In the case of the lens, the optic vesicle provides the essential inductive stimulation in most am- phibian species, but in at least some species it is preceded by a conditioning influence of the head mesoderm. The latter may be suffi- cient to call forth a lens in those species which can form it in the absence of an optic vesicle. Probably both the rostral part of the archenteron roof and the prospective fore- brain are involved in the induction of the nasal placode, and both head mesectoderm and archenteron roof derivatives are con- cerned with balancer induction. In the case of the pituitary gland, the invagination of the ectoderm seems to be induced by oral ento-mesoderm, while its final differentiation depends upon a subsequent contact of this ectoderm with the infundibulum. It has been shown in several of these in- stances that in the absence of one of the two components, the induced structure fails to attain normal shape and _ differentiation. When an inductive substratum is deranged, or when its contact area with the ectoderm is abnormally large, double or multiple formations may result, such as supernumer- ary nasal epithelia (Holtfreter, °36) or medullary plates (Waddington, ’40). In order to call forth normal structures, the con- stituents of the “inductor systems” must be arranged typically with respect to the ectoderm, and they must act at the proper time. To complete this brief survey of ectodermal inductions we merely enumerate others: epi- dermal glands, enamel organs, gills, and fin. In addition, the characteristic perforations which produce mouth opening, gill slits, and anus have sometimes been listed under the term “induction.” It is true that they result from a contact action between specific parts of the intestinal tract and the overlying ecto- derm. Yet, the physiological processes mak- ing for the thinning out and regression of these ento-ectodermal membranes appear to be quite different from those which cause the cytological determination of the other struc- tures dealt with above. HOMOIOGENETIC INDUCTION As pointed out above, certain parts of the nervous system are engaged in the induction of ectodermal organs (nose, lens, otocyst, pituitary). On the other hand, in the early gastrula only the chorda-mesoderm has in- ductive capacity, whereas the ectoderm lacks it. Obviously, the medullary portion of the ectoderm becomes inductive in postgastrula stages. Mangold and Spemann (’27) dis- covered that pieces of the medullary plate of urodeles when transplanted into the blasto- coele of young gastrulae would induce a secondary medullary plate in the host, and Mangold (’29a) showed that all parts of the plate acquire the capacities for self-differen- tiation and induction simultaneously. The authors used the term “homoiogenetic in- duction” for this phenomenon. In accordance with the regional differences in the inductive specificity of the archenteron roof, anterior parts of the plate induce mainly brain and sense organs; middle portions, a spinal cord; and posterior parts, various tail structures (Mangold, ’29a, ’32, ’33b; Bytinski-Salz, ’29; Ter Horst, ’48). The experimental conditions leading to the manifestation of homoiogenetic induction are entirely artificial since, normally, the neural plate has no chance to act upon gastrula ecto- derm. From this viewpoint, the phenomenon belongs in the category of ‘abnormal in- ductions” (see p. 267). However, it is likely that the agents which neuralize gastrula ecto- derm are identical with those which are instrumental in the normally occurring in- duction of the above-mentioned placodal structures, all of which differentiate from neurula and tail-bud epidermis, and that the differences in response are due to changes in the reaction potency (competence) of the ectoderm. In fact, it has been shown that the ectoderm loses its neural competence towards the end of gastrulation, while acquiring local competences for nasal epithelium, oto- cyst, and so forth. If the above interpretation is correct, then the seemingly atypical phe- nomenon of “homoiogenetic” neural induc- tion would be merely an expression of the age differences between inductor and react- ing tissue. CHAINS OF INDUCTION Tissues such as the neural plate which must first be induced in order to become capable of inducing other structures have been termed “secondary inductors.” Subse- quently, the secondarily induced structures may act as “tertiary inductors”: the otocyst induces the surrounding mesenchyme to form a cartilaginous capsule, and it appears that the lens stimulates the overlying ecto- derm to adopt the characteristics of a cornea. Thus differentiation, particularly of the ecto- dermal structures, proceeds through the AMPHIBIANS mediation of several generations of properly distributed inductors which succeed each other and may combine in various ways to act as coordinated determinative systems. This chain of inductive processes continues to operate beyond embryonic stages, involv- ing the determination of the gonads, the mesonephros and various morphogenetic proc- esses which occur during the metamorphosis of the amphibian larvae. Thus the principle of induction is of great importance for the organization of the amphibian body and, probably, of all other vertebrates. COMPETENCE Some districts of the early amphibian embryo, particularly the entodermal ones, do not seem to require specific external stim- uli for their normal differentiation, while the mesodermal ones do require them to some extent. This dependence is most pronounced in the prospective ectodermal districts whose typical differentiations do not arise in the absence of exogenous inductive factors. On the other hand, it has been found that the differential fate of the ectoderm depends not only upon regional differences of the in- ductors but just as much upon a proper state of ectodermal responsiveness which varies with the developmental stage. The physio- logical state of a tissue which permits it to react in a morphogenetically specific way to determinative stimuli has been termed “Reaktionsfahigkeit” (Mangold, ’29a), ““Reak- tionsbereitschaft” (Machemer, ’32) or “‘com- petence”’ (Waddington, °32, ’40). Like the capacity of certain cell groups to differenti- ate normally without specific external stim- uli, competence is of course primarily determined by genetic factors. The term “competence” is useful if one wishes to dis- tinguish between cell-inherent and exoge- nous factors of cell determination. In using this term, it is necessary to specify embryonic area, stage of development, and the differentiation process to which compe- tence refers. We can define competence only in terms of the products of differentiation because the inner conditions which permit an embryonic area to pursue a certain differ- entiation when properly stimulated can be grasped today only in a very hypothetical way. It would be advisable to confine the term to a limited phase and not to the termi- nal steps of differentiation. As an illustration of some fundamental aspects of competence, data collected on the ectoderm have been selected for discussion. Do TIME PATTERN OF COMPETENCE It is well known that the ectoderm from early gastrulae can respond to different in- ductive stimuli by forming any one of a number of different ecto- or mesodermal derivatives (p. 241). These and other data refute the notion of a dichotomy of progres- sive differentiation (Lillie, 27), according to which only one of two directions of differen- tiation is open to an embryonic area at a given stage. With progressive development, the compe- tence to form a diversity of structures be- comes as a rule gradually restricted. Neural induction of the ectoderm occurs normally in advanced gastrula stages. Spemann (718), Mangold (29a), Lehmann (’28, ’29), and others have shown that the competence of the prospective epidermis for neuralization is lost shortly after this event has taken place. In special studies of this problem, Machemer (32), using urodeles, and Schechtman (’38a), using Hyla, transplanted the upper blasto- poral lip of early gastrulae under ventral and flank epidermis of various stages. They found that the capacity for neuralization decreases markedly in the late gastrula stage, and is completely absent in the early neurula ectoderm. Raunich (’42b), using adult liver as an inductor, obtained similar results. In a different set-up, Holtfreter (38a) removed pieces of prospective ectoderm from early Triturus gastrulae, reared them in standard solution for varying lengths of time, and then implanted them homo- or xeno- plastically into the dorsolateral regions of neurulae where a variety of inductors were known to be present. It was found that competence changes with time in both a quantitative and a qualitative sense. With increasing age, the grafted ectoderm forms increasingly smaller and at the same time less complex neural structures. Furthermore, new directly induced structures arise. Older ectoderm corresponding to the neurula stage forms only tissues which normally arise from the placodes and neural crest, until finally it becomes entirely refractory to in- ductive stimuli. The experiments indicate that the gradual change and final loss of competence is due to an autonomous process of ageing within the ectoderm (see also Waddington, *36b). It is not yet possible to state at which stage the ectoderm acquires its responsiveness to inductive stimuli. Mangold (’26, ’29a) and many subsequent workers observed that when living or dead inductors are grafted into 258 an early gastrula, the induced medullary plate appears synchronously with that of the host. Assuming that these grafts begin to re- lease the neuralizing agent immediately after implantation, a precocious medullary plate might have been expected. Since this did not happen, it has been inferred that the ectoderm acquires neural competence not before the middle or late gastrula stages. In view of the following data, the interpreta- MESODERMAL NEURAL Intensity oo EMBRYOGENESIS: PROGRESSIVE DIFFERENTIATION nized with these two inductive events are successive states of competence: the ecto- derm of the gastrula and early neurula will respond only to the mesodermal inductor, but is refractory to the neural inductor. During neurulation, the ectoderm acquires the capacity to form otocyst in response to the neural inductor while gradually losing re- sponsiveness to the mesodermal inductor (Fig. 89). INDUCTION ACTIVATION —— RESPONSE -=---- __ DIFFERENTIATION Soe aan OF EAR ECTODERM ~~ ~~ . s~ . ~ 7s. ome, - a ~ wae Mt Maton Sarma: 26 28 30 32 Fig. 89. Graphic representation of relative intensities of various components of ear inductions in Ambly- stoma. The maxima are placed at the same level arbitrarily. Abscissa: Harrison stages (12 = small yolk plug; 32 = advanced tail bud). Ordinate: Degree of ear differentiation in arbitrary units. (From Yntema, ’50.) tion may be somewhat different. Neuraliza- tion could be produced in the ectoderm of early gastrulae by exposing it for only a few minutes to cytolyzing solutions (p. 272). Nevertheless, the induced neural plate ap- peared synchronously with that of the donors of the ectoderm. Evidently, the early gastrula ectoderm is already competent for neuraliza- tion, but there was a latent period between the time of stimulation and visible differen- tiation. The analysis of otocyst induction by Yntema (50) provides a good example of the synchronization of competence and induc- tions in the normal embryo. Unspecified gastrula ectoderm, or prospective gill ecto- derm of older stages, was transplanted to the ear region, including a wide variety of donor and host stages in altogether more than 100 combinations. It has been known that both mesoderm and prospective hind- brain are instrumental in otocyst induction (see Harrison, 45). Yntema showed that the mesodermal action which occurs first is maximally effective in early neurula stages, whereas the hindbrain operates somewhat later with a maximum of activity in the early tail-bud stages. Remarkably synchro- REGIONAL PATTERNS OF COMPETENCE The prospective ectoderm of the early gas- trula shows no regional differences in neural or other competences (see p. 273), but such differences do develop gradually in the em- bryo. For instance, in the late neurula and early tail-bud stages, the lens-competence of the ectoderm is in some species narrowly restricted to the normal lens-forming area and its adjacent ectoderm; in Bombinator pachypus it is limited to the head epidermis, and in a number of species of Rana and Triturus it extends over the entire head and trunk (reviews of this much-discussed problem in Mangold, ’31a; Spemann, ’38; Needham, *42). In the same stages, the com- petence for balancer formation in A. puncta- tum is confined to the balancer-forming epi- dermis and its immediate neighborhood (Harrison, °25). Similar studies have been made with regard to the competence for mouth and gill formation, but the exact pat- tern of these competences has not been in- vestigated, and the suggestion of Spemann (12) that the reaction potencies for dif- ferent structures may represent overlapping gradient fields (“Reaktionsfeld,’ Mangold, AMPHIBIANS ’°31a) has never been subjected to a rigorous test. The observations of Yntema (733) con- cerning the distribution of otocyst compe- tence in the late neurula do not substantiate this concept. We have little information concerning the specific factors responsible for regional dif- ferences in competence. In the case of the ear (Yntema, 750) and of the lens (Liedke, 51) it seems that the head mesoderm pro- vides the primary specific stimuli for ear and lens differentiation which are locally differ- ent. In all probability, this initial step of de- termination (“Bahnung,” Vogt, 28a) involves a loss of reactive potencies for other stimuli. It is conceivable that in this way differentials are set up, at first between head and trunk epidermis in general, and then between different head regions. This problem, which has been touched upon already in connection with the two-step “inductor-systems” (p. 255), awaits clarification by further system- atic analysis. The data presented in tnis chapter indicate that the competence of the ectoderm changes with time and, in a rather involved way, with the region of the embryo. Since the entoderm and, to a lesser extent, the meso- derm, develop rather independently of ex- ternal determinative stimuli, we refrain from discussing their particular age-conditioned states of competence. As far as the ectoderm is concerned, it is clear that for the achieve- ment of its typical differentiations, two sets of temporally and spatially matching, yet relatively independent, mechanisms are re- quired: (1) a proper distribution of inductors which are in some measure specific; (2) a proper stage-specific state of responsiveness of the ectoderm. As was stated above, the genetic constitu- tion determines decisively the response of a given tissue to external stimuli. This point has been mentioned above in connection with the regional distribution of the competence for lens formation. REGIONAL HOST INFLUENCE When living or dead inductors were grafted into early gastrulae, results were obtained which could not always be ac- counted for by inductive specificity of the graft but which indicated the interference of regional host influences. Such effects were first shown in the experiments of Spemann (31). Whereas “head organizer” would in- duce brain structures in any body level of the host, “trunk organizer” did not express 259 its specificity when transplanted into the head level, since there it induced eyes and parts of brain which were fused with those of the host. Various explanations have been offered for this host influence, but before discussing them, it should be pointed out that certain phenomena sometimes listed under this head- ing may be due to purely mechanical inter- actions between transplant and host. (1) The fact that the structures induced by grafted blastoporal material tend to appear at the same body level as the corresponding host structures may be due to a deflection of the invagination movements of the graft by those of the host whereby both become concordant. (2) Double or multiple brains and sense organs can arise if the graft acts as a mechanical obstacle, by splitting or deranging the inductors of the host. Grafts which are otherwise rather inactive can produce this effect (Raven, °33; Holtfreter, 34a), and it is not unlikely that some of Spemann’s results can be explained in this way. The elusiveness of this problem is indicated by the fact that normal trunk-tail organizer (Mangold, ’33b; Holtfreter, ’36) as well as adult organs (Chuang, 739) can induce a tail in the host’s head region. Obviously, a clarification of this problem can only come from experiments which deal with it on a statistically satisfactory basis and avoid the situations outlined above. Besides, an analy- sis of the host influences would require a comparison of the inductions obtained by the same grafts in the whole embryo and in isolated ectoderm. Normal head and trunk inductors express their regional specificity in every case when they are grafted into isolated ectoderm vesi- cles (Holtfreter, ’36; Okada and Hama, 43, 45; Fig. 83). Comparable results were ob- tained when fresh or boiled adult tissues were implanted into such explants. For in- stance, Chuang (738, 39) found that mouse kidney, when acting upon explanted ecto- derm, would induce exclusively brain and sense organs, whereas Triturus liver would induce trunk and tail in addition to head structures (Figs. 93-96; Table 14, p. 269). Since the inductions obtained in such ex- plants could be just as complex and well organized as those produced by the same in- ductor in whole embryos, these findings in- validate the idea of some authors that atypical inductors merely “evoke” a general- ized structure such as neural tissue and that an “individuation field” of the host is re- % in Explants 4 Brain Otocyst Musculature Musculature Notochord Tail Balancer Notochord Nose Eye Levels ——»Head Heart Ant. Post. trunk trunk Fig. 90. Influence of host levels on the inductions of different organs by fresh liver of Triturus in Triturus embryos. Abscissa: host levels; left ordinate: frequency of inductions of different organs; right ordinate: fre- quency of the same organs when induced by fresh Triturus liver in explants (ectodermal vesicles). (After Chuang, ’39.) Yo 100 90 80 70 60 50 40 30 Balancer Notochord Forebrain Levels —> Head Heart Ant. Post. trunk trunk Fig. 91. Influence of host levels on the inductions of different organs by alcohol-treated livers and kidneys of different vertebrates in Triturus embryos (see text). Abscissa: host levels; ordinate: frequency of induc- tions. This graph is based on the figures given in Table 10, Toivonen (’40). AMPHIBIANS 261 Fig. 93. Fig. 94. Fig. 92. a, Inductive effects of the “head organizer’; b, inductive effects of the “trunk-tail organizer”; c, heteroplastic induction of a complete secondary embryo, oriented opposite to the main axis of the host (from Holtfreter, ’?33d). Fig. 93. Induction of headlike structure with three balancers, by fresh mouse kidney implanted in ecto- dermal vesicle of Triturus alpestris (from Chuang, ’39). Fig. 94. Induction of neural structures and eye with lens, by fresh mouse kidney implanted in ectodermal vesicle of Triturus alpestris (see Fig. 93) (from Chuang, ’38). quired to organize this tissue into regionally specific tissue patterns, such as brain with sense organs (Needham, Waddington and Needham, ’34; Woerdeman, ’36). Evidently, typical head and trunk patterns can arise within the stimulated ectoderm itself, by way of self-organization, and whatever the host influence may be, it can only modify the effects of the graft. Using the same adult tissues as in the preceding experiments, Chuang (739, ’40) grafted them into the ventral region of vari- ous levels of whole embryos. To simplify matters we shall discuss only his results ob- 262 EMBRYOGENESIS: PROGRESSIVE DIFFERENTIATION tained with Triturus liver which are based upon a statistically sufficient number of cases (185). As far as the host influence is concerned, the extensive data on mouse kidney gave much the same results. In Figure 90 the distribution of the various in- duced structures over the host is plotted according to four cephalocaudal levels. No distinction was made between direct and indirect inductions; for instance, the listed ear vesicles could have been induced directly by the graft or indirectly by hindbrain which was induced first. Fig. 95. If one uses the rate of incidence of certain inductions in explants as a control (right ordinate), it is apparent that host influences reinforce the appearance of a given struc- ture in certain levels and reduce the chance of its occurrence in other levels. Brain, nose and eye inductions clearly decrease in a cephalocaudal direction while those of pro- nephros, muscle and tail decrease in the opposite direction. However, as was the case in explants, head and tail structures can be induced simultaneously at any host level. Apparently, specific effects of the graft and regional host influences interact and compete with each other in a complicated way. The former may account for the rather unex- pected frequency of otocysts and balancers in trunk levels. Strikingly similar distribution patterns of specific inductions within a host were ob- served by Holtfreter (34b) and Toivonen (40), who used a great variety of adult tis- sues which had been pretreated in different ways. Unfortunately, the inductive specificity of these tissues had not been tested in ecto- derm explants. Nevertheless the diagram of Figure 91 shows again that cephalic in- ductions tend to increase toward the head region, and trunk-tail inductions toward the tail region of the host. Once more, otocyst and balancer fall out of line, having their maximum of occurrence not in the head but in the anterior trunk region. The available evidence indicates that there does exist a specifying influence of the host level upon accessory inductions, although ae Fig. 96. Fig. 95. Induction by fresh liver of Triturus implanted in ectodermal vesicle of Triturus alpestris. Notice tail and balancer. (From Chuang, ’38.) Tig. 96. Induction by fresh liver of Triturus, implanted in ectodermal vesicle of Triturus alpestris (see Fig. 95). Notice spinal cord, notochord, somites, tail fins. (From Chuang, ’38.) this influence asserts itself only in a certain percentage of cases. This phenomenon does not seem to be covered by our explanations given above and one has to look for other possible interpretations. The following sug- gestions have been made: (1) The head mesoderm differs from trunk mesoderm quantitatively and possibly also in its reaction potency (Chuang, 39); hence the difficulty of the grafts to induce mesoder- mal trunk structures in the head region. (2) The competence of the ectoderm appears to vary along the cephalocaudal axis (Spemann, 31). This point has been discussed before (p. 258). Further indications for such a gradient are found in experiments which show that the ectodermal response to a graft is usually more pronounced in the anterior than in the posterior ventral regions of the embryo (Lehmann, ’32; Machemer, ’32; Holtfreter, °34b; Schechtman, ’38a; Chuang, 39). (3) The primary host inductors may tend to assimilate the accessorily induced AMPHIBIANS structures into their regional fields. We recall the observation that the inductors in the archenteron roof represent overlapping fields which reach far into the lateral parts of the embryo. The normally unexpressed activity of the periphery of these fields may inter- vene and superimpose its regional character istics upon the inductions of the graft. Such an auxiliary effect seems to account for the fact that the closer a graft lies to the dorsal host mesoderm, the more pronounced is its neural induction (Raven, 733). There is also the “bridge phenomenon”: An induced neural plate lying close to that of the host usually links up with the latter, suggesting that sub-threshold inductive fac- tors of the lateral mesoderm can add up with those of the graft to induce the neural bridge. The most impressive indication of such an auxiliary host effect was found by Pasteels ('47a,b, 49a) who produced second- ary axial systems by centrifuging blastula and gastrula stages. The secondary formations did not attain perfect differentiation unless they appeared linked up with the primary axial system (‘‘contagion,” Pasteels). The interpretation of host effects in terms of regional fields meets with the difficulty that the peaks of frequency of the induced organs are not always at the levels of the respective host organs, where one would ex- pect them. Obviously, the term “host influence” com- prises a group of heterogeneous factors, and their analysis offers serious obstacles. In a given instance, it is difficult to decide whether the graft suppresses the activities of the host inductors, or whether it displaces the host inductors or cooperates with them. Furthermore, the regional specificity of the grafts themselves is somewhat variable and can be expressed only in statistical terms. HETEROPLASTIC AND XENOPLASTIC TRANSPLANTATIONS* Born (1897) was the first to accomplish the fusion of parts of amphibian embryos belonging to different species and genera. This new method proved to be of great value in the analysis of some basic problems such as the interactions of genetic and environ- mental factors in growth (reviewed by Har- rison, ’35; Twitty, 40), the formation of pig- ment patterns (see Twitty, 42, 45), and * “Ffeteroplastic” usually refers to tissue combi- nations between different species of the same genus, and “‘xenoplastic” (Geinitz, ’25b) to those between more distantly related forms. 263 embryonic induction. The following discus- sion is concerned only with the last-men- tioned problem. The combination of inductors and reacting tissues differing in hereditary characteristics, such as pigmentation, or size of cells and nuclei, allows for an exact distinction be- tween transplanted and host tissues. There- fore, this method has become an irreplaceable tool for the finer analysis of induction. Ex- periments of this type have established several fundamental principles: 1. Induction operates across the borders of species and even orders. This has been shown in a variety of experiments. For in- stance, the optic vesicle of a urodele embryo can induce a lens in competent epidermis of other urodele species or of anurans, and the chorda-mesoderm can induce a_ secondary embryo in the gastrula of a foreign species or order. It follows that the inductive agents are not species-specific. 2. The responses to inductive stimuli are in accordance with the genetic potentialities of the reacting material; they are species- specific with respect to form, growth rate, and other tissue characters. For instance, a balancer induced by head structures of Triturus cristatus in transplanted prospective belly ectoderm of 7. taeniatus is a typical taeniatus-balancer (Rotmann, *35a; see Fig. 106). The genetic constitution is thus recog- nized as one of the most important limiting factors for the competence of the reacting tissue. 3. The larvae of urodeles and anurans dif- fer in various order-specific structures: The urodeles are equipped with a pair of bal- ancers on the ventrolateral side of the head, whereas the anurans have a pair of adhesive glands (‘suckers’) in a more ventral position. The larvae of urodeles possess dentine teeth, whereas the anuran tadpoles do not develop them until metamorphosis; instead, tad- poles are equipped with horny denticles which are not homologous with the former. There are conspicuous differences in the shape, number and topographic relations of the cartilages of the visceral skeleton, in body pigmentation and other characters. The in- duction of these structures which are diver- gent in the two orders has been analyzed by means of xenoplastic transplantations. It was found that the inductors are responsible for the regional specificity of the induced organs, but the pattern and structural characters of the latter are determined by the inherent properties of the reacting material. The fol- lowing experiments illustrate this point. 264 EMBRYOGENESIS: PROGRESSIVE DIFFERENTIATION When belly ectoderm of an anuran gas- trula is transplanted to the prospective head region of a urodele gastrula, it is in- duced to form suckers on the ventral side of the head (Spemann and Schotté, *32) and horny denticles in the mouth region (Holt- freter, ’35b; Schmidt, ’37; Spemann, ’38; see Fig. 97). In the reciprocal experiment, head structures of the frog or toad induced bal- ancers (Fig. 98) and dentine teeth in the pros- pective neural ectoderm of a salamander evolution. This aspect has been stressed particularly by Baltzer and his associates, in their extensive transplantation experiments (Baltzer, °41, ’50a,b; Andres, *49; Wagner, 49; Roth, *50). It seems that when structural divergencies originated in the course of evolution, the underlying inductive mechanisms were re- tained, in part, and they have remained a common property of both urodeles and anurans. The term “homodynamic” was pro- Fig. 97. Fig. 97. Xenoplastic substitution of prospective ventral head ectoderm of Triturus by belly ectoderm of Rana esculenta, in early gastrula stage. The urodele head structures have induced mouth implements of anuran type: horny “teeth” and suckers. (From Spemann, ’38, after Schotté.) Fig. 98. Xenoplastic substitution of prospective ventral head ectoderm of the toad (Bombinator) by belly ectoderm of Triturus taeniatus, in early gastrula stage. The anuran head structures have induced a urodele balancer. (From Spemann, ’38, after Rotmann.) (Holtfreter, ’°35a; Rotmann, °35b). The re- spective inductors are regionally specific since they induce head implements, but they are at the same time sufficiently general to call forth the formation of structures which do not occur in the genetic repertory of the inducing host. Spemann and Schotté (32; see also Spemann, 38) used the term “gen- eral situation stimulus” in describing this condition. The explantation experiments of Holtfreter (36) have demonstrated that re- gionally specific structures can be induced xenoplastically, outside of a whole embryo. All these results brought sharply into focus the “release” character of the inductive mech- anism, and the important role which the self- organizing capacity as well as the genetic constitution of the reacting tissue play in the inductive process. The induction of organs which urodeles and anurans do not share has interesting implications for problems of homology and Fig. 98. posed for factors which are equivalent in different taxonomic groups, and the term “specific” for factors in which they differ (Baltzer, “50b). For instance, the cephalic inductors of balancers and suckers would be homodynamic, while the competence of the ectoderm would represent the specific factors. The same situation was revealed in exchange transplantation between JT. taeni- atus, which is equipped with balancers, and the axolotl, which lacks them. It was found that both possess the balancer inductors but that the axolotl ectoderm has lost the cap- acity to respond to them (Mangold, ’31b). The following example may illustrate the point that inductive agents as well as com- petences can be homodynamic. When anterior neural crest was exchanged between neurulae of 7. alpestris and the toad, Bombinator pachypus, its derivatives partici- pated in the formation and induction of nu- merous structures in the foreign head AMPHIBIANS (Wagner, 49). Dermal bones of the visceral skeleton, which are characteristic of early larvae of urodeles but absent in anuran tad- poles, were formed by the mesectoderm of Triturus in the anuran head in typical loca- tion and at the normal time. Conversely, the mesectoderm of Bombinator formed typical rostral cartilages in the urodele head which normally lacks these structures. These results imply that in both orders the entomesodermal head tissues are homodynamic in their capac- ity to provide a favorable substratum for the migration and final topographic localization of the skeletogenous neural crest derivatives. In addition, the visceral entoderm supplies an agent necessary for chondrogenesis (Horstad- ius and Seilman, °45) which is equally effective (that is, homodynamic) in both orders. Anuran mesectoderm formed tooth papillae in the larval Triturus head, although normally the anuran tadpoles do not form teeth until metamorphosis. Bombinator and Triturus mesectoderm are thus shown to be homodynamic in their competence to stimuli which induce papillae. Baltzer (50a, b) has pointed out that the homodynamic components of the develop- mental mechanisms which are a common stock of large taxonomic units represent an as yet unexhausted reservoir for future evo- lutionary divergencies. The specific compo- nents, on the other hand, indicate where genetic changes have occurred in the past. The ectoderm appears to be predominantly a carrier of specific factors, while the in- ducing mesoderm and entoderm carry the homodynamic factors, although this is by no means a general rule. In several instances, the evolutionary change seems to have in- volved shifts in the time patterns of differen- tiation, as is illustrated in the delay of dermal bone and tooth formation in anurans (for a discussion of these and related prob- lems, see DeBeer, 751). ANALYSIS OF THE PHYSIOLOGICAL MECHANISM OF INDUCTION From the preceding account rather com- plete information has been gained concerning the morphological phenomena of induction in the amphibian embryo. Departing from this level, further research proceeded to examine the induction mechanism from a more physiological and biochemical angle. Since it is impossible in this context to re- view adeauately the vast literature in this field, we shall consider only those data which are relevant to our topic. 265 INDUCTIVE CAPACITY OF DEAD EMBRYONIC TISSUES In 1932, a short collective report by Bautz- mann, Holtfreter, Spemann, and Mangold announced that organizer and medullary plate retained inductive power after they had been killed by heat, cold, or alcohol (Fig. 99). Holtfreter showed in addition that non-inductors, such as ectoderm and entoderm of the gastrula, acquired induc- tion capacity through these treatments. More intensive studies were then devoted to the effects of extreme temperatures and various chemical agents on the inductive capacity of embryonic tissues (Holtfreter, ’33e, 34a). The test method consisted largely in grafting the killed tissues into the blastocoele of early Triturus gastrulae. One of the first surprises in these experi- ments was the observation that the neuraliz- ing capacity of the inductors was hardly affected by drying, or boiling at 100° C., whereas the mesodermizing capacity disap- peared rapidly through these treatments. At temperatures between 100° and 150° C. the neuralizing effect was likewise progressively reduced to zero. After it had been either dried, heated, frozen, or treated with acid, alcohol, or boil- ing ether, the posterior part of the medullary plate plus subjacent archenteron roof which, in the living state, can induce trunk and tail tissues, had not only lost its mesodermiz- ing activity but also changed its regional specificity: it now induced brain portions which could be associated with various sense organs (see also Barth and Graff, ’38; Okada and Takaya, ’42; Holtfreter, ’48a). Likewise, ectoderm and ventral entoderm from a gas- trula and early neurula which were rendered active by extreme temperatures or fat solvents had a strong tendency to induce brain por- tions and sometimes free lenses, but they never induced mesoderm. Thus, through the different killing methods, the original re- gional specificity of the normal inductors was abolished, and all parts of the gastrula or neurula became inductors of cephalic structures. Also inductive were all cyto- plasmic regions and the nucleus of the coagulated ovum (Holtfreter, ’34a; Wadding- ton’, o8b). In order to disarm the possible criticism that the high degree of organization found in the induced structures might have been due to an activation of the host’s determination fields (Weiss, ’35; Woerdeman, °36), the host influence was excluded in the following 266 ; EMBRYOGENESIS: PROGRESSIVE DIFFERENTIATION Fig. 99. a Fig. 100. Fig. 99. Induction of neural structures and eye with lens, obtained by implantation of heat-killed anterior medullary plate into the blastocoele of an early gastrula of Triturus (from Holtfreter, *33e). Fig. 100. Induction of a secondary embryo obtained by implantation of a piece of boiled human thyroid gland into the blastocoele of an early gastrula of Triturus. Notice neural tube, notochord and otocysts in b. (From Holtfreter, ’34b.) two experimental modifications (Holtfreter, ’33e): (1) the heat-killed tissues were sand- wiched between two sheets of isolated pro- spective epidermis from an early gastrula; (2) pieces of prospective epidermis were loosely placed upon the dead tissues. In both cases, voluminous inductions were obtained in the explants representing brainlike forma- tions (Fig. 101) which were sometimes ac- companied by mesenchyme, melanophores, or an olfactory placode. Graded inactivation of neural inductors by heat was as a rule expressed in three inter- related phenomena: relative frequency, vol- ume, and organological complexity of the induced structures. If large enough, the in- AMPHIBIANS duced cell mass could organize itself into brain and sense organs. Very small masses formed nondescript neural cell aggregates or vesicles. Finally, there were intermediary cell types between epidermis and neural tissue known as ‘“neuroid” or ‘“palisade” structures which were produced by the weakest inductors (Holtfreter, ’33c, °34a; Fig. 101. Fig. 101. Induction of neural structures, obtained by placing a piece of prospective epidermis of the early medullary plate into the blastocoele of an early gastrula of Triturus (from Holtfreter, ’33e). Fig. 102. Neural differentiations in fused pieces of prospective ectoderm of middle gastrulae of Amblystoma punctatum, reared (without inductor) in salt solution (from Barth, ’41). Needham, Waddington and Needham, ’34). It seems that the size of the induced area plays a more important role in determining the degree of complexity of the induced structure than does any special property of the inductor (see p. 278). The following conclusions may be drawn: a chemical stimulus which is relatively stable in extreme temperatures and in fat solvents, rather than mere physical stimulation, is instrumental in neural induction. Heating, freezing and drying abolish the mesodermiz- ing but not the neuralizing inductivity. Furthermore, neuralizing agents are present in all germ layers but are normally inactive in the ectoderm and entoderm, either be- cause they cannot diffuse out while these tissues are alive (Holtfreter, ’33e; Toivonen, 40) or, more likely, because they occur there in a physically masked, or chemically bound form (Needham, Waddington and Needham, °34) and are liberated only after the cells are killed. 267 DISTRIBUTION OF INDUCING AGENTS IN THE ANIMAL KINGDOM It has been shown that inductions are possible in combinations of tissues from dif- ferent species or orders (p. 263). It was known, furthermore, that the inducing ca- pacity of at least some tissues is retained in Fig. 102. stages in which the ectoderm is no longer competent to react to them. For instance, brain from a swimming larva (Mangold, °29b) and somites and notochord from tail- bud stages (Holtfreter, ’33c, ’36) proved to be inductors. Woerdeman (’33b) reported that even skeletal muscle and sarcoma from adult rat, chick and man can neuralize gas- trula ectoderm. Holtfreter ((34b) made an extensive study of the inducing capacity of a variety of tissues from larval and adult organisms. Fragments from practically every organ or tissue from various amphibians, reptiles, birds and mammals, including man, were inductive (Figs. 100, 103, 104). The re- sults were about the same whether the grafts were fresh, or had been dried, or briefly ex- posed to higher temperatures. Likewise ac- tive were cell-free coagulated homogenates from amphibian and chick embryos and from Daphnia and beef liver. Most tissues tended to induce neural structures with cephalic characteristics, while others (e.g., liver, kid- 268 EMBRYOGENESIS: PROGRESSIVE DIFFERENTIATION Fig. 105. Fig. 104. Fig. 106. Fig. 103. Induction of symmetrical brain parts and eye by mouse kidney, boiled for 15 minutes and im- planted in the blastocoele of an early gastrula of Triturus (from Chuang, ’39). Fig. 104. Induction of tail in posterior trunk region of Triturus, obtained by implantation of mouse kidney which had been killed by dipping briefly in boiled water (from Chuang, ’39). Fig. 105. Induction of a neural tube by 3,4-benzpyrene, using coagulated egg albumen as a carrier. Axial organs of host above, and induced neural tube adjacent to transplant below. (From Needham, ’42, after Waddington. ) Fig. 106. Heteroplastic induction of balancer. a, Larva of Triturus cristatus. The prospective head epider- mis on the right side had been replaced in the early gastrula by prospective belly epidermis of Triturus taeniatus. The right balancer induced in the transplant by underlying T. cristatus structures is of T. taeniatus type. Notice its size, shape, direction of outgrowth, absence of terminal club. b, Normal larva of Triturus taeniatus, for comparison. (From Rotmann, ’35a.) ney, thyroid of some mammals) induced preferentially spinocaudal structures. Less conspicuous and complex inductions were ob- tained with grafted tissues from various in- vertebrates. Inactive were agar, starch, glycogen, wax, charcoal. Similar results were obtained by Need- ham, Waddington and Needham (’34) and Wehmeier (’34) on urodeles, and by Rau- nich (’42b) on anurans. The implantation of plant tissues (cambium, root tip) pro- duced merely ectodermal thickenings (Rago- sina, °37; Toivonen, ’38); but Brachet (43, °50) obtained neuralizations by applying nucleoprotein fractions from yeast and wheat embryos. AMPHIBIANS DIFFERENCES IN INDUCTIVITY OF ADULT TISSUES The phenomenon that some adult tissues imitate the action of normal head inductors and others that of trunk-tail inductors was investigated further by Chuang (738, 739), who used mouse kidney and Triturus liver as inductors, testing their effects both on whole Triturus gastrulae and on isolated gastrula ectoderm which served as a jacket for the grafts (Figs. 93-96, 103, 104). Table 14 shows that, whereas Triturus liver induced head as well as_ trunk-tail structures in both experimental series, mouse 269 into two groups: liver, thymus and kidney of certain animals induced almost exclusively “archencephalic” structures* (tel- and dien- cephalon with eyes, nose, balancers), where- as the kidneys of other animals acted as “deuterencephalic” (rhombencephalon, oto- cysts) and spinocaudal inductors. Undoubt- edly, regional influences of the host com- plicate the situation. However, even if due allowance is made for this factor, differ- ences of different organs with respect to their archencephalic and spinocaudal induc- tivities are uncontestable. Yet no relationship of a general kind can be established between the histological type of certain adult organs Taste 14. Inductive Specificity of Fresh Mouse Kidney and Triturus Liver (from Chuang, ’39) RESULTS IN PERCENTAGE NO. OF BAL- EAR NOTO- PRO- GRAFT CASES BRAIN NOSE EYE ANCER VESICLE TAIL MUSCLE CHORD NEPHROS In Explants Kidney 97 100 MAL} Pf il. DN} Ih — — = = Liver 63 a5) ihifec! (Geo) 9.0) 68.4 41.2 5206 44.4 — In Whole Gastrulae Kidney 186 97.8 9) 1 9.6 8.6 63.9 B19) 7 27.9 1.07 13.9 Liver 198 Ws MBs 9.6 9.0 46.4 47.0 Bil 3} BAG 1356 kidney would do so only when grafted into a whole gastrula but confined itself to purely ectodermal cephalic structures when enclosed in isolated ectoderm. This difference can be explained by assuming that mouse kidney is unable to induce mesodermal structures out of ectoderm, but can induce them from the mesoderm which is available in whole embryos. Triturus liver, on the other hand, can induce mesodermal trunk structures from either ectoderm or mesoderm. These results show that differences do exist in the effect- specificity of adult organs. In agreement with the data on heat-treated embryonic tissues, Chuang (739, ’40) found that in both kidney and liver the mesodermizing capacity is much more readily abolished by boiling water (within two seconds) than is the neuralizing capacity (still present after one hour of boiling, Fig. 103). This indicates the presence of different inducing factors which are possibly similar to those present in embryonic inductors. Inductor-specific differences were brought out even more strikingly in the experiments of Toivonen (’40), who tested various adult vertebrate tissues which were first treated with 70 per cent alcohol and then implanted in whole Triturus gastrulae. The tissues fall (such as liver or kidney) and the type of their inductions. For instance, the alcohol- treated kidney of the viper induced archen- cephalic structures, that of the guinea pig spinocaudal structures and that of the beef induced both (Toivonen, ’40; Toivonen and Kuusi, 748). If one compares these results with those of Holtfreter ((34b), Chuang (°39) and Hama (44) on the inductivity of liver and kidney from various species, it becomes even more difficult to draw general conclu- sions as to the inductive specificity of adult tissues. In addition to species differences, such factors as age (Rotmann, ’42) and star- vation (Toivonen, ’51) of the donor animal, as well as pretreatment of the inducing material and the competence condition of the reacting tissue, influence the results. This applies not only to such complex inductions as head and trunk structures but also to single tissues such as a free lens, which can be induced by a variety of fresh or killed embryonic and adult tissues. * A terminological inconsistency should be pointed out: Lehmann (’42b) introduced the terms “archencephalic” and ‘‘deuterencephalic” as des- ignations of parts of the brain, whereas Toivonen and others use them to designate different regions of the head, including non-neural structures. DA EMBRYOGENESIS: PROGRESSIVE DIFFERENTIATION ATTEMPTS TO ISOLATE AND IDENTIFY THE INDUCTIVE AGENTS Several groups of workers set out to sub- ject embryonic amphibian tissues and in- ductive adult organs to various extraction methods in the hope that the active agents could be isolated and chemically identified. In addition, a large number of well charac- terized chemical compounds were tested. The extracts and substances were usually in- cluded in supporting material, such as agar or albumen, which was then grafted into the blastocoele cavity of early urodele gastrulae. Needham (’42) and Brachet (’45) have dis- cussed these experiments extensively. We shall confine ourselves to some pertinent points. At the beginning, we wish to emphasize the methodological difficulties involved in this kind of investigation. They arise mainly from the fact that the ectoderm and ento- derm of the gastrula contain neuralizing agents in a masked form which can _ be liberated or activated by practically any treatment that kills the tissue. The im- planted tissue or substance may be toxic and kill surrounding host cells which, in turn, would release their liberated agents into the adjacent, non-injured' ectoderm (Barth and Graff, ’38; Okada, °38; “relay mechanism,” Holtfreter, ’45b). Furthermore, as Brachet points out, implanted dead or adult tissue is subject to degradations by the enzyme systems of the host. For these reasons, it is impossible, in principle, to obtain conclu- sive direct evidence that any of the tissues or substances which were implanted contains, or is identical with, the neuralizing agent of the living archenteron roof. The following “inductors” operate prob- ably indirectly (“relay”), by cytolyzing host tissue: inorganic substances, such as silica and kaolin (Okada, ’38), cephalin and digito- nin (Barth and Graff, ’38; Barth, ’39). Like- wise methylene blue, claimed by Wadding- ton, Needham and Brachet (’36) and by Beatty, deJong and Zielinski (’39) to have neuralizing effects, was found to be effective only in toxic concentrations (Holtfreter, *45b).* This was true also of sulfhydryl compounds for which Brachet and Rapkine (39) claimed a neuralizing activity. As to the inductivity of thiocyanate (Ranzi and Tamini, °39), the evidence suggests that the observed neuralizations were either due to a toxicity of the chemical or to contamination of the ectoderm with mesoderm. The much- * Pasteels (751) reports that data of his own do not support the latter conclusion. discussed question whether or not glycogen can induce is still open; very likely, the posi- tive cases reported can be attributed to im- purities of some glycogen preparations. The Cambridge workers contended that the naturally occurring neural inductor (“evocator”) is a steroid compound. This claim was based on the effectiveness of the unsaponifiable fraction of ether extracts from embryos and adult tissues, and further- more, on the effectiveness of a variety of chemically identified steroids (see Fig. 105; for references, see Waddington, ’40; Need- ham, °42). Concerning the first point, the evidence is not very striking with respect to the frequency and the histological type of inductions observed. Unfortunately, control experiments showed that coagulated egg al- bumen, which usually served as carrier, could by itself induce similar palisade struc- tures and neuroid cell aggregations, which further weakens the case. Fischer (°35) and H. Lehmann (738) found the unsaponifiable fraction to be inactive when freed from acids and salts, but the saponifiable fraction, containing fatty acids, to be strongly active. This was confirmed by Barth and Graff (38) and Toivonen (’49, 50), who found, furthermore, that the extracted residues were far more active than their ether extracts. The second point, the effectiveness of chem- ically identified polycyclic hydrocarbons, most of which had been prepared syn- thetically (Waddington, ’38a), deserves more serious consideration. Some of these com- pounds are known as estrogenic, others as carcinogenic agents; however, some other- wise biologically inactive substances did in- duce as well, and there was no parallelism between the neurogenic and the estrogenic or carcinogenic activity of these compounds. Experiments of Shen (739) with a water- soluble polycyclic hydrocarbon (a carcino- gen) have been used as the main argument in support of the steroid nature of the normal neural evocator. The substance, in very low doses, was mixed with crystalline al- bumen which was then implanted into the blastocoele. The alleged correlation between dosage and percentage of inductions is not clear: a peak for neural inductions seems to be at a concentration of 0.0125y,- while the percentage of neuroid structures simply in- creases with decreasing dosage. The effectiveness of very low concentra- tions of this hydrocarbon was taken by Waddington (40) and Needham (’42) as an + This peak has statistically little significance (chi square test). AMPHIBIANS indication that it has a direct effect, similar to that of the normal neural inductor, rather than a “relay” effect. However, Shen did observe necrotic cells in all of his experi- mental series. Furthermore, as has been pointed out by Needham himself (39, °42) and by Brachet (745), some substances, at extremely low concentrations, may be com- paratively more toxic, diffusible, or bio- logically active than others. It appears, there- fore, that the questions of whether poly- cyclic hydrocarbons act directly or indi- rectly and whether such substances are at all involved in normal neural induction cannot yet be answered. The same uncertainty seems to apply to fatty acids and nucleic acids. Their neuraliz- ing action was on the whole more pro- nounced than that of steroids. Since such entirely different substances as purified fatty acids from plants (e.g., oleic acid) or ani- mals (e.g., muscle adenylic acid), and com- pletely lipid-free nucleoproteins and nu- cleotides were equally effective, the inves- tigators (Fischer and collaborators, ’33, °35; H. Lehmann, ’38) refrained from identify- ing any of them with the hypothetical neu- ralizing agent of normal development and ascribed their common effect to an unspec- ified “‘acid stimulus.” Possible cytolytic ef- fects of the substances used have not been considered by these workers, although they have been demonstrated (Holtfreter, *45b; Brachet, ’49). A relay-effect of the “acid stimulus” is therefore within the range of possibility. FRACTIONATION OF TISSUE EXTRACTS One might hope to break the deadlock by comparing the activities of various tissue extracts, obtained by different fractionation methods, with the inductive capacity of the residues. The pioneering experiments were done by Toivonen and Kuusi (748) and Toivonen (749, °50), who analyzed the spe- cific inductive capacities of guinea pig liver (‘archencephalic inductor’) and _ kidney (“‘spinocaudal inductor”) by testing a great variety of fractions of these tissues obtained by extraction with alcohol and petroleum ether, dialysis, treatment with salt solutions, and heat treatment. All implants were made in whole Triturus embryos. The authors suggest that there exist two qualitatively different inducing substances, an archen- cephalic agent which is dialyzable, thermo- stable and ether-soluble, and a spinocaudal agent which is thermolabile and not ex- Af tractable in ether. Kuusi (’51) continued the analysis of the same tissues, testing tissue homogenates, isolated nuclei, protein frac- tions, cytoplasmic granules, ribonuclease- treated homogenate, formol-treated tissue and others. The capacity for spinocaudal inductions was not associated with cyto- plasmic granules or with extracted nucleo- proteins but seemed to be linked with protein and was easily lost by fractionation proced- ures. The capacity for archencephalic in- ductions seemed to be rather stable. How- ever, no clear-cut chemical separation of the two hypothetical agents has been ob- tained. In order to avoid the interference of regional host effects, Yamada (personal com- munication) used isolated gastrula ectoderm as the reacting material. He found that a 0.14 M sodium chloride extract (supernatant) of guinea pig kidney induces somites and spinal cord. The extract after heat treatment, and an RNA-protein fraction isolated from the original extract, induced archencephalic structures. DNA-protein of the same tissue gave also archencephalic inductions. The re- sults are interpreted in terms of a “dorsaliz- ing” and a “caudalizing” factor. Although the techniques employed by these workers were not yet adequate for a chemical characterization of the different agents, this type of experiment seems to be a particularly promising approach to the problem of induction. RELATION BETWEEN INDUCTION AND NUCLEIC ACIDS Brachet (45) has marshalled a variety of data in support of his contention that nucleic acids, and particularly RNA found in small granules of the cytoplasm, are responsible for inductions by both normal and abnormal inductors. In sectioned amphibian embryos stained with pyronin or toluidine blue, Brachet (40, °43) found an abundance of basophilic cytoplasmic elements whose stainability was lost after treatment with ribonuclease, sug- gesting that the basophilia was due mainly* to the presence of RNA. Microscopic study of the distribution of basophilic elements in amphibian embryos showed a high con- centration in the upper blastoporal lip which decreased during invagination of this mate- * Toluidine blue, unless applied within a strongly acid range, stains proteins in addition to nucleic acid (Herrmann, Nicholas, and Boricious, 50). Brachet does not mention having controlled the pH. 22 EMBRYOGENESIS: PROGRESSIVE DIFFERENTIATION rial.* A simultaneous increase in stainable elements in the medullary plate suggested a transfer of RNA from the inductor to the neural plate. Later, basophilia was again strong in the notochord, decreasing in the mesoderm in a ventral direction, but the epidermis contained these granules as well.+ Exact quantitative determinations of this dis- tribution pattern have not been made. Using the ultracentrifuge, Brachet and cvu- workers (40, ’42, ’44) isolated from homog- enized embryonic and adult tissues small cytoplasmic granules which contained all or most of the RNA, besides phosphatides, pro- teins with —SH groups, and certain enzymes which varied with the type of tissue. These granules which were assumed to be identi- cal with the basophilic elements in sections proved to be inductive when applied as coag- ulated grafts, as was also tobacco mosaic virus. Treatment of organizer tissue, adult organs, or tobacco mosaic virus with ribo- nuclease resulted in a conspicuous decrease or loss of their inductive capacity. All these data suggested that either RNA or its nu- cleotides are instrumental in normal and experimental neuralization. Persuasive as these data may appear, they are subject to several objections which re- duce their conclusiveness. (1) The basophilic elements of the amphibian embryo occur over much wider ranges of tissues and de- velopmental stages than are normally in- volved in induction. Their distribution seems to reflect rates of general metabolic activity rather than inductive activity. (2) The scanty histological evidence offered by Brachet (45) indicates that the implanted granules are much less effective inductors than are whole tissues. Kuusi (751) found that the isolated granules from guinea pig liver and kidney are not more potent than are the cell residues, or the supernatant of the homogenates. Contrary to Brachet, Kuusi did not find any relationship between the RNA content of the various preparations * Pasteels ('49b) found a steady increase of baso- philic granules in the prechordal mesoderm. + In late gastrulae and neurulae of T. alpestris, Cagianut (°49) found pyronin-stainable cell struc- tures distributed in both the neural plate and the archenteron roof along a cephalocaudal gradient and diminishing in density toward the ventral regions of the ectoderm and mesoderm. These struc- tures consisted mainly of perinuclear and cortical caps, and of long filaments such as Brachet (°40, ’43) has already described. It appears rather hazard- ous to identify these gross structures with the gran- ules of various sizes that have been isolated from centrifuged homogenates. and their inductive capacity. The results ob- tained with tissues that had been hydrolyzed with crystalline ribonuclease led Kuusi to the conclusion that this treatment fails to reduce their inductivity both in a quantita- tive and in a qualitative sense. (3) Doubts about a correlation between RNA content and inductivity of a graft are strengthened by the observation of Toivonen and Kuusi (48) that proteolytic enzymes practically in- activate inductive tissues. (4) Different prep- arations of crystalline ribonuclease contain proteolytic enzymes (Cohen, °44; McDon- ald, °48), and it appears that the prepara- tions used by Brachet have not been assayed for such impurities; on the other hand, its efficiency in removing the RNA present in isolated granules or in tobacco mosaic virus has been questioned (Loring, °42; Claude, 44). (5) Observations of Brachet himself (49) make it likely that the neuralizing activity of pure nucleic acids and nucleotides is due to a relay mechanism because these substances in low concentrations cytolyze ectoderm explants of the axolotl. The above studies have the merit of draw- ing attention to the possible significance of cytoplasmic granules in morphogenetic proc- esses, but they do not settle the problem of the chemical nature of the neuralizing agent in favor of nucleic acids or any of the other components of the basophilic elements. NEURALIZATION IN RESPONSE TO A TRANSIENT CELL INJURY A new approach to the analysis of induc- tion was opened by the discovery of Barth (41) that the isolated ectoderm of A. punctatum, in contrast to the ectoderm of other species, can form neural structures in the absence of a tangible inductor, namely when reared in standard solution (Fig. 102). Holtfreter (44b) confirmed this result. He showed that it is correlated with cellular disaggregation of the explant, the degree of disaggregation being proportional to that of subsequent neuralization. Obviously, A. punctatum is more susceptible than other urodeles to ordinary standard solution which hitherto had been considered as “neutral” with regard to factors that direct cell de- termination. One could anticipate that if ectoderm of A. punctatum were cultured in a less injurious medium it would become epi- dermal, and that the ectoderm of other spe- cies, for which standard solution is not injurious, could be neuralized if it would undergo a transient disaggregation. Subse- AMPHIBIANS quent experiments on A. punctatum and T. torosus verified this assumption (Holtfreter, "45b, *47b). Cellular adhesion and integrity of the cell membrane in isolated amphibian tissues de- pend largely on the pH and the presence of an adequate amount of calcium ions in the isotonic salt solution. A lowering of pH to 4.5 or the addition of glucose, sucrose or histone to the culture medium usually re- sulted in an epidermal differentiation of A, punctatum explants, probably owing to a fortification of the cell membrane. On the contrary, neural differentiation was obtained in gastrula ectoderm of T. torosus which had been briefly treated with inorganic acids, or alkali, alcohol, distilled water, or cal- cium-free standard solution. These agents are known to increase cell permeability; they cause swelling and ameboid dispersal of the cells, and finally cytolysis when ap- plied for longer periods. Reintegration of the dispersed cells was produced by their transfer to a neutral balanced medium. The closer the cells were brought to the brink of death, the more pronounced was their tendency to become neural. Yamada (’50a) obtained similar neuralizations in isolated ectoderm of 7. pyrrhogaster by exposing it briefly to ammonia. In the above experiments the prospective epidermal and medullary areas of the gas- trula reacted alike. Any portion of the ectoderm could form brain-like structures as- sociated with olfactory placodes, rudimen- tary eyes, mesenchyme and pigment cells. Since this pattern of cephalic structures de- veloped in the absence of a structured in- ductor and of any determination field of the host, it must have arisen by self-organiza- tion of the stimulated explants. The absence of a locally applied inductor expressed itself in the fact that the brain diverticula were not bilaterally symmetrical but multiple for- mations; they could be associated with as many as twelve olfactory placodes in one explant. It was thought at first that we were deal- ing once more with a “relay” mechanism, killed cells acting as inductors for the intact cells. If this were the case, then all further progress would have been stalled. However, neuralization was also observed in the ab- sence of any permanently damaged or dying cells (Holtfreter, ’47b). Obviously, the ex- ternal stimulus liberated the inductive agent within the reacting cells themselves by caus- ing initial and reversible steps of cytolysis. The similarity between this process in living 203 cells and the “unmasking” of the neuraliz- ing agent in killed cells is underlined by the fact that in either case many different injurious treatments can cause this libera- tion. The essential and common mechanism by which the cytolytic agents initiate neu- ralization of the explants seems to be: (1) increase of permeability of the cell mem- brane; (2) flooding of the cytoplasm with water and the electrolytes of standard so- lution; (3) dissociations and, after recovery from the shock, new combinations of certain cytoplasmic compounds whose specific syn- thetic activity would shift differentiation from epidermal to neural. It is unlikely that the initial steps of this mechanism occur in normal development also. Some of the inductive effects discussed above can conceivably be reinterpreted in the light of these experiments. For instance, neuralization by methylene blue, organic acids, water-soluble carcinogens or sulf- hydryl compounds could be due to a reversi- ble cell injury rather than to a relay mech- anism. The mechanism by which certain carcinogens exert their neuralizing effect may be similar to the unmasking effect of subcytolytic agents. Waddington and Good- hart (49) found that such a hydrocarbon becomes selectively fixed to the large gran- ular cell inclusions (lipochondria) which may entail the liberation of the neurogenic factor. The fact that an abnormally high water imbibition of the cell, caused by unspecific injurious agents, is apparently sufficient to elicit neuralization demonstrates once more the futility of the attempts to identify any of the experimentally applied chemicals with the normal neuralizing factor. On the other hand, our conviction is strengthened that the key to an understanding of the induc- tion phenomena is to be sought in the re- acting cells rather than in the inductors. This supposition led to a comparative study of some cytological phenomena in differen- tiating epidermal and neural cells. CELLULAR POLARITY, MOTILITY AND ADHESIVENESS AS RELATED TO INDUCTION The recognition of the polar structure of early embryonic cells is of fundamental im- portance for an understanding of neural in- duction. The entire surface of the embryo is coated with a protective film having a low permeability. Cellular motility is mainly confined to the inner uncoated surface, 274 which becomes the ameboid anterior pole when a cell migrates or exhibits “out- growth.” Inductive agents, whether repre- sented by formed inductors or by injurious aqueous media, act only when applied to the uncoated cell surfaces or after the coat has been dissolved (Holtfreter, "43a, ’48b). As a consequence of neural induction, ectodermal cells acquire the tendencies (1) to elongate reversibly into cylindrical bodies EMBRYOGENESIS: PROGRESSIVE DIFFERENTIATION uptake, suggested by Glaser (14, 716), do not control the neurulation process (Brown, Hamburger and Schmitt, ’41). Determination of a cell to become a neu- ron is immediately reflected in new tenden- cies for adhesiveness and ameboid motility (Holtfreter, °47a). An isolated Rohon-Beard neuroblast adopts a pear shape, its anterior pole projecting mobile filopodia into the liq- uid medium and its tapering posterior pole Fig. 107. Kinetic tendencies of isolated embryonic cells. a, b, c, Ectodermal cells, determined to become epidermis, exhibit flattening (epiboly) over the substrate; d, an isolated cell of the medullary plate elongates into a cylindrical body; e, f, isolated neuroblasts form dendritic processes. (After Holtfreter, ’48b.) (Fig. 107d), (2) to move from the periphery into the depth, (3) to detach themselves from non-neural cells. The same three tendencies appear in other induced ectodermal structures (lens, otocyst, pituitary, placodes). By con- trast, ectoderm cells which have not re- ceived inductive stimuli and are destined to form epidermis become spherical when float- ing in a culture medium; they flatten out when provided with a proper surface of contact, and they do not invaginate (Holt- freter, °47a, °48b; see Fig. 107a,b). These dif- ferent kinetic tendencies are due to forces inherent in each isolated cell and they are the first cytological indication that induction has occurred. According to the measurements of Gillette (44) these kinetic properties are instrumental in normal development: the medullary plate arises as the result of a proximodistal elongation and corresponding contraction of the outer surface of the dorsal ectoderm cells, while the prospective epi- dermis cells undergo a compensatory flatten- ing. Volume changes by differential water having a pronounced adhesiveness (Fig. 107e,f). Given an appropriate contact surface, a neuroblast may migrate ameba-fashion, but if the cell body remains attached, the ad- vancing ameboid projections are spun out to form branching processes, and the config- uration of a dendritic neuron is obtained (Harrison, °10). In a similar way, the morphogenesis of other cell types, their properties of aggrega- tion or dispersal, and the direction of their movements, are largely determined by their inherent axial organization. Stage-specific and tissue-specific differences of motility and adhesiveness are instrumental not only in cyto-differentiation but also in organ forma- tion. Holtfreter (46a, ’48b) has suggested that shape and kinetic properties of embry- onic cells result primarily from local and temporal variations of the expansibility of the cell membrane, reflecting changes in the composition and molecular arrangement of this structure which may be considered as being essentially an organized lipo-protein AMPHIBIANS film subject to various states of hydration. He assumed furthermore that induction and cytological determination in general involve the elaboration of new tissue-specific com- pounds, some of which become integrated localized elements of the cell membrane and are thus responsible for the different new types of cellular shape, motility, selective ad- hesiveness, and metabolism. The problem of how, and in which parts of the cell, these specific morphogenetic changes originate is a matter of conjecture (see discussion be- low). GENERAL CONSIDERATIONS Before entering into a discussion of some general notions on amphibian embryogene- sis, we should point out that our preceding presentation is somewhat unbalanced: much emphasis has been placed upon the phe- nomenon of induction, but other, equally important principles of organogenesis have not received their due share of consideration. We did stress the importance of morpho- genetic movements and of self-organization, but we have neglected such aspects as selec- tive cell adhesion; structural, mechanical, hydrostatic factors of development; space- time patterns of physiological processes and identified substances; problems of energy metabolism; phenomena of _ differential growth; genetic and comparative-embryolog- ical considerations. For information on these latter aspects, which are in many ways interconnected with the problems dealt with above, the reader is referred to other chap- ters of this book. In the following discussion our attention shall be mainly focussed upon the concepts of embryonic fields and gradients which have played a predominant role in the analy- sis of induction, and of organogenesis in general. GENERAL CHARACTERISTICS OF INDUCTION PHENOMENA Since the term “induction” has sometimes been used in a loose way, let us try to cir- cumscribe it for our present purposes. Unlike hormones, inductive stimuli operate only at certain stages, as a rule, during early de- velopment, and they are normally ineffective unless there is an intimate contact between inducing and reacting tissues. The effects of the inductive tissues are undeniable, since in their absence none of the ectodermal and probably few of the mesodermal differentia- tions would ever arise. Once stimulated, the 215 cells proceed along their new course of dif- ferentiation independently of a continued application of the inducing stimulus. The newly acquired characteristics are self-main- taining and handed on to subsequent cell generations. In this respect, too, the induc- tive stimuli differ from hormones which must be applied continuously in order to sustain the differentiations initiated by them (for further discussion of these and related problems, see Medawar, ’47). Thus, normally, “inductors” are living and as a rule embryonic tissues which deter- mine the cytological fate of the reacting, adjacent cells. This inauguration of a new trend of differentiation is almost invariably associated with new trends of kinetic activi- ties of the induced cells, such as invagina- tions, delaminations, new rates of cell divi- sion, etc. But kinetic activities may arise independently of inductive stimuli, as for instance when the blastoporal rim invagi- nates into the blastocoele, or when certain cells establish an epithelium in response to a favorable non-specific substratum. In the same sense, the transformation of cuboidal into temporarily elongated cells, which is practically always associated with ectoder- mal inductions (medullary plate, lens and other placodes: Ruffini, ’25; Lehmann, ’29; McKeehan, ’51), merely indicates that these cells proceed to invaginate. The invaginating cells of the blastoporal lip likewise stretch into long bodies; their change of shape is not due to an inducing substratum but to an inherent migration tendency of these cells (Holtfreter, ’44a). Columnar cell elonga- tions also appear frequently in intestinal epithelia and their glandular evaginations. Therefore, we cannot subscribe to the idea that there is necessarily a causal relation- ship between cell elongation and induction mediated by attracting and orienting mo- lecular forces acting across the inductive contact surface (Weiss, ’47, ’50). Hence, when speaking of induction, em- phasis should be placed upon the “material” and irreversible rather than upon the “dy- namic” and temporary cellular changes (Vogt, ’23—24) following this kind of stimu- lation. In the terms of Weiss (’39), the latter are merely “modulations,” that is, reversible manifestations of one among a variety of possible cellular adaptations, the range of which is determined genetically and often with the cooperation of inductors. Therefore, we are reluctant to call the epithelial per- forations produced by contact action in the mouth, pharynx and anus regions “induc- 276 EMBRYOGENESIS: PROGRESSIVE DIFFERENTIATION tions,” just as one would hesitate to apply this term to the secondary formative effects of the archenteron roof upon the bilaterality of the neural system, or to the accumulation of limb blastema cells following amputation or the insertion of a graft. The kinetic re- sponse of cells to certain stimuli in itself does not indicate the presence of inductive stimuli (see inflammatory reactions). Like- wise, to extend the term “induction” to the effects of external factors upon the rate of tissue growth (as opposed to cell deter- mination) would result in a confusion in terminology. There is no doubt that in the course of development new kinds of specific inductors come into play. They may involve the de- rivatives of all three germ layers, and these primordia may cooperate so as to form synergistic “inductor systems.” The differ- ent tissue components of such a system may act simultaneously or in succession but they must be arranged in a typical pattern in order to induce a single structure, such as a typical neural tube or an acoustic organ. At this point, we should remind the reader that the outcome of an induction is not only determined by the specific properties of the different inductors but just as much by those of the stimulated cells. The reaction potencies of the latter were found to be lim- ited by the following inherent factors: ge- netic potentialities, stage-specificity, and tissue-specificity. The two independent sets of properties, namely those of the inductors and those of the reacting cells, must be inter- locked in space and time in order to insure normal development. This is principally achieved by directed cellular mass-move- ments. COMMENTS ON THE TRANSMISSION OF INDUCTIVE STIMULI Although we can no longer doubt that induction involves some sort of an activating chemical mechanism, we really do not know the chemical nature and the mode of trans- mission of these stimuli. Schmitt (41) drew attention to the pos- sibility that the columnar shape of the neural plate cells might result from a mo- lecular zipper effect operating by means of dehydrating agents between the adjoining surfaces of these cells. This interpretation has become improbable in view of the fact mentioned above that isolated medullary cells can elongate reversibly without any surface of contact. But Schmitt applied considerations of molecular surface interactions also to the problem of the inductive mechanism, and other workers have speculated along similar lines. On the basis of physicochemical models, Needham (’42, p. 289) suggested that in- duction may not actually involve the trans- mission of a chemical agent but may be pri- marily a matter of polar orientation and attraction of long-chain molecules at the inductor-facing side of the reacting cell; this effect might reach into the deeper zones of the cell, creating an increasing complexity of structuration which would lead to neural differentiation. Holtfreter ('44b) also believed that interfacial attraction forces and surface adsorptions might be instrumental in the induction process. Weiss (47, ’49a,b, °50) has elaborated and expanded the concept of intercellular surface actions, making it part of a general hypothesis on the role of “molecular ecology” in morphogenesis. We cannot discuss here the full background and the ramifications of this highly interesting hypothesis but shall deal only with its ap- plication to induction. The concept of “molecular ecology” is stated by the author as follows: “Each cell and organized cell part (nucleus, chromo- some, etc.) consists of an array of molecular species whose densities, distribution, ar- rangement and groupings are determined by their mutual dependencies and interactions as well as by the physical conditions of the space they occupy. These species range from the elementary inorganic compounds to the most complex ‘key’ species characteristic of a given cell. Chemical segregation and _ lo- calization within the cell result from free molecular interplay, as only groups of ele- ments compatible with one another and with their environment can form durable unions. ... Among the principal segregative factors of molecular mixtures are interfaces. Inter- facial forces between partly immiscible molecular populations concentrate certain se- lected molecular species of the interior along the border. By their surface positions, these border species acquire power over the further behavior of the enclosed system. . . . Various considerations suggest that in biological sys- tems fixation of a given molecular species in a surface is not due solely to unspecific factors, such as surface tensions, adsorption, etc., but that in addition, highly selective chemical affinities are involved. These may be based on the steric interlocking of char- acteristically shaped end groups of the sur- face molecules of adjacent systems” (Weiss, ’49b, p. 476). AMPHIBIANS The highly specific molecular configura- tions at the cell surface are thought to play a particularly important role in differentia- tion, tissue affinities and embryonic induc- tion. Assuming that the cell surfaces of inductive tissue are characterized by specific, oriented end groups, “the molecular surface film of the inducing layer (e.g., retina) would selectively attract key molecules of complementary or otherwise conforming con- figuration from the interior of the cells of the adjacent layer (e.g., epidermis) and thus decide the further differentiation of these cells” (Weiss, “49b, p. 478). The incorpora- tion of certain key molecules in the surface would result in the competitive removal of others from the surface, and this redistri- bution would initiate a specific trend of differentiation within the reacting cell. The inductor itself would operate only on the interface of inducing and reacting tissue by creating on the surface of the latter a “steric conformance, i.e., complementary spatial configurations between molecules or certain exposed atomic groups of them enabling them to conjugate in key-lock fashion” (Weiss, °47, p. 256). A diffusion of molecules from the inducing cells to the interior of the reacting cells is not considered as an alternative mechan- ism. Since attempts to test the diffusibility of inductive agents of normal inductors by the use of separating membranes of known porosity have failed so far (Brachet, 50; McKeehan, °51), the old assumption that under normal conditions close contact be- tween inducing and reacting cells is neces- sary, remains unchallenged. * The above concept of Weiss meets with serious difficulties when we consider that freely floating ectodermal explants can be neuralized by different liquid media which have no fixed molecular configuration. This can occur in the absence of any disintegrat- ing cells (Holtfreter, ’47b). Consequently, there is no likelihood that, as Weiss (49a) suggests, an “inductive protein film” might become adsorbed at the surface of the float- ing explant. Furthermore, it is difficult to apply this hypothesis to the following ob- servations: transformation of noninductors, such as ectoderm or entoderm, into neural inductors by different killing procedures; * Very recently, Niu and Twitty (53) demon- strated that the inductive factors of living tissues are indeed diffusible. Ectoderm explants differentiated into neural and mesodermal tissues through the application of the salt solution in which pieces of embryonic inductors had previously been cultured. 277 the neuralizing capacity of such diverse sub- strata as the normal archenteron roof, adult organs of all sorts and their cell-free extracts. It can hardly be assumed that all of them possess the same specific molecular surface pattern to account for their common neural- izing effects. It appears, therefore, that the kinetic and differentiation tendencies of induced cells can arise independently of a molecule-orient- ing inductor and that the new cell proper- ties are evoked by the introduction of non-oriented chemical agents rather than by the orienting or attracting forces of a spe- cifically structured substratum. A cytotypical molecular configuration of the cell surface which seems to be very important for cell- specific motility, adhesion and differentia- tion, may very well arise secondarily and independently of a contact with other cells. Hence the burden of organizing the cellular constituents for cytodifferentiation would be carried by the reacting rather than by the inducing cells, a relationship similar to that between sperm and egg development or be- tween a hormone and the differentiations initiated by it. We do not question the fact that polarity, shape, arrangement and move- ments of all embryonic and adult cells in- cluding cellular “modulations” (Weiss, ’39) are subject to controlling factors of the en- vironment. However, the latter are merely subsidiary and not determinative factors. Their presence enables the cells to manifest their inherent potentialities which have been previously determined by cell-inherent or extraneous factors. THE QUESTION OF QUANTITATIVE VERSUS QUALITATIVE CHEMICAL DIFFERENCES BETWEEN THE DIFFERENT INDUCTORS Once the chemical nature of inductive processes was established, there arose the question: how does the diversity of induced structures come about? This question is very complex, as may be deduced from the fact that histologically different primordia can induce the same kind of tissue, and that a single primordium can induce one structure in one germ layer and another one in an- other germ layer. In a simplifying fashion this problem has been reduced to the ques- tion of whether the stimuli of the different inductors vary in a quantitative or in a qualitative sense. It should be realized that this question cannot be answered satisfactorily by any ex- 278 EMBRYOGENESIS: PROGRESSIVE DIFFERENTIATION periments with normal living inductors. The elucidation of this problem must wait until more is known about the chemical nature of the inducing agents, of their distribution in the embryo, and of their effects at dif- ferent concentrations. At present, only more or less plausible assumptions can be made. Quantitative Aspects. In connection with what will become evident from our later discussion, one may accept the view that differences in the size of the inductive sub- stratum, or in the concentration of the in- ductive agent, can bring about different induction patterns. Thus the formation of a brain, in contrast to that of a spinal cord, is perhaps due in part, though certainly not entirely, to differences in the size of the stimulated area. It is conceivable, further- more, that the different components of a synergistic “inductor system” release the same kind of inductive agent and that dif- ferential effects arise from differences in concentration, or in the spatial or temporal application of this agent. The fact that a lens normally arises in contact with the optic vesicle, but that, in the absence of an eye, a lens can be induced by various other tissues of the head (Ikeda, ’°38) may be ascribed to the existence of a widely extend- ing gradient of the lens-inducing agent. Within this gradient, the optic vesicle would constitute the dominating center which nor- mally prevents the appearance of accessory lenses. This consideration may be applicable to other structures, such as balancer or sucker. There is no convincing evidence that quantitative variations of one and the same inductive agent can change its effect-spec- ificity. It is true that progressively heat- inactivated neural inductors induce increas- ingly smaller and less organized formations. Nevertheless, the grafts remain neural in- ductors. This applies also to living induc- tors, such as the medullary plate: when increasingly smaller grafts of this tissue are used, the resulting inductions become smaller and less complex, but no shift to other types of inductions, as for instance mesodermal ones, occurs. When a neuraliz- ing hydrocarbon was applied at different concentrations, the histological type of the inductions remained unchanged (Shen, ’39). Many attempts have been made to deter- mine in what respect neuralizing grafts dif- fer from lens-inducing grafts, but no defi- nite conclusions could be drawn from these experiments. Pasteels (’40, °45) has inter- preted the successive cephalocaudal dis- appearance of entomesodermal tissues in centrifuged or chemically treated embryos by assuming that the treatments produce a progressive lowering of a “morphogenetic potential” or, in other words, a depression of the concentration gradient of “organi- sine.” This interpretation, however, still lacks a biochemical foundation. Similarly, at- tempts at ascribing the induction of neural tissue, as against neural crest derivatives, to different concentrations of “organisine” have remained questionable. Altogether, it appears unlikely that merely quantitative differences of one type of agent can account for the diversity of the induc- tion phenomena observed. The fact that of the wide range of reaction capabilities only one becomes manifest in a given case seems to indicate that at least some of the inductive stimuli differ from each other qualitatively and that they act selectively upon the array of multiple capabilities (competence). Qualitative Aspects. When normal trunk inductor or certain adult tissues are de- vitalized by heat, cold, or drying, they lose abruptly their capacity of inducing meso- dermal structures, but their neuralizing ca- pacity is fully retained, if not increased, and only after prolonged boiling in water, or exposure to temperatures above 100°, does the latter decrease slowly and finally disappear. On the other hand, treatment with various fat solvents does not appreciably reduce the mesodermizing but may strongly diminish the neuralizing capacities of adult tissues. Extraction with salt solutions or proteolytic enzymes nearly abolished both capacities. It appears, therefore, that some tissues contain simultaneously several kinds of inducing agents which can be eliminated selectively. On the basis of these and other findings, Holt- freter (34a), Chuang (739) and Toivonen (40) have distinguished between a “neuraliz- ing” and a “mesodermizing” agent. In most instances when mesodermal tis- sues arose under the influence of either normal or atypical inductors, they did not appear singly but as an array of different tissues which could exhibit the character- istic pattern of a trunk or tail (Fig. 104). These experiments therefore could not deter- mine whether or not the emergence of the different components of the chorda-meso- derm material is due to the action of differ- ent stimuli. However, the data obtained from the treatment of prospective mesoderm with lithium, urea, or ammonia have shown that some of these chemicals have a specific chor- AMPHIBIANS dalizing and others a somite-producing ef- fect. This seems to suggest that qualitative rather than quantitative differences in the inductive agents cause these divergent dif- ferentiations. It should be pointed out that, so far, no purified tissue extracts or chemicals have been found which transform ectoderm into mesodermal tissues, and the claim that the mesodermizing agent may be represented by proteins and the neuralizing agent by nu- cleic acids (Brachet, °45, ’50; Kuusi, °51) is based on inconclusive data. Since it has not been possible to convert neuralizing tissues or chemicals into in- ductors for non-neural structures, e.g., for those of lens, ear or teeth, it is likely that the latter owe their emergence to the action of still other agents. No doubt, temporal and regional changes of competence are involved in the distribution of these sec- ondary inductions, but it is questionable whether the principle of competence is suf- ficient to account for this regional diversity. Especially in cases of a two-step inductive mechanism, as is exemplified in the deter- mination of teeth, branchial cartilage and pituitary, where the histogenetic effect of the first stimulus differs strikingly from that of the subsequent stimulus, it seems plausi- ble to assume that successively different in- ductive agents are engaged. THE “ORGANIZER” AND THE INDUC- TIONS CONSIDERED AS MORPHO- GENETIC FIELDS The main difficulty in the analysis of in- duction arises from the fact that in most cases both the inductors and the induced material originally constitute “morpho- genetic fields” which defy further break- down into localizable subunits. We are dealing with complex dynamic systems and not with an assembly of independent fixed primordia. Let us briefly retrace these dif- ficulties and then try to untangle them as far as it seems possible at this moment. Field Characteristics of the “Organizer.” The concept of ‘morphogenetic fields” (Spemann, ’21a; Gurwitsch, ’22; Weiss, ’26, °39) emerged from the older concept of “harmonious-equipotential systems” (see Driesch, ’29), which stressed the regulative capacities inherent in the majority of em- bryonic systems. However, the doctrinal formulations of Driesch barely did justice to the actual conditions in the embryo (see the critical discussion of Needham, ’42, p. 219 119). A typical example of a morphogenetic field is the chorda-mesodermal area of the early gastrula: any isolated part of it tends to regulate into a well-proportioned axial system comprising considerably more kinds of tissues than would arise from this mate- rial in a normal embryo; parts of this field can be removed or interchanged without causing abnormal development; two chorda- mesoderm fields can be fused to form a sin- gle axial system. This means that the histological fate of any part of the field de- pends upon its topographic relationship to the other parts of the field. There are many other examples of such regulative fields both in vertebrate and invertebrate development (see Huxley and De Beer, ’34; Weiss, 739). We do not know how the chorda-meso- derm field comes into existence, but we do know that there are complicated determina- tive inter-relationships between the constitu- ents of this field leading to its segregative differentiation. This poses the question of whether the determinative effects within the field involve principally the same physi- ological mechanisms as are operative in in- duction. The latter is characterized by a unidirectional action of inducing upon reacting tissues. In a morphogenetic field, however, there seem to exist reciprocal de- terminative actions which are not strictly confined to close contact-relationships. Fur- thermore, there was evidence that the ectoderm of the gastrula, itself not an in- ductor, contributes to the fixation of the boundaries of the chorda-mesoderm field. Contrary to the conditions in ordinary in- duction, this ectodermal effect would have to operate in a tangential direction and not be- tween two superimposed cell layers. These considerations are a challenge to find common principles which govern both the inductions and the self-organization which occur within a morphogenetic field. It is a peculiar property of the chorda- mesoderm field (“‘trunk-tail organizer”) and of the dorsal blastoporal lip proper (“head organizer”) to induce adjacent tissues, par- ticularly the overlying prospective ectoderm, to start new trends of differentiation. This capability of induction should not be in- cluded in a general definition of morpho- genetic fields (see Waddington, 34; Weiss, 35). However, in vertebrate development, these two properties—self-organization of the primary inductors, and their regionally specific inductive capacities—are intricately associated. It would be entirely misleading to con- 280 EMBRYOGENESIS: PROGRESSIVE DIFFERENTIATION ceive of the “organizer” material as of a kind of general manager which determines the destiny of the entire remainder of the embryo. Large portions of the embryo are not subject to its influence and, as has been pointed out already, the demarcation and invagination of the “organizer” itself are largely controlled by factors outside this area. The “organizer” has suffered further de- valuation. It is true that when part of this area is grafted into another gastrula it not only reorganizes itself into an incomplete axial system but it supplements its own dif- ferentiations by way of “assimilative in- duction,” i.e., adjacent host cells (mainly mesoderm) become harmoniously incorpo- rated into the system. Rather than consider- ing this integrating action as a unique prop- erty of the “organizer,” one may interpret it as just another demonstration of the char- acteristics of a morphogenetic field. The graft extends its field properties into the adjacent non-determined host cells, probably because it is merely a fragment of the “or- ganizer” and as such in an “unsaturated” condition. In the unfragmented normal em- bryo there is no necessity for assimilative induction. There, the organizing effect of the “organizer” is confined to its own ma- terial, namely the chorda-mesoderm man- tle.* While the “organizer” becomes segregated into distinct tissue primordia it loses the regulation tendencies of a field and, at the same time, the capacity of assimilating ex- perimentally supplied additional material into its system. Yet these primordia, let us say of an advanced neurula, retain strong and complex inductive powers. When ex- perimentally confronted with gastrula ec- toderm, they may induce a complete tail, or brain and sense organs. However, the struc- tural pattern of these inductions has no re- lationship to that of the inductors; it is autonomously organized. Thus the deriva- tives of the “organizer” behave essentially like dead or adult tissues which can likewise induce highly complex organ systems but do not contribute to the patterning of their inductions. Field Characteristics of the Inductions. It has been argued that there are two kinds of in- ductive agents: (1) “evocators” (Needham, * Some difficulty of interpretation arises in the case of the induction of tail somites by the posterior part of the archenteron roof. This process might be considered as a rather belated assimilative field effect of the “organizer.” Waddington and Needham, °34), which are supposedly present in normal inductors as well as in dead tissues and certain chemi- cals, and which evoke merely amorphous cell masses; (2) “modulators” (Waddington, 40) or “eidogens” (Needham, *42), which are present in normal inductors only and are thought to specify the induction so as to acquire organotypical patterns (‘‘indi- viduation”’). This concept has been discussed recently by Holtfreter (51). We shall con- fine ourselves to brief statements which are largely based upon the data presented in the preceding chapters. It has been shown that fresh or dead adult tissues can call forth perfectly individuated inductions outside of a whole host, namely in an isolated piece of ectoderm (Figs. 93-96). This excludes the necessity of an interven- tion of “eidogens” for the emergence of highly complex and normal anatomical pat- terns. Since the atypical inductors possess no structure which they have in common among themselves or with the inductions, it seems reasonable to conclude that they simply ac- tivate the ectoderm to establish a new head or tail field which then organizes itself into typical tissue patterns. If there are such sub- stances as individuating “eidogens,” they did not come from the inductors but arose de novo within the induced ectoderm. This would make it futile to attempt a distinc- tion between “inducing” and “evocating” agents. There is good reason to assume that in normal development the archenteron roof acts not much differently from a dead in- ductor. The neural plate is induced in the form of new complex field systems which have no point-to-point relationship to the inductors and are capable of organizing themselves into anatomical patterns. This independence applies also to secondary in- ductions (lens, ear vesicle), since their complex structures have evidently no coun- terpart in the structural properties of their inductors. Striking evidence for this notion that independent fields and not specific or- gans or structures are induced has been provided by the xenoplastic experiments. They showed that although the parts of the archenteron roof have regionally specific effects, the elaboration of anatomical and histological patterns is due to inherent prop- erties of the reacting material. In this re- spect, therefore, the inductive action of the “organizer” may be compared with the ac- tion of hormones which can stimulate certain tissues to undergo specific differentiations AMPHIBIANS but which have no control over the organ- ized pattern of these differentiations. To conclude, then, the “organizer” has the characteristics of a morphogenetic field which is, however, not really harmonious- equipotential in the strict definition of Driesch, and it induces another, or several other, fields, which are likewise capable of regulation and of self-organization. Self-organization of Fields. We have had several occasions to emphasize the important role which the self-organization of mor- phogenetic fields plays in progressive dif- ferentiation: an area which is capable of self-differentiation and of regulation breaks up into smaller units which may represent fields on a smaller scale and with more re- stricted differentiation potencies. They, in turn, may subsequently be subjected to further segregation, until the final organiza- tion of an organ is achieved. One of the outstanding characteristics of the process of “self-organization” is its autonomous char- acter; it is illustrated by the morphogenetic behavior of the chorda-mesoderm field (Fig. 81) and the limb field, and by the highly organized differentiations induced by adult organs (Figs. 100, 103, 104) and in xeno- plastic combinations (Fig. 97). This issue, and its significance for verte- brate development, has been recognized early by different investigators. Weiss has stressed its importance under the heading of “autonomization” ((26) and “emancipa- tion” (735, °39); Lillie (27) has used the term “embryonic segregation” and Lehmann C42b) the term “autonomous self-organiza- tion.” The earlier cell-lineage studies and isolation experiments on invertebrate eggs had already focussed the attention to this principle. In contrast to the extensive analysis to which embryonic induction has been sub- jected in the last 50 years, our information on the mechanisms involved in the process of self-organization is negligible. Tentative approaches to this problem have been made. In connection with his investigations on the development of the limb and otocyst in Amblystoma, Harrison (745) has suggested that the emergence of axial structuration in a field might be based on the presence of a supracellular paracrystalline lattice of di- polar molecules. Spiegelman and Steinbach (45) have interpreted the differentiation of morphogenetic fields in terms of physiologi- cal competition between transforming cells, in the framework of a more general and more elaborate theory which we cannot dis- 281 cuss in detail. Weiss (750) points out that neither electrodynamic theories nor simple concentration gradients nor differences in position and exposure of cell groups can ac- count for the complex behavior of segregat- ing fields. He suggests (’50, p. 194) that “It may become necessary to assume that at any given point of a particular field, conditions of such specific constellation arise that cer- tain molecular groupings will be selectively favored or energized by a sort of resonance relation between field and molecular pat- tern.” The lack of precision and the diversity of these notions testify to our present-day ignorance of the physicochemical factors op- erating in the process of field segregation. But the progress which is being made in the elucidation of some instances of embry- onic fields (see below) is encouraging enough to support the belief that the crucial and universal problem of self-organization is not entirely refractory to further analy- sis. Morphogenetic Movements as an Organizing Principle. The patterning of an embryonic field involves more mechanisms than deter- minative (inductive?) interactions of its parts. In some instances, implication of the latter is even doubtful. For example, the re-individuation of a disaggregated adult sponge is mainly, if not entirely, mediated by the principles of directed movements and selective adhesion of the different types of cells. Corresponding morphogenetic cell movements are of equal importance for the tissue patterning of the chorda-mesoderm field. That these movements of invagination, spreading or stretching which make for the regroupings and final segregation of the tissue components are inherent properties of the cells concerned is clearly demonstrated in explants from the blastoporal region, especially if such explants have been previ- ously disaggregated. Embryonic Fields as Related to Their Initial Mass. Graded inactivation by heat of either normal or atypical inductors reduces their effectiveness in a quantitative as well as a qualitative sense. The inductions not only become progressively smaller but they ex- hibit less organized patterns until they have the aspects of non-specific neural or neuroid cell groups. This reduction and final loss of patterning seems to show that the organo- logical complexity of a neurogenic field de- pends primarily on the number of cells which have experienced stimulation. In other words, the size of the field appears to 282 EMBRYOGENESIS: PROGRESSIVE DIFFERENTIATION be a determining factor in its future organ pattern. Other observations tie up with this notion. Lopashoy (’35a) states that the differentia- tions produced by explants from the dorsal blastoporal region increase in complexity with the number of fused identical explants. It is well known, furthermore, not only that if the cephalic entomesoderm is progressively reduced in mass it induces increasingly smaller head organs, but also that there are lower limits beyond which well proportioned and complete head structures can no longer be formed. Instead of decreasing proportion- ally in all its parts, the tissue pattern of the reduced heads changes qualitatively. Thus, in the graded series of micro- to anencephalic animals, the lateral structures, such as gills, balancers, eyes, or ear vesicles, become shifted mediad, then appear as single organs, and finally drop out entirely. But even in the complete absence of eyes, the dien- cephalon may still exhibit a rather normal structure (Nieuwkoop, °47). These reces- sions in a lateromedian direction are invari- ably associated with a stepwise reduction and final absence of all head structures in a cephalocaudal direction. Lehmann (745, °48) circumscribes these changes of pattern in terms of “Realisations- stufen” and he points out that corresponding situations exist in other blastemas, such as those of the ear vesicle (Andres, ’48) or the limb (Bretscher, 49). As a matter of fact, this relationship between initial size of the field and the diversity of its final differentiations has been found previously in many other in- stances, especially in invertebrates, where this phenomenon was described under head- ings such as vegetative budding, regenera- tion, or reconstitution of disarranged organis- mic fields (for references see Child, °41; Huxley and De Beer, *34; Berrill, 41; Holt- freter, 51). In all these instances, the initial size of the tield seemed to determine its fu- ture organological complexity, each field having its own series of critical size thresh- olds which determine type and configura- tion of the tissues that will emerge from the originally pluripotential mass of cells. AMPHIBIAN DEVELOPMENT AS RELATED TO CHEMICAL GRADIENTS Child’s Gradient Theory, according to which structural patterns are preceded and deter- mined by axial gradients of metabolic activ- ity, is well known and need not be elaborated here. This theory, which has its main and legitimate field of application in reconstitu- tion processes of invertebrates, has also been applied to amphibian development (general references: Huxley and De Beer, *34; Child, ’41, °46). From what follows it appears, how- ever, that the factual basis for this latter application is rather slim. Let us first acknowledge that not all em- bryonic fields originate through induction. In an aggregation of a_ sufficiently large number of amebocytes of Dictyostelium (Raper, *41), a field arises “autonomously,” that is, under “unspecific” external condi- tions. A corresponding self-establishment of a field appears to occur in an isolated piece from any region of the gastrula ectoderm of anurans which tends to segregate into epi- dermal and sucker cells (Holtfreter, *33a, 36, ’38c; Yamada, ’38; Raunich, 42a). As in the classical experiments on hydrozoans and worms, the mere act of “physiological isolation” from controlling factors of the whole organism (Child, °15, ’41) seems to create the new field. No other examples of this sort are known in amphibian development, and it has been mentioned already that no new fields could be produced simply by raising the rate of metabolism of an embryonic district. The production of a brain-field in ectoderm ex- plants by means of injurious treatments cer- tainly resembles the individuations obtained by Child (41) in injured hydrozoa, but it is doubtful that Child’s rather vague con- cepts of differential susceptibility, recovery and dominance are of much help for a con- crete physiological interpretation of the above results. Principally on the basis of the data of Bellamy (719) on the regionally different susceptibility of amphibian gastrulae to the cell-dispersing action of potassium cyanide, Child postulated a region of ‘“‘physiological dominance” in the ectoderm around the ani- mal pole, and a dorsoventral gradient with its center above the dorsal blastoporal lip. The differential susceptibility to potassium cyanide was interpreted as indicating dif- ferences of oxygen consumption. However, later workers refuted this interpretation (Buchanan, ’29; Holtfreter, ’43a). Subse- quent quantitative determinations did reveal regional differences, although not exactly gradients, in the distribution of certain com- pounds, or physiological processes, within the early amphibian embryo. This was shown in the case of sulfhydryl compounds, RNA, alkaline phosphatase, degree of reducing power, glycolytic activity, and rate of oxy- AMPHIBIANS gen consumption, all of which follow more or less the same pattern of distribution (for references see Brachet, ’45; Boell, ’48). These patterns, however, do not coincide at all with those of the inducing fields or their induc- tions. All they appear to show are regional differences in the rate of metabolic activity of the various germ layers or primordia. These differences seem to be largely a re- flection of tissue-specific differences in the ratio between inert yolk and physiologically active cytoplasmic components. Some parts of the embryo—notably the ectoderm, which is poor in yolk reserves—differentiate com- paratively faster than others, but the bio- chemical data have so far failed to throw any light upon the phenomena of induction, regulation or tissue determination. Even if there were a close parallelism between met- abolically especially active and morphogenet- ically “dominant” regions in the sense of Child, one may still advance the argument of Spemann (38) and others that it is diffi- cult to decide whether such metabolic pat- terns are the cause or the effect of tissue determination. The Double Gradient Theory of Daleq and Pasteels (737, °38; Dalcq, 41a) follows similar lines. These authors have postulated that all embryonic tissue differentiations, including the intramesodermal segregations, are caused by one and the same hypothetical agent (“‘or- ganisine”) and that qualitative differences between the inductions are due to different concentrations of this agent. The premises of this gradient concept and some of its ap- plications to embryological problems have already been discussed in previous chapters. No doubt this concept aims to be all-embrac- ing. It attempts to attribute such diverse phenomena as egg organization, morphoge- netic movements, field segregation, and re- gional induction to the interplay of just two hypothetical factors, a cortical and a vitel- line factor, whose interactions would result in the establishment of the aforementioned gradient of “organisine” that pervades the whole embryo. It is impossible in this re- view to evaluate critically the factual and theoretical aspects of this hypothesis. It is based upon assumptions which seem to be controversial or arbitrary, and some of the interpretations offered are merely circum- scriptions of the problems to be solved. Similar criticisms were raised by Rotmann (43). We assert again that in most instances when this concept has been applied to cer- tain observations, other hypotheses would serve as well, if not better. It seems that too 283 many unrelated, though partly overlapping, processes are engaged in embryogenesis to allow for their unitary interpretation in terms of an oversimplified gradient concept. Yamada (’50a, b) has proposed a different version of a double gradient theory. He postulates two qualitatively different activ- ities distributed in a gradient fashion and changing their values with time. One of these “morphogenetic potentials’ would be re- lated to the dorsoventral pattern of organiza- tion in ectoderm and mesoderm and have its highest concentration at the dorsal side. The other would be related to stretching and convergence activities; it would be repre- sented by a cephalocaudal gradient with its peak at the caudal end. The specific differen- tiation of a germ area would be determined by the combined effects of both. The spatiotemporal patterns of both poten- tials are thought to be controlled by extrinsic as well as intrinsic factors and therefore modifiable experimentally. For instance, the dorsoventral and cephalocaudal potentials of ventral ectoderm remain low when the ectoderm is isolated, resulting in atypical epidermis. A change of pH results in brain- like differentiations which are interpreted as a raise of the d-v potential (“dorsalization”’ ) without a change of the cephalocaudal poten- tial. The same ventral ectoderm, when com- bined with guinea pig kidney, differentiates into tail-like structures including spinal cord and somites, which is considered to be the result of the raising of both potentials (‘“dor- salization” and ‘“‘caudalization”). Other ex- periments are interpreted along the same lines. However, the experimental data on nor- mal and atypical inductions can be inter- preted equally well in terms of “neuralizing” and “mesodermizing” inductors, and more factual evidence would be required for a critical evaluation of this hypothesis. Inside-Outside Gradients as Determining Factors in the Organization of Embryonic Fields. Concentration gradients involving not just one, but many kinds of chemical sub- stances as well as more complex organic entities are bound to develop in any embry- ological system in response to its environ- ment (Gibbs phenomenon). The concentric organization of an amphibian egg into ovo- plasm, pigmented cortex and coat can be partly attributed to this principle of sorting out of surface-active substances and their subsequent reactions with each other and with external factors. Such inside-outside gradients are significant even in the multi- cellular embryonic stages where the inward 284 EMBRYOGENESIS: PROGRESSIVE DIFFERENTIATION movement of certain primordia and the out- ward movement and epibolic spreading of others can hardly be explained without in- voking chemical gradients of some sort which control the direction of the cellular migra- tions (Holtfreter, 44a). In a stimulating article dealing with the individuation of slime molds from the aggre- gation of equipotential myxamebae, Cohen (42) points out that in response to the ex- ternal inorganic medium, an initially homo- geneous cell aggregate most likely elaborates concentric patterns of distribution and con- centration of different substances. Critical levels of pH, oxygen tension, salts and other diffusible compounds will be established in certain regions; in this way, the necessary conditions are provided for chemical proc- esses that cannot occur in other regions. From these reactions new patterns of com- plexity will arise which in turn initiate further local structurations. Provided the differentially distributed new compounds are of cell-determinative significance, it is quite conceivable that the initially merely quanti- tative gradations play a fundamental role in the segregation of an originally equi- and pluripotential cell-aggregate into definite tissue patterns. We believe that such considerations will be very helpful in the analysis of the factors responsible for the self-organization of the various fields in the embryos of higher organ- isms. Clearly, the location of a prospective field within the organism and the size of its cell population would have a_ decisive bearing on the establishment and effective- ness of such inside-outside gradients. Further- more, regional differences in the accessibility and composition of the environmental fac- tors would determine the outcome. It is important to realize that the struc- turation of a field can be inaugurated both by “unspecific” and “specific,” that is, induc- tive, factors. Globular cell-aggregates of a po- tential field which float freely in an “‘unspec- ific’” medium may be expected to elaborate a concentric and radially symmetrical organi- zation; this is exemplified in the multiple brain formations of shock-activated ectoderm explants. Local attachment to a substratum, even if it is inert, would introduce polarity into the system. This is demonstrated not only in the basal-apical polarization of at- tached cell aggregates of sponges, slime molds and hydrozoa (Child, ’41), but also in the axial organization of embryonic fields of vertebrates which are exposed to a dif- ferential of external conditions. The neu- ralized ectoderm explants develop an apical- basal polarity when attached to glass. Simi- larly, the elaboration of a bilateral symmetry in glass-attached explants of blastoporal ma- terial (Fig. 81) is undoubtedly enhanced by external differentials, since freely float- ing explants of this kind fail to manifest their tendency for axial organization. When inductors of any kind are grafted into a whole embryo their action is more or less unidirectional upon the overlying tissues, which simulates the conditions in normal development. In either case, the induction tends to establish a bilateral symmetry and dorsoventral polarity. But such patterns are very rare when the same inductors are placed into a mantle of isolated ectoderm, thus affecting the latter throughout its cir- cumference; under these topographic con- ditions an irregular multiplicity of struc- tures tends to appear. Obviously, and this is in accord with Child’s ideas, the emergence of axial patterns in a prospective field re- quires the application of external factors in certain directions. This seems to be true not only of “unspecific” environmental condi- tions but still more so of inductive stimuli. CYTOPLASMIC FACTORS OF DIFFER- ENTIATION It would be desirable to connect the data and concepts derived from the study of am- phibians with those obtained in other fields of biology so as to arrive at a generalizing hypothesis of the factors that determine cellular differentiation. This enterprise is, however, too involved to be tackled satis- factorily within the limited space available. We propose, nevertheless, to venture a few steps into this little explored yet very stimu- lating field of speculation. There seems to be agreement among em- bryologists that the regionally different fate of the parts of an egg, and of its subsequent embryonic fields, is determined by cyto- plasmic rather than nuclear differences of the cells concerned. One cannot doubt that the nuclear genes control the emergence and maintenance of the cytoplasmic differentia- tions, yet there is an impressive body of evidence indicating that once local cyto- plasmic differences are established, they may for long periods become relatively inde- pendent of continued gene control. (See, for instance, Hadorn, °36; Sonneborn, °47; Rhoades, ’49.) This has led to the supposi- tion that, apart from self-reproducing genes, there are cytoplasmic entities, designated AMPHIBIANS by the various authors as plasmones, cyto- genes, plasmagenes or gene products, which become likewise capable of reduplication and mutation, and which are directly re- sponsible for the physiological and morpho- logical properties of the various types of differentiating cells (Wright, ’41, 45; Dar- lington, ’44; Haddow, °44; and others). It has been suggested by several authors that the hypothetical plasmagenes are com- parable to, or located within, the basophilic eranules (mitochondria, microsomes) re- ferred to above, which in turn share some properties with the virus (for further ref- erences see Brachet, ’45, °50). Needham (42) compares homoiogenetic induction with the mechanism of virus infection. Many workers are inclined to think that some tumors orig- inate from the transformation of normal microsome-like entities into virus particles. This idea seems to be supported by the find- ings of Claude (740, °41), who isolated from normal chick tissues small nucleoprotein particles which could not be distinguished chemically from the infective virus which renders these tissues malignant. Thanks to the investigations of Claude (46), Brachet and collaborators (40, 42, 44) and many others, it has been established that cyto- plasmic granules of a comparable composi- tion, though of various sizes, are present in at least the majority of embryonic and adult tissues. Many of the essential enzymes have been found to be localized in the larger granules (mitochondria), and it has been suggested that the alleged capability of re- duplication of these granules is due to their richness in ribonucleic acid. It will be recalled that Brachet attaches great importance to the function of the nucleoprotein granules in morphogenesis. Al- though we were reluctant to accept the data now available as evidence to show that any of the described constituents of these gran- ules are specifically engaged in neural in- duction, one can hardly doubt that in view of their being the carriers of so many impor- tant compounds, the granules play an essen- tial role not only in the _ metabolic specification of adult tissues but also in the developmental elaboration of the different tissues. On the basis of his studies on enzymatic adaptation of yeasts, Spiegelman (’48) has arrived at similar ideas and has proposed a concrete scheme which he considers to be applicable to embryonic differentiation. The chief merit of this scheme consists in its attempt to translate the notions of potential- 285 ity, competence and induction into the more tangible terms of enzymology. As is pointed out, it is the uniqueness of the enzyme patterns more than anything else that dis- tinguishes the tissues from each other. Ac- cording to Spiegelman, genes determine merely the potentiality of enzyme formation, but whether or not a particular enzyme is actually formed in the cytoplasm depends upon other factors, of which the substrate is obviously one. It is assumed that en- zyme formation is governed by autosynthetic reduplication and that in this _ process the various enzyme-forming precursors or ‘“plasmagenes” compete with each other for nitrogenous compounds. An externally applied specific substrate is presumed to combine with one of the inactive and rather unstable precursors to form a specific plasma- gene-enzyme system which thereafter can reproduce itself faithfully. Thus the sub- strate enhances the continued production and accumulation of one type of enzyme at the expense of other potential enzyme-precursors which through competition are more or less crowded out. However, owing to cytoplasmic patterns of distribution and to certain “sym- biotic” relations and ecological interactions within the population of the competing en- tities, there remains room for the co-exist- ence of different enzymes within a cell. Just the same, “the fate of any given cell during morphogenesis will be determined by the outcome of the competitive interactions amongst the initial plasmagene population.” (Spiegelman, °48; see also Steinbach and Moog, Section III, Chap. 2 of this book.) If plasmagene population means “com- petence,” and activating substrate “induc- tor,’ then the above considerations may serve as a model for interpreting the relations between these phenomena. Competence may then be ascribed to the stage-specific pre- ponderance of certain enzymatic or morpho- genetic precursors which can combine se- lectively with certain inductive agents. By doing so, the newly formed self-reproducing compounds would gain an advantage over the other competing precursors, and the cell would progressively narrow down its competence to become eventually unipotent. But since competence changes with time independently of external stimulation, either progressive transformation of the plasma- genes themselves, or their selective elimina- tion through cell-bound competition must be assumed. The prevailing irreversibility and “canalization” of differentiation (Wad- dington, 48) would result from the “survi- 286 val of the fittest,” namely the accumulative and self-reinforcing propagation of tissue- specific compounds extending over successive cell generations. Yet some cells, especially those capable of sexual or asexual reproduc- tion, remain omnipotent throughout the life cycle. How did they avoid specialization? Other cells, having become incapable of shifting into a new trend of normal differ- entiation, may still react to proper stimuli by “mutating” cytoplasmically into cancer- ous tissues (Graffi, °40; Haddow, ’44; Potter, 45; Holtfreter, 48a). This comparison of enzymatic adaptation of yeasts with tissue determination has some weaknesses: whereas the former requires con- tinued external application of a_ specific substrate (as in the case of hormones), em- bryonic tissues proceed to differentiate inde- pendently of the initial inductive stimulus. Moreover, it remains to be explained how ectoderm explants can be switched into neural differentiation simply by an unspe- cific injury. Finally, some primordia, espe- cially those of the entoderm, do not seem to require exogenous inductive stimuli for their normal differentiation. To fit these data into the above concept it seems necessary to assume that the sub- strates required for the autocatalytic syn- thesis of specific tissue proteins pre-exist in the reacting cells, and that the inducing stimulus acts neither like a virus nor as an enzymatic substrate but as an agent which activates or liberates certain intra- cellular substrates. The “unmasking” of the neuralizing agent in killed or slightly in- jured ectoderm may be due to autolytic dis- sociations comparable to those which occur in dying amphibian cells (Holtfreter, ’48a; Brachet, ’49, °50) and in the isolated nucleo- protein granules (Claude, 46). It has also been suggested that some enzymes are in- active in the living cell because they are separated from their proper substrates. This may give a clue to the observations of Pas- teels (47a,b; 49a) that centrifuged blastu- lae give rise to accessory neural and meso- dermal structures in the affected ectoderm. The cell content of the latter becomes clearly stratified, suggesting that these ‘‘auto-induc- tions” (Holtfreter, *47b, ’48a) may origi- nate from a mechanically produced union and reaction of morphogenetic precursors with certain substrates present in the same cell. In a thoughtful article, Waddington (’48) has discussed these problems from a some- what different angle, emphasizing likewise EMBRYOGENESIS: PROGRESSIVE DIFFERENTIATION that one must assume the existence of dif- ferent intracellular substrates and gene prod- ucts, subject to competitive interactions, in order to account for the autocatalytic syn- thesis of tissue-specifying protein compounds. At present, such attempts to arrive at gen- eral concepts by coordinating genetic, physi- ological and embryological data are liable to be of a highly speculative nature. How- ever, they suggest outlines for future research, REFERENCES Adelmann, H. B. 1930 Experimental studies on the development of the eye. III. The effect of the substrate (‘““Unterlagerung”) on the heterotopic development of median and lateral strips of the anterior end of the neural plate of Amblystoma. J. Exp. Zool., 57:223-281. 1932 The development of the prechordal plate and mesoderm of Amblystoma punctatum. J. Morph., 54:1-67. 1934 A study of cyclopia in Amblystoma punctatum with special reference to the meso- derm. J. Exp. Zool., 67:219-281. 1936 The problem of cyclopia. Quart. Rev. Biol., 77:161-182. 1937 Experimental studies on the devel- opment of the eye. IV. The effect of the partial and complete excision of the prechordal substrate on the development of the eyes of Amblystoma punctatum. J. Exp. Zool., 75:199-227. Alderman, A. L. 1935 The determination of the eye in the anuran, Hyla regilla. J. Exp. Zool., 70: 205-232. 1938 A factor influencing the bilater- ality of the eye rudiment in Hyla regilla. Anat. Rec., 72:297-302. Ancel, P., and Vintemberger, P. 1948 Recherches sur le déterminisme de la symétrie bilaterale dans Voeuf des amphibiens. Bull. Biol. France et Belg., Suppl. 37:1-182. Andres, G. 1948 Realisationsgrade bei der Ent- wicklung des Amphibienlabyrinths. Arch. J. Klaus-Stiftung., 23:562-568. 1949 Untersuchungen an Chimiaren von Triton und Bombinator. I. Entwicklung xeno- plastischer Labyrinthe und Kopfganglien. Gen- etica, 24:1-148. Atlas, M. 1935 The effect of temperature on the development of Rana pipiens. Physiol. Zool., 8: 290-310. Aufsess, A. von 1941 Defekt- und Isolationsver- suche an der Medullarplatte und ihrer Unter- lagerung an Triton alpestris- und Amblystoma- Keimen, mit besonderer Beriicksichtigung der Rumpf- und Schwanzregion. Roux’ Arch. Entw.- mech., 141:248-339. Balinsky, B. I. 1947 Kinematik des entoder- malen Materials bei der Gestaltung der wichtig- sten Teile des Darmkanals bei den Amphibien. Roux’ Arch. Entw.-mech., 743:127-166. Baltzer, F. 1941 Untersuchungen an Chimaren AMPHIBIANS von Urodelen und Hyla. Rev. suisse Zool., 48: 413-482. Baltzer, F. 1950a Chimaren und Merogone bei Amphibien. Rev. suisse Zool., 57:93-114. 1950b Entwicklungsphysiologische Be- trachtungen iiber Probleme der Homologie und Evolution. Rev. suisse Zool., 57:451—-477. Banki, O. 1927a Die Lagebeziehungen der Sper- mium-Eintrittsstelle zur Medianebene und zur ersten Furche, nach Versuchen mit 6rtlicher Vi- talfarbung am Axolotlei. Anat. Anz., 63: Erg. H., 198-209. 1927b Die Entstehung der ausseren Zeich- en der bilateralen Symmetrie am Axolotlei; nach Versuchen mit 6rtlicher Vitalfarbung. Verh. X. Internat. Zool. Kongress Budapest, pp. 375-384. Barth, L. G. 1939 The chemical nature of the amphibian organizer: III. Stimulation of the pre- sumptive epidermis of Ambystoma by means of cell extracts and chemical substances. Physiol. Zool., 12:22-29. 1941 Neural differentiation without or- ganizer. J. Exp. Zool., 87:371-384. , and Graff, S. 1938 The chemical nature of the amphibian organizer. Cold Spring Harbor Symp. Quant. Biol., 6:385-391. Bautzmann, H. 1926 Experimentelle Untersuch- ungen zur Abgrenzung des Organisationszen- trums bei Triton taeniatus. Roux’ Arch. Entw.- mech., 708:283-321. 1928 Experimentelle Untersuchungen liber die Induktionsfahigkeit von Chorda und Mesoderm bei Triton. Roux’ Arch. Entw.-mech., 114:177-225. 1929 Uber Induktion durch vordere und hintere Chorda der Neurula. Roux’ Arch. Entw.- mech., 779:1-46. 1933. Uber Determinationsgrad und Wir- kungsbeziehungen der Randzonenteilanlagen (Chorda, Ursegmente, Seitenplatten und Kopf- darmanlage) bei Urodelen und Anuren. Roux’ Arch. Entw.-mech., 128:666—765. , Holtfreter, J., Spemann, H., and Mangold, O. 1932 Versuche zur Analyse der Induktions- mittel in der Embryonalentwicklung. Natur- wiss., 20:972-974. Beatty, R. A., DeJong, S., and Zielinski, M. A. 1939 Experiments on the effect of dyes on in- duction and respiration in the amphibian gastrula. J. Exp. Biol., 76:150-154. Bellamy, A. W. 1919 Differential susceptibility as a basis for modification and control of early development in the frog. Biol. Bull., 37:312-361. Berrill, N. J. 1941 Spatial and temporal growth patterns in colonial organisms. Growth (Suppl.), 5:89-111. Bijtel, H. 1931 Uber die Entwicklung des Schwanzes bei Amphibien. Roux’ Arch. Entw.- mech., 725:448-486. 1936 Die Mesodermbildungspotenzen der hinteren Medullarplattenbezirke bei Ambly- stoma mexicanum in Bezug auf die Schwanzbil- dung. Roux’ Arch. Entw.-mech., 734:262-282. Boell, E. J. 1948 Biochemical differentiation dur- ing amphibian development. Ann. New York Acad. Sci., 49:773-800. 287 , and Shen, S. C. 1944 Functional differ- entiation in embryonic development: I. Cholin- esterase activity of induced neural structures in Amblystoma punctatum. J. Exp. Zool., 97:21-41. Born, G. 1897 Uber Verwachsungsversuche mit Amphibienlarven. Roux’ Arch. Entw.-mech., 4: 349-465. Brachet, A. 1911 Etudes sur les localisations germinales et leur potentialité réelle dans l’oeuf parthenogénétique de Rana fusca. Arch. de Biol., 26:337-363. Brachet, J. 1940 Etude histochimique des pro- teines au cours du développement embryonnaire des poissons, des amphibiens et des oiseaux. Arch. de Biol., 57:167-202. 1941 La localisation des acides pentose- nucléiques dans les tissus animaux et les oeufs d’Amphibiens en voie de développement. Arch. de Biol., 53:207-257. 1943 Pentosenucleoproteides et induction neurale. Bull. de l’Acad. Roy. de Belgique, 5th Serie, 29:707-718. 1945 Embryologie Chimique. Masson, Paris. (English edition, 1950.) 1949 Le réle et la localization des acides nucléiques au cours du développement embryon- naire; in Acidi Nucleici, Proteine e Differenzia- mento Normale e Pathologico, pp. 1-25. Rosen- berg and Sellier, Torino. 1950 Characteristiques biochimiques de la compétence et de linduction. Rey. suisse Zool., 57:57-75. , and Chantrenne, H. 1942 Nucleopro- teides libres et combinés sous forme de granules chez l’oeuf d’Amphibiens. Acta biol. Belg., 4: 451-454. , and Jeener, T. 1944 Recherches sur les particules cytoplasmiques de dimensions macro- moléculaires riches en acide pentosenucléique. Enzymologia, 77:196-211. , and Rapkine, L. 1939 Oxydation et ré- duction d’explantats dorsaux et ventraux de gas- trulas (Amphibiens). Compt. rend. Soc. de biol., 131:789-791. Brandes, J. 1938 Modification de la morpho- génese primordale chez les Amphibiens, par l’ac- tion précoce des rayons ultraviolets. Acad. roy. de Belgique Sci., 24:92-108. 1940 Action des rayons ultraviolets sur la morphogénése des Amphibiens. II. Arch. de Biol., 57:219-292. 1942 Action des rayons ultraviolets sur la morphogénése des Amphibiens. II. Arch. de Biol., 53:150-206. Bretscher, A. 1949 Die Hinterbeinentwicklung von Xenopus laevis und ihre Beeinflussung durch Colchicin. Rev. suisse Zool., 56:33-96. Brown, M. G., Hamburger, V., and Schmitt, F. O. 1941 Density studies on amphibian embryos with special reference to the mechanism of or- ganizer action. J. Exp. Zool., 88:353-372. Bruns, E. 1931 Experimente iiber das Regula- tionsvermogen der Blastula von Triton taeniatus und Bombinator pachypus. Roux’ Arch. Entw.- mech., 123:682-718. Buchanan, J. W. 1929 The relation between em- 288 EMBRYOGENESIS: PROGRESSIVE DIFFERENTIATION bryo volume and the susceptibility of Ambly- stoma punctatum embryos to potassium cyanide. Physiol. Zool., 2:125-147. Burt, A. 1943 Neurulation in mechanically and chemically inhibited Amblystoma. Biol. Bull., 85: 103-115. Bytinski-Salz, H. 1929 Untersuchungen iiber die Determination und die Induktionsfahigkeit eim- iger Keimbezirke der Anuren. Roux’ Arch. Entw.- mech., 778:121-163. 1931 Untersuchungen iiber die Induk- tionsfahigkeit der hinteren Medullarplattenbe- zirke. Roux’? Arch. Entw.-mech., 723:518-564. 1936 Kombinative Einheitsleistungen in der Entwicklungsgeschichte. Compt. rend. XII. Congrés Intern. de Zool. Lisbonne, 1935, pp. 595— 618. Cagianut, B. 1949 Zur Wirkung von Sexualhor- monen auf die Primitiventwicklung von Triton alpestris. Zeitschr. Zellf., 34:471-501. Carpenter, E. 1937 The head pattern in Ambly- stoma studied by vital staining and transplanta- tion methods. J. Exp. Zool., 75:103-129. Child, C. M. 1915 Individuality in Organisms. University of Chicago Press, Chicago. 1929 Physiological dominance and phys- iological isolation in development and reconstitu- tion. Roux’ Arch. Entw.-mech., 777:21-66. 1941 Patterns and Problems of Develop- ment. University of Chicago Press, Chicago. 1946 Organizers in development and the organizer concept. Physiol. Zool., 79:89-148. Chuang, H. H. 1938 Spezifische Induktionsleis- tungen von Leber und Niere im Explantations- versuch. Biol. Zentrbl., 58:472-480. 1939 Induktionsleistungen von frischen und gekochten Organteilen (Niere, Leber) nach ihrer Verpflanzung in Explantate und verschie- dene Wirtsregionen yon Tritonkeimen. Roux’ Arch. Entw.-mech., 739:556-638. 1940 Weitere Versuche iiber die Veran- derung der Induktionsleistungen von gekochten Organteilen. Roux’ Arch. Entw.-mech., 740:25- 38. 1947 Defekt- und Vitalfarbungsversuche zur Analyse der Entwicklung der kaudalen Rumpfabschnitte und des Schwanzes bei Uro- delen. Roux’ Arch. Entw.-mech., 743:19-125. Claude, A. 1940 Particulate components of nor- mal and tumor cells. Science, 97:77-78. 1941 Particulate components of the cyto- plasm. Cold Spring Harbor Symp. Quant. Biol., 9:263—271. 1943 Distribution of nucleic acids in the cell and the morphological constitution of cyto- plasm. Biol. Symp., 70:111-125. 1944 Distribution of enzymatic activities in fractions of mammalian liver; in A. A. A. S. Research Conference on Cancer. Washington, D. C., pp. 223-226. 1946 Fractionation of mammalian liver cells by differential centrifugation. J. Exp. Med.., 84:51-89. Cohen, A. L. 1942 The organization of proto- plasm: a possible experimental approach. Growth, 6:259-272. Cohen, S. S. 1944 The enzymatic degradation of thymus nucleohistone. J. Biol. Chem., 758:255- 264. Copenhaver, W. M., and Detwiler, S. R. 1944 Developmental behavior of Amblystoma eggs subjected to solutions of indolebutyric acid. Anat. Rec., 79:247-261. Dalcq, A. 1933 La détermination de la vésicule auditive chez le discoglosse. Arch. d’Anat. Micr., 29:389-420. 1940 Contribution a l’étude du potentiel morphogénétique chez les Anoures. I. Arch. de Biol., 57:387-586. 1941a L’Oeuf et son Dynamisme Organ- isateur. A. Michel, Paris. 1941b Contributions a l’étude du poten- tiel morphogénétique chez les Anoures. III. Arch. de Biol., 53:2-124:. 1943 Le phénomeéne d’induction en em- bryologie. Bull. Acad. Roy. Méd. de Belgique, Ser. VI., 8:300-312. 1946 Recent experimental contributions to brain morphogenesis in Amphibians. 6th Growth Symposium, pp. 85-119. 1947 Sur Vinduction de l’épiphyse et sa signification pour la morphogénése du cerveau antérieur. Arch. Portug. Sci. Biol., 9:18-41. , and Lallier, R. 1948a Transplantations de territoires variés de la zone marginale médio- dorsale de la jeune gastrula chez le Triton. Arch. de Biol., 59:268-378. , and Lallier, R. 1948b Neuralisation directe de greffons chordo-mésoblastiques chez le Triton. Compt. rend. de l’Assoc. des Anatomistes Strasbourg, pp. 160-163. , and Pasteels, J. 1937 Une conception nouvelle des bases physiologiques de la morpho- génese. Arch. de Biol., 48:669-710. , and Pasteels, J. 1938 Potentiel morpho- génétique, régulation et “axial gradients” de Child. Bull. Acad. Roy. Méd. de Belgique, Ser. VI, 3:261-308. Damas, H. 1947 Effet de la suspension précoce du flux inducteur sur la détermination du neu- rectoblast medullaire. Arch. de Biol., 58:15-57. Daniel, F. J., and Yarwood, E. A. 1939 The early embryology of Triturus torosus. Univ. Calif. Publ. Zool., 43:321-356. Darlington, C. D. 1944 Heredity, development and infection. Nature, 754:164-169. DeBeer, G. R. 1947 The differentiation of neural crest cells into visceral cartilage and odontoblasts in Amblystoma, and a reexamination of the germ- layer theory. Proc. Roy. Soc. B., 134:377-398. 1951 Embryos and Ancestors. Clarendon Press, Oxford, England. Dollander, A. 1950 Etude des phénoménes de régulation consécutifs 4 la separation des deux premiers blastomeéres de l’oeuf de Triton. Arch. de Biol., 67:1-110. , and Derby, G. 1949 Données complé- mentaires sur la régulation consécutive a la liga- ture frontale de l’oeuf aux stades jeunes chez le Triton (T. helveticus). Ann. Soc. Roy. Zool. Belg., 80:9-19. Driesch, H. 1929 ‘The Science and Philosophy of AMPHIBIANS the Organism. 2d ed. Black and Co., London. Diirken, B. 1936 Uber Bestrahlung des Organ- isatorbezirkes im Tritonkeim mit Ultraviolett. Zeitschr. wiss. Zool., 747:295-356. Du Shane, G. 1943 The embryology of vertebrate pigment cells. Part I. Amphibia. Quart. Rev. Biol., 78:108-127. Ekman, G. 1936 Beobachtungen iiber den Bau durch halbseitige obere Urmundlippe induzierter Embryonen bei Triton. Ann. Acad. Sci. Fenn., Ser. A., 45:1-100. 1937 Zwei bemerkenswerte induzierte Embryonen bei Triton. Acta Soc. pro. Fauna et Flora Fenn., 60:113-128. Endres, H. 1895 Uber Anstich- und Schniirver- suche an Eiern von Triton taeniatus. Schles. Ges. Vaterland. Kultur. 73. Fankhauser, G. 1925 Analyse der physiologi- schen Polyspermie des Triton-Eies auf Grund von Schniirungsexperimenten. Roux’ Arch. Entw.- mech., 705:501-580. 1930 Zytologische Untersuchungen an geschniirten Triton-Eiern. I. Die verzédgerte Kernversorgung nach hantelf6rmiger Eimschnii- rung des Eies. Roux’ Arch. Entw.-mech., 122: 117-139. 1948 The organization of the amphibian egg during fertilization and cleavage. Ann. New York Acad. Sci., 49:684-708. Fautrez, J. 1949 De “chordaliserende” invloed van urea up het ei van Rana temporaria. Medel. Koninkl. Vlaamse Acad. Wetensch., 77:1-30. Fernald, R. L. 1943 The origin and development of the blood island of Hyla regilla. Univ. Calif. Publ. Zool., 57:129-148. Fischer, F. G. 1935 Zur chemischen Kenntnis der Induktionsreize in der Embryonal-Entwick- lung. Verhandl. dtsch. Zool. Ges. (Zool. Anz. Suppl.), pp. 171-176. , Wehmeyer, E., and Jithling, L. 1933 Zur Kenntnis der Induktionsmittel in der Embryonal- entwicklung. Nachr. Ges. Wiss. Gottingen, VI. Biologie, 9:394—400. , Wehmeyer, E., Lehmann, L., Jiihling, L., and Hultzsch, K. 1935 ‘Zur Kenntnis der In- duktionsmittel in der Embryonal-Entwicklung. Ber. dtsch. Chem. Ges., 68:1196-1199. Gallera, J. 1947 Effets de la suspension précoce de Vinduction normale sur la partie préchordale de la plaque neurale chez les Amphibiens. Arch. de Biol., 58:221-264. 1948 Recherches comparées sur le dével- oppement du neurectoblaste préchordal trans- planté sur l’embryon ou enrobé dans l’ectoblaste in vitro (Triton alpestris). Rev. suisse Zool., 55:295-303. Geinitz, B. 1925a Embryonale Transplantation zwischen Urodelen und Anuren. Roux’ Arch. Entw.-mech., 706:357-408. 1925b Zur weiteren Analyse des Organ- isationszentrums. Zeitschr. f. ind. Abstamm.-L., 37:117-119. Gilchrist, F. G. 1928 The effect of a horizontal temperature gradient on the development of the ege of the urodele, Triturus torosus. Physiol. Zool., 1:231-268. 289 Gillette, R. 1944 Cell number and cell size in the ectoderm during neurulation (Amblystoma mac- ulatum). J. Exp. Zool., 96:201-221. Glaser, O. C. 1914 On the mechanism of mor- phological differentiation in the nervous system. Anat. Rec., 8:527-551. 1916 The theory of autonomous folding in embryogenesis. Science, 44:505-509. Goerttler, K. 1925 Die Formbildung der Medul- laranlage bei Urodelen im Rahmen der Verschie- bungsvorgange von Keimbezirken wahrend der Gastrulation und als entwicklungsphysiologisches Problem. Roux’ Arch. Entw.-mech., 106:503-541. 1927 Die Bedeutung gestaltender Bewe- gungsvorgange beim Differenzierungsgeschehen. Roux’ Arch. Entw. Mech., 112:517-576. Graffi, A. 1940 Einige Betrachtungen zur Aeti- ologie der Geschwiilste, speziell zur Natur des wirksamen Agens der zellfrei iibertragbaren Hiihnertumoren. Zeitschr. f. Krebsforsch., 50: 501-551. Gurwitsch, A. 1922 Uber den Begriff des embry- onalen Feldes. Roux’ Arch. Entw.-mech., 57:383- 415. Gustafson, T. 1950 Survey of the morphogenetic action of the lithium ion and the chemical basis of its action. Rev. suisse Zool. (Suppl.), 57:77-91. Haddow, A. 1944 Transformation of cells and viruses. Nature, 754:194-199. Hadorn, E. 1936 Ubertragung von Artmerk- malen durch das entkernte Eiplasma beim mero- gonischen Triton-Bastard, palmatus-Plasma X cristatus-Kern. Verh. Dtsch. Zool. Ges., pp. 97- 104. Hall, E. K. 1937 Regional differences in the action of the organization centre. Roux’ Arch. Entw.-mech., 735:671-688. Hama, T. 1944 On the inductive specificity of fresh and boiled tissues of vertebrate liver and kidney. Annot. Zool. Japon., 22:165-172. 1949 Explantation of the urodelan or- ganizer and the process of morphological differ- entiation attendant upon invagination. Proc. Jap. Acad., 25:No. 9. Hamburger, V. 1942 A Manual of Experimental Embryology. University of Chicago Press, Chi- cago. Harrison, R.G. 1910 The outgrowth of the nerve fiber as a mode of protoplasmic movement. J. Exp. Zool., 9:787-848. 1925 The development of the balancer in Amblystoma, studied by the method of trans- plantation and in relation to the connective tissue problem. J. Exp. Zool., 47:349-427. 1935 Factors concerned in the develop- ment of the ear in Amblystoma punctatum. Anat. Rec., 63:(Suppl. 1) 38-39. 1938 Die Neuralleiste. Anat. Anz., 85: Erg. Heft:3—30. 1945 Relations of symmetry in the de- veloping embryo. Trans. Connecticut Acad. Arts Sci., 36:277-330. 1947 Wound healing and reconstitution of the central nervous system of the amphibian embryo after removal of parts of the neural plate. J. Exp. Zool., 106:27-84. 290 Herlitzka, A. 1897 Sullo sviluppo di embrioni completi da blastomeri isolati di uova di Tritone (Molge cristata). Roux’ Arch. Entw.-mech., 4: 624-658. Hermann, H., Nicholas, J. S., and Boricious, J. K. 1950 Toluidine blue binding by developing muscle tissue. J. Biol. Chem., 784:321-322. Hoadley, L. 1938 The effect of supramaximum temperatures on the development of Rana pipiens. Growth, 2:25-48. Horstadius,S. 1944 Uber die Folgen von Chorda- Exstirpation an spaten Gastrulae und Neurulae von Amblystoma punctatum. Acta Zool., 25:1-13. 1950 The Neural Crest. Oxford Univer- sity Press, Oxford, England. , and Sellman, S. 1945 Experimentelle Untersuchungen iiber die Determination des knorpeligen Kopfskelettes bei Urodelen. Nov. Acta Soc. Scient. Uppsala., Ser. IV, 73:1-170. Holmdahl, D. E. 1939 Die Morphogenese des Vertebratorganismus vom formalen und experi- mentellen Gesichtspunkt. Roux’ Arch. Entw.- mech., 139:191-226. Holtfreter, J. 1931 Uber die Aufzucht isolierter Teile des Amphibienkeimes. II. Roux’ Arch. Entw.-mech., 124:404—465. 1933a Die totale Exogastrulation, eine Selbstabl6sung des Ektoderms yom Entomeso- derm. Roux’ Arch. Entw.-mech., 129:669-793. 1933b Nicht typische Gestaltungsbewe- gungen, sondern Induktionsvorgange bedingen die medullare Entwicklung von Gastrulaek- toderm. Roux’ Arch. Entw.-mech., 127:591-— 618. 1933c Der Einfluss von Wirtsalter und verschiedenen Organbezirken auf die Differen- zierung von angelagertem Gastrulaektoderm. Roux’ Arch. Entw.-mech., 127:620-775. 1933d Organisierungsstufen nach region- aler Kombination von Entomesoderm mit Ekto- derm. Biol. Zentrbl., 53:404—431. 1933e Nachweis der Induktionsfahigkeit abgetoteter Keimteile. Roux’ Arch. Entw.-mech., 127:584-633. 1934a Der Einfluss thermischer, mech- anischer und chemischer Eingriffe auf die Indu- zierfahigkeit von Tritonkeimteilen. Roux’ Arch. Entw.-mech., 732:225-306. 1934b Uber die Verbreitung induzier- ender Substanzen und ihre Leistungen im Triton- Keim. Roux’ Arch. Entw.-mech., 132:307-383. 1934c Formative Reize in der Embryo- nalentwicklung der Amphibien, dargestellt an Explantationsversuchen. Arch. exp. Zellf., 15: 281-301. 1935a Morphologische Beeinflussung von Urodelenektoderm bei xenoplastischer Trans- plantation. Roux’ Arch. Entw.-mech., 133:367- 426. 1935b Uber das Verhalten von Anuren- ektoderm in Urodelenkeimen. Roux’:Arch. Entw.- mech., 133:427-494, 1936 Regionale Induktionen in xeno- plastisch zusammengesetzten Explantaten. Roux’ Arch. Entw.-mech., 134:466-550. 1938a Verdnderungen der Reaktions- EMBRYOGENESIS: PROGRESSIVE DIFFERENTIATION weise im alternden isolierten Gastrulaektoderm. Roux’ Arch. Entw.-mech., 738:163-196. 1938b Differenzierungspotenzen _isolier- ter Teile der Urodelengastrula. Roux’ Arch. Entw.-mech., 738:522-656. 1938c Differenzierungspotenzen _ isolier- ter Teile der Anurengastrula. Roux’ Arch. Entw.- mech., 138:657-738. 1939a Studien zur Ermittlung der Ge- staltungsfaktoren in der Organentwicklung der Amphibien. I. Roux’ Arch. Entw.-mech., 739: 110-190. 1939b Studien zur Ermittlung der Ge- staltungsfaktoren in der Organentwicklung der Amphibien. II. Roux’ Arch. Entw.-mech., 139: 227-273. 1939c Gewebeaffinitat, eim Mittel der embryonalen Formbildung. Arch. f. exp. Zellf., 23:169-209. 1943a Properties and functions of the surface coat in amphibian embryos. J. Exp. Zool., 93:251-323. 1943b 5 02 0% 620 a) ° a) 0° o ae % ° ° Peery: at ° 33.9530 200 0 ES 208 °. 0: o e2%o ° ° Da 2anole o © po ef? Po °°? So) 000 0 00° “00 0 Ses LZ Zan () ( ae 301 longitudinal line and its closure hence en- tirely dissociated from the overgrowth of the yolk by the periphery of the blastoderm. The diagrams of Figure 111A, taken from the sequence in Fundulus, can serve as a model for teleost development only in a very eg 299 0° o 39 uh Pn r Ay ee gas Fig. 111. Diagrams comparing major prospective divisions of the germ in (A) teleost (Fundulus: Oppen- heimer, ’37) and (B) chick (Rudnick, ’48; Spratt, 52) during gastrulation and later. For obvious reasons most of the yolk mass is omitted in the chick figures. White: ectoderm. Radiating lines in B/ and B2: non- medullary ectoderm. Stippled: superficial material which will later be invaginated. Broken lines: invaginated material. Mesh: extraembryonic ectoderm. Circles: yolk. 7, Blastula stage; 2, late gastrula, blastopore nearly closed; 3, embryo formed. position of the material to be invaginated, in the teleost blastoderm, as contrasted with its central-posterior position in the bird; con- versely, the central position of the extra- embryonic material in the fish, and the peripheral location of all extraembryonic material in the bird. The blastopore in the first case is the periphery of the disc; in the second case, it is reduced to a central general way. Fish eggs vary widely in their proportion of protoplasm to yolk, in the relative time necessary for closure of the blastopore, and hence in the relation of axis formation to invagination. The two best-known forms experimentally, Salmo and Fundulus, are widely separated in the series. Figure 112 diagrams comparable stages as regards blastopore closure in these eggs; 302 EMBRYOGENESIS: PROGRESSIVE DIFFERENTIATION the marked precocity of axis formation in the large-yolked Salmo is thus made clear. On the enlarged diagrams of the early gas- trula at the top, prospective areas have been projected in lateral view. As in the case of all vertebrates so far studied, invagina- FUNDULUS AEE Reece LOL] TIT T 8 ——=za_if Fig. 112. Comparative lateral diagrams of pro- spective areas in the early gastrula and diagrams of gastrulation stages. White: prospective nervous system. Stippled: prospective chorda. Oblique lines: future somites. Shaded: extraembryonic area (yolk sac). TB, tailbud material; Ent, entoderm. The boundary between chorda and entoderm in the Fundulus diagram is purely conventional; experi- mentally it has not been specified. The asterisk in the gastrulation series represents the migration of a group of cells on the germ ring into the somite mesoderm. (From Pasteels, ’36a; Oppenheimer, 136bs, 737.) tion of material to lower layers is accom- panied by marked convergence of both in- vaginating and superficial material to the dorsal midline of the future embryo. Hence, prospective areas for nervous system, chorda and other mesoderm extend at first some dis- tance laterally to the position they will later occupy. The convergence movements in the mesoderm are summarily indicated by migration of the asterisk in the lower figures. Contrary to what might be expected a priori, the trout blastodisc, which has so much centrifugal growth and stretching to accomplish, likewise undergoes mutch more marked convergence movements. Meso- dermal trunk material in Fundulus lies originally closer to the embryonic shield than it does in Salmo. In both, ultimately, the ventral lip enters the tailbud. For the avian embryo, we have almost no comparative material for this period; our information derives almost exclusively from the egg of the domestic fowl. Surface dia- grams are presented in Figure 113 for the stages corresponding to those just indicated for teleosts. What emerges from the some- times conflicting studies is that the following order of gastrulation takes place: (1) The hypoblast is formed by a process of sporadic delamination from the posterior region of the pellucid area; this process is not shown in the diagrams of Figure 113. (2) In the same general region, the primitive streak appears soon afterward as an irregular thick- ening, followed by invagination and emigra- tion of mesoderm beneath the superficial layer. The first invaginated mesoderm moves laterally and posteriorly, becoming extra- embryonic. The posterior part of the streak through which it invaginates thus corre- sponds roughly to the ventrolateral blasto- pore lips. (3) Gradually, more anterior ma- terial is incorporated in the streak and anterolateral emigration of mesoderm en- sues; the front end of the streak thus corre- sponds to the dorsal half of the blastopore. The anteriormost axial mesoderm remains in the lower layers of the anterior streak for considerable time: this concentration of material constitutes Hensen’s node. Lateral and extraembryonic mesoderm continues to invaginate through the posterior three- fourths of the streak even after the axial mesoderm, as notochord and somites, is elongating and differentiating anterior to the node. This order of events in the chick is clearly associated with convergence of superficial areas toward the invaginating zone, the most extensive migration being performed by the more peripheral area pellucida re- gions which enter the early (posterior) prim- itive streak. The arrows in Figure 111B-/ are intended to show the approximate extent of convergence. The movements thus differ in order, tempo and magnitude from those outlined for the teleost. TELEOSTS AND BIRDS In both types, it should be noted, con- vergence and concentration of material that is to enter the embryonic axis is in sharp contrast to thinning and rapid _ epibolic spreading of material destined to be extra- embryonic. Luther (37) was able to show an antagonism between embryo formation 303 tion from the rest of the blastoderm. Future ventral or extraembryonic areas react to ex- plantation in vitro by spreading and vesicu- lating (Rudnick, ’38b). In embryonic and extraembryonic areas, in teleost and chick, before and during gastru- lation, the innate migration tendencies of HYPOBLAST DELAMINATED FROM POSTERIOR HALF LATE PRIMITIVE STREAK MESODERM POSTERIO GROWTH EARLY PRIMITIVE STREAK ECTODER MESODERM HEAD —PROCESS Fig. 113. Gastrulation stages in the chick: prospective areas, from marking experiments by Wetzel (29), Pasteels (°36b), Spratt (’46, 47, ’52), Spratt and Condon (’47). In all but the unincubated stage, the left haif shows superficial areas, the right half, invaginated mesoderm. The following prospective areas are indicated: extraembryonic ectoderm, shaded. Embryonic ectoderm, white: the medullary plate is enclosed by a heavy line and its anteroposterior levels demarked in the head-process diagram as P, prosencephalon; M, mesen- cephalon; R, rhombencephalon; Sp, spinal cord; TJ, tail nerve cord. Chorda and prechordal plate, stippled. Mesodermal somites, heavy parallel lines. Heart, cross-hatched. Lateral plate and extraembryonic mesoderm, light horizontal lines. The blastopore or primitive streak is indicated by vertical shading. The boundaries of the posterior and lateral mesodermal areas are conjectural; those of the anterior axis (notochord and somites) as well as of the ectodermal areas have been carefully mapped by Spratt and his co-workers, on explanted blastoderms. In the head-process stage, the region of greatest anteroposterior growth, most pronounced in the median line, is bracketed. (Spratt, 47; Gaertner, ’49.) and epiboly by defect experiments in Salmo. In this same form, Devillers (’48a) has shown that the onset of gastrulation introduces a marked localization of the spreading tend- ency, which becomes restricted to the fu- ture extraembryonic region at that time (cf. Fig. 114). This may indicate the first appearance of a relatively persistent physio- logical differentiation between these two major areas. In the chick, explantation experiments show a similar characteristic tendency of axial materials to condense, even in isola- the epithelia do not interfere with wound healing, or with regulation of excised parts, provided the latter are not disproportion- ately large. The role of the surface coat in this process has been studied for the trout ege by Devillers (48a, ’51b), im Fundulus by Trinkaus (749, 51). No surface coat as such has been demonstrated in the chick blastoderm or yolk, but the behavior of the former after lesions would indicate that a continuous intercellular material must be present and must function similarly after wound healing. 304 EMBRYOGENESIS: PROGRESSIVE DIFFERENTIATION It has been emphasized for the chick that invagination in the primitive streak must involve a de-epithelization process (Wadd- ington and Taylor, ’37) where individual cells lose connection with other cells (detach from the surface coat?) and become rounded and heaped up in the streak. Evidence from a wide variety of histological and_histo- chemical studies is continually accumulating to indicate that this invagination process is accompanied by a whole series of cellular Fig. 114. Healing of trout blastodisc after removal of one-half its substance. Left side, morula stage. The lower figure shows that wound healing and closure take place evenly from all sides. Right, gastrula stage. Only the ventral (top) half stretches to heal the wound (see lower figure). (After Devil- lers, 48a, Fig. 55.) changes: increased affinity for dyes in the upper layers of the streak at least; increased reducing capacity for oxidation-reduction in- dicators (Rulon, °35); loss of lipids upon invagination (Jacobson, ’38); increased ribo- nucleic acid content followed by its loss in laterally migrating mesoderm—not, evi- dently, in the axial mesoderm which mi- grates anteriorly (Gallera and Oprecht, ’48) ; increased sulfhydryl (reported by Brachet; cf. also Bufio, 51); concentration of indo- phenol oxidase and phosphatase (Moog, 743, 44). The anteroposterior gradient of activity observed so frequently in these studies on the streak is of course morphologically a dorsoventral gradient of the blastopore and may be at least in part a function of differential massing and tempo of invagina- tion of cells in the different parts of the streak. In the teleost, invagination is primarily an ingression of a continuous sheet of cells; Devillers (51b) has figured in Salmo the process of loss of connection with the surface coat, as in the amphibian gastrula, and has also described loss of basophilia during the process. According to Oppenheimer (’36b), some individual detachment of cells from anterior portions of the embryonic shield epiblast, to join the mesoderm, occurs in addition to ingression at the dorsal lip. It is difficult to avoid comparing these de- epithelizations to the more radical experi- mental situation demonstrated by Holtfreter (47) in explants of amphibian ectoderm, where neural differentiation appears to be associated with the margin of the epithelial mass, under conditions promoting slight cytolysis and eventual sloughing of cells. The experimental analysis of the am- phibian egg has led to the identification of the region actively concerned in induction of central nervous system, with the in- vaginating chordamesoderm. It would be expected from the distribution of the latter areas that Salmo and Fundulus (cf. Fig. 112) might also differ in spatial extent of material capable of performing the primary induction. This property has not been com- pared directly in the two forms. It is known, for both, that a bit of dorsal lip or already invaginated material will induce a super- numerary axis from material that would not normally form dorsal axis. A phenomenon that would seem to bear some relation to such direct tests of induction is the differ- entiation of axial tissues in isolated pieces or grafts taken from regions which normally do not contribute to the dorsal axis. In Salmo in the blastula stage, according to Luther (?36a), all quadrants of the blastoderm are equipotent for the formation of axial structures when grafted to the yolk sac of hatched fry. It is hard to imagine that the neural tubes in these grafts arose by inde- pendent differentiation of ventral ectoderm: we are forced to interpret such formations as “self-organizations” involving release, among other things, of some latent neural-plate— inducing potency somewhere in the isolate. The ability of the ventral side of the trout blastoderm to form an axis when released from the influence of the dorsal side was confirmed (Luther, ’37) by combining two ventral half-blastulae: an axis regularly formed, usually from the half that was slight- ly younger. Devillers (*51a) in extending this work has made the striking discovery that if the preparation is made using as a host TELEOSTS AND BIRDS a ventral half-blastula deeply stained with neutral red, the resulting embryo arises from the ventral pole, i.e., in reversed orien- tation. However, if a stained ventral trans- plant replaces the host ventral half, an unstained axis arises from the dorsal pole in the original direction. Thus staining seems SALMO FUNDULUS tw wo 305 dulus, according to Oppenheimer, no grafts of lateral or ventral germ ring in gastrula stages have been found to form axial tissues unless implanted directly into the embryonic shield, where inducing power is already known to exist. Thus in both forms, com- petence to form nervous system exists all BIRD C ve Vv Fig. 115. Comparison of distribution of inducing potency (vertical lines) and neural competence (hori- zontal lines) in Salmo, Fundulus and the bird, during gastrulation. Approximately comparable stages are shown for each form: blastula, early gastrula, late gastrula. In Salmo (Luther, ’35, ’36a,b, ’37) all sectors of the germ ring can form axial structures in grafts, at least part-way through gastrulation: this is interpreted as showing even distribution of inducing potency in early stages, normally expressed only at the dorsal pole. This potency is gradually lost in ventral and lateral sectors as gastrulation proceeds. In Fundulus, no data for the blastula are available. No data show inducing power outside the embryonic shield in gastrulation stages (Oppenheimer, ’34, 35, ’36a, ’38). In both forms, the capacity to respond to induction is apparently not restricted to any locality during gastrulation. In the bird, all sectors of the blastoderm at the blastula stage (Lutz, 48) can form axes; hence both potencies must be evenly distributed. Data from the chick (Waddington, ’32, etc.; Woodside, ’37) show that the primitive streak or its precursors can induce axes throughout gastrulation, and that all ectoderm (includ- ing extraembryonic) can give the neural response until fairly late in gastrulation, when this response becomes gradually restricted to the prospective medullary region as suggested by the bottom figure of the series. only to reverse polarity and convey domi- nance in an undisturbed ventral half-blas- tula. Devillers emphasizes the relation be- tween periblast and blastoderm and tends to look on the former as a repository, or chief effector, of the hypothetical ‘“organiz- ing substances” postulated by Tung (see p. 300). As gastrulation begins, the latent capacity for “self-organization” drops sharply in ven- tral sectors, in the trout, and disappears entirely before mid-gastrulation. In Fun- around the blastopore, as well as centrally, during the blastula and at least in the early gastrula stages; while inducing power, overt or latent, becomes gradually limited to the dorsal lip in Salmo; in Fundulus its exist- ence has apparently never been demonstrated outside the embryonic shield sector. Luther observes that the loss of potency to differentiate axial structures from ventral grafts can by no means be explained by migration of a localized cellular material from the ventral lip region. Convergence 306 EMBRYOGENESIS: PROGRESSIVE DIFFERENTIATION movements are not rapid enough to account for such results; the most striking potency change occurs at the onset of gastrulation, before convergence becomes really noticeable (cf. Fig. 115). This is the latest stage when an axis may be induced in the ventral lip by transplantation of a piece of dorsal lip; it is, also, the stage when defects in the dorsal lip region begin to be incompletely regulated (Devillers, *48b). In the avian egg, organization of the blas- tula stage has been most successfully studied by means of sectioning the blastoderm in situ, thus isolating various portions mechani- cally. In the duck egg, which is relatively younger at the time of laying than the chick, and may be classified as a blastula, Lutz (48) has accomplished the startling feat vf obtaining four embryos from one blastoderm. This result could be obtained by making two perpendicular or three paral- lel cuts in the blastoderm. It would seem that all peripheral areas of the pellucid area (at least) were equally able to regulate a whole when isolated; any fixed pre-locali- zation of an inducing center during cleavage would thus be excluded, for this form. It must be recalled that this sectioning of the avian area pellucida is not the morphologi- cal equivalent of Luther’s experiments on the trout: the relation to the future blasto- pore is quite different in the two (cf. Fig. GD. For the chick, the transplantation experi- ments of Butler (735) and the injury experi- ments of Twiesselmann (’38) on early gastrula stages indicate that at this time axial organ- izer material becomes localized in the poste- rior sector of the pellucid area. At the same period, Waddington (’33) has demonstrated that an axis may be induced in the anterior epiblast, which normally would become ex- traembryonic, by contact with the posterior hypoblast. As the primitive streak elongates, it ap- pears to contain the material directly capa- ble of performing inductions of neural axis; it must be noted that regions outside the streak seem not to have been systematically tested for this action, so that there is really no experimental justification for stating that inducing action is restricted, in the living blastoderm, to the primitive streak. Wood- side (°37) has shown that the capacity to re- spond to induction—i.e., neural competence —at first evenly distributed throughout the whole disc, falls off with progressing gastrula- tion, becoming by degrees restricted to the area which normally forms medullary plate. Figure 115 is intended to summarize the substance of the discussion in the preceding paragraphs. It has been emphasized that it has not been possible to apply strictly com- parable tests in all cases. The general trend, however, is alike in all forms: namely, that both induction potency and neural com- petence are at the start of gastrulation widely distributed; that the induction center early becomes restricted to the invaginating area; and that considerably later neural compe- tence likewise becomes restricted to the medullary plate region. By the neurula stage, responsiveness to the axis-inducing stimulus evidently is at an end; the ability to induce lingers on in axial tissues and may be un- masked in many adult tissues by coagulation. PRIMITIVE ORGANIZATION OF THE EMBRYO THE ECTODERM Events of the gastrulation period are most strikingly viewed if attention is focused, as has been done in the preceding section, on the primary embryonic areas (the invaginat- ing and the non-invaginating regions), and on the primary induction of the dorsal axis. That this is an oversimplification is obvious. Each of these primitive areas is to play a role as germ layer first, segregating sub- sequently along anteroposterior and medio- lateral axes into definite regions which are traceable directly to the organs of the adult body. Differentials in the future germ layers may be recognizable before, and especially during, the invagination process. In the late gastrula and neurula stages, transplantation experiments in both teleost and chick agree in showing a more or less fixed anteroposterior arrangement of medul- lary ectoderm. Figure 116 has been pre- pared to summarize the results of numerous investigations of this nature; the axiation is clearly indicated in all cases that have been reported. In many of these cases, how- ever, the transplanted piece contained both archenteron roof and overlying ectoderm; the subsequent differentiation into appro- priate levels of the central nervous system might thus be attributed to specificity of either layer, or of both. The anterior tip of the neural axis, in particular, presents special problems. It would be expected that the anteriormost invaginated material (prechordal plate) would be the normal inductor of forebrain tip and optic vesicles. The elegant experi- ment of Luther (’36b) on Salmo, in which TELEOSTS AND BIRDS ectoderm that would normally have re- mained extraembryonic was placed in the path of otherwise undisturbed invaginating archenteron roof, resulted in induction of forebrains that were normal in many cases. Some abnormal inductions resulted, and some undersized forebrains as well. For Fun- dulus, as Figure 116A shows, forebrain and eye differentiate from transplants of anterior embryonic shield pieces even if mesoderm is not included, although better if the under- lying material is present. It should be men- tioned that in Fundulus the anterior end of the notochord lies at the hindbrain level, so that the prechordal inductor must be relatively extensive. The above-mentioned grafts, however, developed in the host em- bryonic shield, and were thus exposed to undefined host induction fields. In the ex- periments cited in the preceding section, where an axis was induced by transplanted dorsal lip, forebrain defects were reported to be of almost universal occurrence in the induced axis. Under the hypothesis that the anteriormost cells invaginated are capable of inducing forebrain, specifically, these de- fects would be attributed to the absence of a definite region of the transplanted in- ductor. Less compatible with this hypothesis is the failure (Fig. 116B) of anterior archen- teron roof to induce eye or forebrain under conditions where posterior levels of the same structure perform appropriate inductions. It will be recalled that a highly effective method of producing cyclopia in Fundulus involves exposure to deleterious agents in early cleavage stages (Stockard, ’07). These considerations make it necessary to believe, at the very least, that if the localized in- ductor hypothesis is true, there must also be special additional factors in the anterior tip of the axis, much more easily disturbed by environmental changes than are the rela- tions in the rest of the dorsal axis. In the chick, peculiarities of the anterior end of the neural axis appear in the period of the formation of the primitive streak (Fig. 116D). At this time, isolated anterior pieces of epiblast form unspecific medullary plate or tube when grown in vitro, even before mesoderm has begun to underlie the upper layer in that region. In such preparations, mesenchyme is invariably present; one pos- sible interpretation is that a sporadic in- vagination from the upper layer may simu- late mesoderm formation together with its induction effect. In the late gastrula stages (Fig. 116), as the anterior ectoderm be- comes fully underlaid, forebrain and optic 307 vesicles tend to form recognizably in vitro even when isolated from posterior levels; this tendency becomes progressively fixed as the medullary plate becomes visible. During this period, potency to form retinal tissues in grafts is evenly distributed along the medio- ar NO BRAIN HINDBRAIN BRAIN FOREBRAIN MIDBRAIN [+ corp A B 2 Jax CORG EYE CNS. EYE MIDBRAIN C.N.S. C D FOREBRAIN EYE ====—> FOREBRAIN i> MIDBRAIN 5 HINDBRAIN > CORD Fig. 116. Axial differentiation of central nervous system. A, Results of transplants or excisions from the axial keel of Fundulus (after Oppenheimer, °36b, Fig. 21). B, Results of inductions by three levels of the archenteron roof of Salmo, associated with extraembryonic ectoderm (after Eakin, ’39, Fig. 1). C, Results of transplantation of parts of the unincubated chick blastoderm to the chorioallantois (Butler, ’35). D, Explantation of pieces of the chick blastoderm: pre-groove primitive streak stage (Spratt, 42). EH, Localization of central nervous system levels in the definitive primitive streak stage of the chick (data from Hunt, ’32; Clarke, ’36); chorioallantoic transplantation. F, Same for the head-process stage (from Clarke, ’36; Rawles, ’36). lateral axis of the eye-forming area; the localization of histogenetic potency lags be- hind morphogenetic localization. Ultimately the ability to differentiate retina is localized in the optic vesicle areas, shortly before they themselves become morphologically patent. In the definitive primitive streak and head-process stages, transplantation experi- ments reveal anteroposterior axiation of the future medullary plate (Fig. 116#, F), with 308 EMBRYOGENESIS: PROGRESSIVE DIFFERENTIATION specific morphogenetic and histogenetic po- tencies corresponding to the successive levels of the central nervous system. These ex- periments have all involved transplanting mesoderm along with ectoderm, and so give no information on the relationship of these layers. No experiments have shown so clearly as those of Clarke (736), cited for the retina, the exact limits and internal differences of these areas; but it seems likely that they also partake of the field character so well illus- trated there. The posteriormost end of the neural axis seems, like the foremost, to depend for dif- ferentiation on special conditions not easily reproduced in transplantation or explantation experiments. These conditions evidently dif- fer in various teleosts: in Fundulus, ex- planted blastodiscs do not form tail at all; in Carassius, tail-like bodies may be formed from isolated or partly isolated blastodiscs, as well as from egg fragments where the injury is not defined. In Epiplatys, Oppen- heimer (738) found that isolated ventral halves will undergo a tail-like morphogene- sis, but without differentiating any true neu- ral or mesodermal axis, until closure of the blastopore. Removed after this event, an isolate will form a_ histologically normal tail. This suggests of course that the dorsal lip material is the differential factor. In the chick, the situation bears certain resemblances. The posterior levels of the medullary region in primitive streak and head-process stages show considerable un- willingness to differentiate neural tissue un- der various experimental conditions (Wadd- ington, °35; Rudnick, °38a), but may occasionally do so. This differentiation, evi- dently difficult to realize in primitive streak stages, becomes more regular and normal as corresponding parts of older blastoderms are isolated, after the midline axis is brought into place and involution of mesoderm more nearly completed. Spratt’s (52) most recent study on the localization of neural plate gives a clear explanation of these differences on the basis of morphogenetic movements. Small neural masses are regularly formed from appropriate regions of the sinus rhom- boidalis (Seevers, ’32; Rudnick, ’45). Regu- lation may occur after removal of tail-form- ing material from the sinus; in the tail-bud stage some of the blastema must be left if regulation is to occur (Zwilling, ’42). The peculiar mechanics of neural tube formation in the tail region has been pointed out many times; the induction pattern prevailing there remains unanalyzed. In addition to the primary neural induc- tion, it is usually believed that the pattern of embryonic organs that soon emerges from the newly-formed germ layers is also a re- sult of induction: that is to say, of contact- or proximity-effect of one layer on another, or of one region on adjacent ones of the same germ layer. No clear formulation of these beliefs has yet been made, or seems possible at this stage of our knowledge. Individual cases, such as the classic eyecup- lens relation, seem in the chick to follow the pattern found in some Amphibia, where the lens may be induced from ectoderm of other body regions by a structure (the future iris border) which itself retains some capacity for lens differentiation (Alexander, 37). McKeehan (51) has made a study of the fine scale orientation and even fusion of cells during this induction, which may have much wider application than to this one case. The chick otocyst (Waddington, °37; Levi-Montalcini, ’46) evidently depends for its origin on several regional factors, rather than on any one tissue. In the fish, a wide variety of embryonic regions may be induced to form ear (Oppenheimer, 738) when transplanted to the embryonic shield— not necessarily in a region corresponding to the host ear. In Salmo, Eakin (’39) found a considerable extent of the archenteron roof capable of inducing (directly or indirectly) otic vesicles in extraembryonic ectoderm. Luther (36a) found ears in grafts from various sectors of the gastrula, approxi- mately corresponding to the distribution of nervous tissue, although less frequent in occurrence. In the chick, the prospective nasal placode appears to depend for differentiation upon both brain and mesoderm, in early stages (Street, °37). The origin and dependence of other extramedullary contributions to the nervous system—various placodes, neural crest—have not been studied. The epidermis, too, has been so little investigated that it is perhaps well to state here that it is not a negative tissue, the product of differentia- tion by default of any ectoderm not receiving another induction stimulus. The epithelium found in hyperblastulae of teleost eggs, or in chick blastoderms in which the axis has been suppressed, is not epidermis; it more resembles the epithelium of extraembryonic membranes. Special relations to dermatome and other mesoderm, and to neural crest, must be factors in producing the differenti- ated tissue and its derivatives. The studies of Saunders (’48) on the remarkable pe- TELEOSTS AND BIRDS culiarities of the ectoderm at the tips of the limb buds in the chick, point to how much there remains to learn about the prop- erties of non-neural ectoderm. THE LOWER LAYERS As gastrulation progresses, not only do in- duction capacity, neural competence, and the specific competences for different levels in the ectoderm become gradually restricted to the appropriate embryonic area: the invaginating layers themselves ultimately must become a mosaic of areas of definite differentiation tendencies. It is believed that all these areas have at first the properties of fields, of indefinite boundaries, possibly overlapping one another. Toward the time of beginning morphogenesis, the boundaries of the fields evidently become more restricted and more precise, as if brought into focus by some microscope capable of reading the future; each field then coincides with the corresponding organ-forming area. In the forms under discussion, our experi- mental knowledge of this course of events in the lower layers is based only on the results of transplantation experiments in which at least two layers were involved, and which resulted in differentiation of complexes of several tissues. The desirable isolation ex- periment has not yet been technically pos- sible. The succession of figures which form the matter of the present section is an ab- straction, presented in order to help the reader to visualize, for each major embryonic structure, the spatial distribution of its origin in grafts, delimited as well as the experimental material allows. Many of the studies on which these abstractions are based have not been carried through to the stage where the potency field corresponds with the actual organ-forming area; but the tendency in that direction is clearly observ- able. Figure 117 deals with the axial mesoderm, in so far as notochord and somites (or skele- tal muscle) are found to occur differentially in grafts of parts of blastoderms. The re- ported data are unsatisfactory for the blas- tula stage; in particular the distributions indicated for the chick (Fig. 117D,J) in the unincubated stage are based on a few positive cases and may not be valid. The trout, however, shows an understandable sequence, with a notochord field converging toward the dorsal lip (Fig. 117B,C), fol- lowed by later invaginated somite material (Fig. 117H,1). The strict delimitation of 309 the chorda field in the chick gastrula (Fig. 117E, F) is in striking contrast, and would indicate that this structure, at least, is local- ized relatively early in the latter form—a CHORDA SALMO 86% i} MYOTOMES CHIC il," Fig. 117. Progressive changes in fields for pro- duction of chorda and somites in grafts from Salmo (Luther, ’36a) and the chick (Butler, ’35; Hunt, *31, °32; Rudnick, ’32; Rawles, ’36). In the diagrams for Salmo, the left half of the figure shows the actual percentages of differentiation in various sec- tors, on which the scheme of the right half is based. No such quantitative estimate has been attempted in the case of the chick. Localization in blastula stages is based on very incomplete information, which may be of no significance. The marked re- striction of chorda-forming material in the chick (E,F) is based on reliable data. The extent of the somite field in the chick (J,K,Z) is based merely on differentiation of skeletal muscle, and so is not critical, nor are the lateral boundaries established. condition not reached even in the late gas- trula in Salmo. The interpretation of this localization in the chick is not easy, as examination of Fig- ure 118 will lead us to reflect. Here we have similar figures showing progressive lo- calization of two more lateral mesodermal 310 EMBRYOGENESIS: PROGRESSIVE DIFFERENTIATION HEART NEPHROS weve ceseosnmcear MM Fig. 118. Progressive localization in the chick of heart muscle and nephros (Butler, ’35; Rudnick, 32, ’38b,c; Rawles, ’43). The situation in the un- incubated blastoderm is somewhat uncertain; it is possible that some localization exists even then. It is established that as soon as the primitive streak appears (B,C), a limited area only can form cardiac muscle in vitro or in grafts (D,E), and that until the head-process stage the median part of this area is most active. In E and G the boundaries of the areas have been carefully determined; in the other stages, they are somewhat conjectural, the shaded areas representing positive cases. structures: heart and nephros. Heart-forming areas (i.e., those which will form cardiac muscle) follow fairly well the course that would be expected of invaginating lateral mesoderm, as probably does the more pos- terior nephric field. A comparison of the various diagrams for the definitive primitive streak stage (Figs. 117E, 118D,F) shows a remarkable concentration of potencies in the immediate vicinity of Hensen’s node: an expression of the experimental fact that grafts from this center develop into larger, more varied masses than do more lateral or anterior or posterior regions; the medianmost region gives the highest proportion of dif- ferentiation of each tissue. It is possible that the thick mesodermal streak in this area contains definitely localized autonomous pri- mordia of notochord and somites, as well as still migrating heart and nephros fields; the results could equally well be interpreted on the basis of high concentration of growth factors plus a few key inducing substances, quite irrespective of morphological localiza- tions. Entodermal derivatives must largely be considered in association with mesodermal ones. Grafts from trout blastodiscs show a distribution not very different from that of somites, for example, for gut tubules (Fig. 119A-C). In the chick, the progressive pat- GUT SALMO A CHICK D LIVER THYROID Fig. 119. Localization of entodermal potencies: A-C from Luther (’36a); D-F from Butler (35), Rudnick and Rawles (’37); G—J from Rudnick (’32) and Rawles (’36); data all from grafts. In the chick material, lateral and posterior limits of areas have not been determined in stages earlier than the head process. In D-F, the shading indicates gut not formed into recognizable organ units; crosses, or- ganized small intestine; circles, identifiable seg- ments of colon. TELEOSTS AND BIRDS terns, if we compare definitive streak stages for entoderm and heart (Fig. 119E, G, /; Fig. 118D) with the corresponding head-process diagrams (Fig. 1197, H, J; Fig. 118£), are un- derstandable on the basis of migration toward definitive positions: thyroid anteriorly, liver laterally, intestine posteriorly. The lateral movement of median material and a sharp anterior stretching of material just in front of the node correspond with known germinal movements. In these cases, at least, the hy- pothesis of migration of definitely localized cell groups in mesoderm and entoderm, to- gether, is tenable and preferable. Of recent years there has been a tendency to focus attention on the heart area not only as an early autonomous region, present at least as early as localized dorsal axial ma- terial, but also as possibly playing an or- ganizing role in the anteroventral region of the body analogous to that of the dorsal lip for the dorsal axis. Direct evidence for this view is lacking; the considerations are early autonomy of the heart, and its tend- ency to be associated with foregut struc- tures (pharynx, thyroid, liver) in experi- mental situations. In the case of the intestine (Fig. 119£, F), no cell migrations have been described that will account for the posterior displacement of the area during the head-process stage. It is only fair to say that they have not been looked for, and that the mesodermal portion of the streak would be the place to look. There is evidence that in gut formation, mesoderm plays a large role (Hunt, ’37) and tnat hypoblast is, to say the least, dispen- sable in the process. It would be very con- venient if such slight but definite posterior displacement of mesoderm in the streak could be found, since it would then explain very well the posterior expansion of the meso- nephric area between the definitive primitive streak and head-process stages, indicated in Figure 118F, G; this material might be con- sidered to undergo a spurt of posterolateral migration at this time, and to push the lateral plate ahead of it, posterior and lateral to the node field. CONCLUDING REMARKS The meroblastic vertebrate eggs, by their pattern of differentiation within the ovary, have realized an almost complete separation of the factors and materials involved in pri- mary differentiation, from those necessary for later growth. Thus most teleost blasto- discs can very early be separated from the Slit yolk and complete most of the histogenetic and embryogenetic sequence in a_ saline medium. The trout and chick blastodiscs differ in requiring at least some carbohy- drate from outside (Spratt, ’49; Devillers, °49) in order to carry through a comparable sequence of events. It is doubtful if these isolates grow to any appreciable degree, if a rigorous definition of growth as increase in mass be insisted on; by contrast, differen- tiation is remarkably complete. Whether this primary semi-independent phase is also rep- resented in some form in holoblastic eggs, and whether a key to developmental kinetics may be seen in this feature and its variants, is not certain. Within the protoplasmic blastodisc of both fish and bird, it seems safe to say that any differentials present during cleavage are fairly easily reversed or negated. That dif- ferentials exist in the fish egg would be indicated by some varieties of experiment where injury or separation results in a de- fective embryo; to balance these, there are clearcut cases of totipotence of halves or even quarters of the blastodisc, continuing up to the blastula stage. Parts of the whole can, when suitably isolated, become wholes themselves. Whether every part of the germ shares this property is still debatable. At some time, during or after the onset of gastrulation, some fairly stable differ- entials appear. Thus in Fundulus the dorsal lip region, as soon as identifiable, has po- tencies sharply different from those of the rest of the circumference of the blastodisc; this situation appears to be achieved only eradually in Salmo. Similarly, the duck egg at laying is evidently largely equipotent, whereas all our information on the unin- cubated chick egg points to the posterior or blastoporal sector as having properties not shared by other regions. Thus within each group are found variations of timing. Between the two groups, bird and fish, is the major and striking difference of spatial pattern: the position of the blastopore, the relative rate of its formation, the continuity of invagination, and complexities of move- ment in the blastoporal area itself. We can contrast the elaborate history of the primitive streak, with its posteroanterior order of ap- pearance, its delayed invagination and emi- eration, and its complex incompletely under- stood pattern of cellular movement, with what seems to be a much more simple and direct invagination pattern in the teleost. It must be said, however, that great technical difficulties have impeded study of details of SIP EMBRYOGENESIS: PROGRESSIVE DIFFERENTIATION invagination in the fish, and that the notion that the process is simple may merely indi- cate our ignorance. Gastrulation, then, shifts germinal regions haying a tentative or reversible pattern into definitive positions related to the blastopore. The axial ectoderm, whatever its primary pattern, is undoubtedly given its final stamp of independent differentiation tendency by the mesoderm that comes to underlie it. As for the invaginated material, there is no evidence in the fish, and but one case in the chick (heart), to show that any specific differentiation tendency is present before the material undergoes invagination. At this stage of our knowledge, then, most of the embryonic fields appear to originate at gas- trulation. Whatever future investigations will show, it is clear that gastrulation is a critical process, separating and reinforcing if not originating the embryonic pattern of potential areas. We know too little of the details of any two of the experimental forms under dis- cussion to compare accurately the course of events whereby any one field moves into definitive position and becomes restricted to the territory of its embryonic anlage. In the teleost neurula we seem to have a fairly well localized system; defects made in the axis may not even be regulated. In the chick, reg- ulation of some parts remains possible for considerable time. Are these differences real, or an accident of the difference in absolute size and scale of organization between the two embryos, whereby it is difficult to make a really minute defect in the teleost, without much subsidiary damage? Again, in gas- trulation stages, it is clear that Fundulus behaves much as the amphibian embryo does, in that transplants to the embryonic shield are induced to form a variety of struc- tures, somewhat but not exactly correspond- ing to the level of the host axis: a result pre- sumably of the wide imprecise nature of the host induction fields. In the chick, no con- vincing experiments show such histological effect of host on transplant, though polarity may be decidedly affected (Waddington and Schmidt, *33). This peculiarity may be only because the usual transplant tested is a piece of primitive streak, the normal fate of which is not known precisely, or because the chick blastoderm reacts in a special way to transplanted tissue: either by joining with it inextricably in a complex induction, or by walling it off, physiologically speaking, so that it develops as a separate intrusive graft. It is obvious that in every way the picture we construct of the progressive invisible changes in any embryo, leading to visible differentiation, depends on the experiments possible with that embryo. Slight differences in texture or in chemical surface, perhaps, may determine whether it is possible to make transplants or explants of any given portion; whether defects will heal or an epithelium respond to induction. These factors are only a small part of the equipment of embryonic cells, but they loom very large experiment- ally. If we are to be able to make valid comparisons, our task must be to disentangle these properties from the intrinsic genetic mechanisms, and to discover the role of both in cellular differentiation. This is for the future. REFERENCES The following list contains only titles actually cited in the text. Recent comprehensive reviews of the material covered are to be found in Oppen- heimer (747), Rudnick (744, 48), and Waddington (G52): Agassiz, L., and Whitman, C. O. 1884 On the development of some pelagic fish eggs. Proc. Am. Acad. Arts & Sci., 20:23-75. Alexander, L. E. 1937 An experimental study of the role of optic cup and overlying ectoderm in lens formation in the chick embryo. J. Exp. Zool., 75:41-68. Bartelmez, G. W. 1912 The bilaterality of the pigeon’s eggs. J. Morph., 23:269-329. Brachet, Jean 1944 Embryologie Chimique. Masson, Paris. Buno, W. 1951 Localization of sulfhydryl groups in the chick embryo. Anat. Rec., 177:123-128. Butler, Elizabeth 1935 The developmental capa- city of regions of the unincubated chick blasto- derm as tested in chorio-allantoic grafts. J. Exp. Zool., 70:357-389. Clapp, Cornelia 1891 Some points in the develop- ment of the toad-fish (Batrachus tau). J. Morph., 5:494-501. Clarke, L. F. 1936 Regional differences in eye- forming capacity of the early chick blastoderm as studied in chorio-allantoic grafts. Physiol. Zool., 9:102-128. Devillers, Ch. 1947 Explantations in vitro de blastodermes de Poissons (Salmo, Esox). Experi- entia, 3:71-74. 1948a Le cortex de l’oeuf de Truite. Ann. Stat. Cent. d’Hydrobiologie appl., 2:29-49. 1948b Suppression du matérial chordal dans la gastrula de Truite. Compt. rend. Acad. sci., 227:1411-1413. 1949 Explantations en milieu synthé- tique de blastodermes de Truite (Salmo irideus). J. Cyto-embryol. Belgo-Neérland., pp. 67-73. 1951a Symétrisation et régulation du germe chez la Truite. Compt. rend. Ass. Anat. Nancy, XXXVIII Réun., pp. 1-7. 1951b Les mouvements superficiels dans TELEOSTS AND BIRDS la gastrulation des poissons. Arch. d’Anat. Micro- scop., 40:298-309. Eakin, R. M. 1939 Regional determination in the development of the trout. Roux’ Arch. Entw.- mech., 139:274-281. Gaertner, R. A. 1949 Development of the poste- rior trunk and tail of the chick embryo. J. Exp. Zool., 111:157-174. Gallera, J., and Oprecht, E. 1948 Sur la distribu- tion des substances basophiles cytoplasmiques dans le blastoderme de la Poule. Rev. suisse Zool., 55:243-250. Hoadley, L. 1928 On the localization of develop- mental potencies in the embryo of Fundulus heteroclitus. J. Exp. Zool., 52:7-44. Holtfreter, J. 1947 Neural induction in explants which have passed through a sublethal cytolysis. J. Exp. Zool., 106:197-222. Hunt, T. E. 1931 An experimental study of the independent differentiation of the isolated Hen- sen’s node and its relation to the formation of axial and non-axial parts in the chick embryo. J. Exp. Zool., 59:395-427. 1932 Potencies of transverse levels of the chick blastoderm in the definitive-streak stage. Anat. Rec., 55:41-70. 1937 The development of gut and its de- rivatives from the mesectoderm and mesentoderm of early chick blastoderms. Anat. Rec., 68:349- 370. Jacobson, W. 1938 The early development of the avian embryo. II. J. Morph., 62:445-488. Levi-Montalcini, R. 1946 Ricerche sperimentali sulla determinazione del placode otico nell’em- brione di pollo. Accad. Naz. Lincei Rendic. Cl. Sci. fis-mat. e nat., ser. VIII, 7:445-448. Lewis, W. H. 1912 Experiments on localization in the eggs of a teleost fish (Fundulus heterocli- tus). Anat. Rec., 6:1-6. 1949 Gel layers of cells and eggs and their role in early development. Roscoe B. Jack- son Mem. Lab. Lect., pp. 59-77. , and Roosen-Runge, E. C. 1942 The for- mation of the blastodisc in the egg of the zebra fish, Brachydanio rerio, illustrated with motion pictures. Anat. Rec., 84:463. Luther, W. 1935 Entwicklungsphysiologische Untersuchungen am Forellenkeim: die Rolle des Organisationszentrums bei der Entstehung der Embryonalanlage. Biol. Zbl., 55:114-137. 1936a Potenzpriifungen an _isolierten Teilstiicken der Forellenkeimscheibe. Roux’ Arch. Entw.-mech., 735:359-383. 1936b Austausch von prasumptiver Epi- dermis und Medullarplatte beim Forellenkeim. Roux’ Arch. Entw.-mech., 735:384-388. 1937 Transplantations- und Defektver- suche am Organisationszentrum der Forellen- keimscheibe. Roux’ Arch. Entw.-mech., 137:404— 424: Lutz, H. 1948 Sur l’obtention expérimentale de la polyembryonie chez le Canard. Compt. rend. Soc. Biol., 142:384-385. McKeehan, M. S. 1951 Cytological aspects of embryonic lens induction in the chick. J. Exp. Zool., 117:31-64. 315 Moog, Florence 1943 Cytochrome oxidase in early chick embryos. J. Cell. Comp. Physiol., 22:223- Silk 1944 Localizations of alkaline and acid phosphatases in the early embryogenesis of the chick. Biol. Bull., 86:51-80. Morgan, T. H. 1893 Experimental studies on teleost eggs. Anat. Anz., §:803-814. Nicholas, J. S., and Oppenheimer, J. M. 1942 Regulation and reconstitution in Fundulus. J. Exp. Zool., 90:127-157. Oppenheimer, J. M. 1934 Experiments on early developing stages of Fundulus. Proc. Nat. Acad. Sci., 20:536-538. 1935 Processes of localization in develop- ing Fundulus. Proc. Nat. Acad. Sci., 27:551-553. 1936a Transplantation experiments on developing teleosts (Fundulus and Perca). J. Exp. Zool., 72:409-437. 1936b Processes of localization in devel- oping Fundulus. J. Exp. Zool., 73:405-444. 1936c The development of isolated blas- toderms of Fundulus heteroclitus. J. Exp. Zool., 72:247-269. 1937 The normal stages of Fundulus heteroclitus. Anat. Rec., 68:1-8. 1938 Potencies for differentiation in the teleostean germ ring. J. Exp. Zool. 79:185-212. 1947 Organization of the teleost blasto- derm. Quart. Rev. Biol., 22:105-118. Pasteels, J. 1936a Etudes sur la gastrulation des vertébrés méroblastiques. I. Téléostéens. Arch. Biol., 47:205-308. 1936b Analyse des mouvements morpho- génétiques de gastrulation chez les oiseaux. Bull. Acad. Roy. Belg., series V, 22:737-752. Rauber, A. 1883 Neue Grundlegungen zur Kenntnis der Zelle. Morph. Jb., 8:233-338. Rawles, M. E. 1936 A study in the localization of organ-forming areas in the chick blastoderm of the head-process stage. J. Exp. Zool., 72:271- Si: 1943 The heart-forming areas of the chick blastoderm. Physiol. Zool., 76:22—41. Roosen-Runge, E. C. 1938 On the early develop- ment—bipolar differentiation and cleavage—of the zebra fish, Brachydanio rerio. Biol. Bull., 75: 119-133. Rudnick, D. 1932 Thyroid-forming potencies of the early chick blastoderm. J. Exp. Zool., 62:287- Silvie 1938a Contribution to the problem of neurogenic potency in post-nodal isolates from chick blastoderms. J. Exp. Zool., 78:369-383. 1938b Differentiation in culture of pieces of the early chick blastoderm. I. Anat. Rec., 70: 351-368. 1938c Differentiation in culture of pieces of the early chick blastoderm. II. J. Exp. Zool., 79:399-427. 1944 Early history and mechanics of the chick blastoderm. Quart. Rev. Biol., 79:187-212. 1945 Limb-forming potencies of the chick blastoderm: including notes on associated trunk structures. Trans. Conn. Acad. Arts & Sci., 36: 353-377. 314 Rudnick, D. 1948 Prospective areas and differ- entiation potencies in the chick blastoderm. Ann. N. Y. Acad. Sci., 49:761-772. , and Rawles, M. E. 1937 Differentiation of the gut in chorio-allantoic grafts from chick blastoderms. Physiol. Zool., 10:381-395. Rulon, O. 1935 Differential reduction of Janus green during development of the chick. Proto- plasma, 24:346-364. Saunders, J. W., Jr. 1948 The proximo-distal sequence of origin of the parts of the chick wing and the role of the ectoderm. J. Exp. Zool., 108: 363-403. Schlesinger, A. B. 1952 Analysis of growth of the chick marginal blastoderm. Science, 116:64— 65. Seevers, C. H. 1932 Potencies of the end bud and other caudal levels of the early chick embryo with special reference to the origin of the metane- phros. Anat. Rec., 54:217-246. Spek, J. 1933 Die bipolare Differenzierung des Protoplasmas des Teleosteer-Eies und ihre Entste- hung. Protoplasma, 18:497-545. Spemann, H., and Falkenberg, H. 1919 Uber asymmetrische Entwicklung und Situs inversus viscerum bei Zwillingen und Doppelbildungen. Roux’ Arch. Entw.-mech. 45:371-422. Spratt, N. T., Jr. 1942 Location of organ-specific regions and their relationship to the development of the primitive streak in the early chick blasto- derm. J. Exp. Zool., 89:69-101. 1946 Formation of the primitive streak in the explanted chick blastoderm marked with car- bon particles. J. Exp. Zool., 103:259-304. 1947 Regression and shortening of the primitive streak in the explanted chick blasto- derm. J. Exp. Zool., 104:69-100. 1949 Nutritional requirements of the early chick embryo. I. J. Exp. Zool., 110:273- 298. 1952 Localization of the prospective neu- ral plate in the early chick blastoderm. J. Exp. Zool., 120:109-130. , and Condon, Leon 1947 Localization of prospective chorda and somite mesoderm during regression of the primitive streak. Anat. Rec., 99: 653. Stockard, C. R. 1907 The artificial production of a single median cyclopian eye in the fish em- bryo by means of sea water solutions of mag- nesium chlorid. Roux’ Arch. Entw.-mech., 23: 249-258. Street, S. F. 1937 The differentiation of the nasal EMBRYOGENESIS: PROGRESSIVE DIFFERENTIATION area of the chick embryo in grafts. J. Exp. Zool., 77:49-80. Trinkaus, J. P. 1949 The significance of the periblast in epiboly of the Fundulus egg. Biol. Bull., 97:249. 1951 A study of the mechanics of epiboly in the egg of Fundulus heteroclitus. J. Exp. Zool., 118:269-320. Tung, T. C., Chang, GC. Y., and Tung, Y.F.Y. 1945 Experiments on the developmental potencies of blastoderms and fragments of teleostean eggs separated latitudinally. Proc. Zool. Soc. London, 115:175-188. , and Tung, T. F. Y. 1944 The develop- ment of egg-fragments, isolated blastomeres and fused eggs in the goldfish. Proc. Zool. Soc. Lon- don, 114:46-64. Twiesselmann, F. 1938 Expériences de scisson précoce de l’aire embryogéne chez le Poulet. Arch. Biol., 49:285-367. Waddington, C. H. 1932 Experiments on the de- velopment of chick and duck embryos cultivated in vitro. Phil. Trans. Roy. Soc. London, B221: 179-230. 1933 Induction by the endoderm in birds. Roux’ Arch. Entw.-mech., 128:502-521. 1935 The development of isolated parts of the chick blastoderm. J. Exp. Zool., 71:273-288. 1937 The determination of the auditory placode in the chick. J. Exp. Biol., 74:232-239. 1952 The Epigenetics of Birds. Univer- sity Press, Cambridge, England. , and Schmidt, G. A. 1933 Induction by heteroplastic grafts of the primitive streak in birds. Roux’ Arch. Entw.-mech., 128:522-563. , and Taylor, J. 1937 Conversion of pre- sumptive ectoderm to mesoderm in the chick. J. Exp. Biol., 14:335-339. Wetzel, R. 1929 Untersuchungen am Hihnchen. Die Entwicklung des Keims wahrend der ersten beiden Bruttage. Roux’ Arch. Entw.-mech., 119: 118-321. Willier, B. H., and Rawles, M. E. 1935 Organ- forming areas in the early chick blastoderm. Proc. Soc. Exp. Biol. & Med., 32:1293-1296. Woodside, G. L. 1937 The influence of host age on induction in the chick blastoderm. J. Exp. Zool., 75:259-282. Yamamoto, T. 1938 Contractile movement of the egg of a bony fish, Salanx microdon. Proc. Imp. Acad. Tokyo, 74:149-151. Zwilling, E. 1942 Restitution of the tail in the early chick embryo. J. Exp. Zool., 97:453-463. Section VI CHAPTER 3 Selected Invertebrates RAY L. WATTERSON ProcressivE differentiation involves changes in the constitution of regions of the egg or of cells or groups of cells which amount to either permanent gains or permanent losses. These changes are accompanied by temporal and spatial restrictions of potencies of the several parts of the egg and embryo as development proceeds. VISIBLE DIFFERENCES ALONG THE ANIMAL-VEGETAL AXIS OF POLARITY The earliest visible differences in the eggs of many aquatic invertebrates are those which occur along one heteropolar axis, the animal-vegetal axis of polarity. Similarly, the earliest invisible differentiation in these eges, as revealed by experimental analysis, usually occurs along this same axis. This axis stands revealed in a variety of ways. Some eggs are more or less elongated along the animal-vegetal axis. The specific weight of the animal half is frequently less than that of the rest of the egg. The germinal vesicle is usually displaced towards the animal pole. Maturation spindles form nor- mally at the animal pole and the polar bodies mark the animal pole insofar as they remain at their point of origin. In many forms a micropyle exists in the chorion at the animal pole, in others at the vegetal pole. More or less striking accumulations of plasm (pole plasms) may occur at the animal pole or at the vegetal pole, or at both. Distinct transverse pigment bands may appear per- pendicular to this axis. In some eggs, whether or not a micropyle exists, the sperm enters preferentially in the animal half, in others at the vegetal half; this is a further indi- cation of structural and/or physiological differences at different levels of this axis. Visible differences along the animal- vegetal axis arise at different developmental stages in eggs of different animals. In some eggs, for example those of Dentalium, a 315 mollusk, such differences are evident before maturation occurs (Wilson, ’04a); in these living eggs there is recognizable a white animal pole region, a broad middle trans- verse pigment girdle, and a white vegetal pole region. In other eggs, changes occur during maturation which reveal the axis of polarity more clearly. For example, in Paracentrotus, a sea urchin, the immature egg contains a uniform distribution of red pigment in the ectoplasm, but after matura- tion this pigment is concentrated and re- stricted to a distinct subequatorial transverse band (Boveri, 01) which has proved to be so useful as an indicator of polarity in the hands of Horstadius. Similarly, in the leech, Clepsine, the visibly distinct animal and vegetal pole plasms are first in evidence after maturation; prior to this time the dis- tribution of yolk and plasm provides no visible evidence of polarity (Schleip, *14). In still other eggs, for example those of the tunicates, the polar axis is only weakly in- dicated in early stages, but fertilization re- sults in redistributions of visibly different constituents within the egg until their strati- fication clearly reveals the polar organiza- tion. Thus in Styela (Conklin, ’05a) the yolk then lies at the animal pole, the yellow ma- terial at the vegetal pole, and in between these two layers there is a zone of clear cytoplasm. Cleavage planes can usually be described readily in relation to the axis of polarity which is revealed as described above: cleav- age planes cut through this axis, run parallel or perpendicular to it, micromeres are cut off at the animal pole (annelids, mollusks, ctenophores) or at the vegetal pole (echino- derms, ctenophores), etc. (see Costello, Sec- tion V, Chap. 2). Even more important than this, the animal- vegetal axis of polarity is the chief axis of differentiation. As a rule the material situ- ated near the animal pole may be traced to 316 EMBRYOGENESIS: PROGRESSIVE DIFFERENTIATION ectodermal structures, whereas material around the vegetal pole becomes incorporated in entodermal structures. A question im- mediately arises as to whether the develop- mental capacities of animal and vegetal halves differ from the earliest stages or whether these differences arise progressively. To date operations have not been possible on ovarian eggs, but certain experiments on uncleaved eggs provide some information bearing on this question. EXTENT OF SPECIFICATION OF CYTOPLASM ENCOUNTERED IN UNCLEAVED EGGS Unfertilized eggs of the nemertine, Cere- bratulus, can be separated into animal and vegetal halves or into meridional halves. Horstadius (’37a, p. 335) summarizes his studies on the developmental capacities of such halves as follows: “Our isolation of animal and vegetative halves prior to fertili- zation confirms the results of Walson (’03) and Yatsu (’10),—that any fragment of the unfertilized egg may develop into a pilid- ium.’’* However, evidence from defect ex- periments on the egg of Cerebratulus indi- cates that cytoplasmic specification increases progressively in the uncleaved egg from fertilization up to the first cleavage (Yatsu, 04). Non-nucleated regions of the cytoplasm were removed at each of the following stages: before dissolution of the germinal vesicle, at the metaphase of the first meiotic division. at the period of conjugation of the egg- and sperm-nuclei, and after constriction of the first cleavage had appeared. For the first three developmental stages listed above the percentages of normal pilidia decreased pro- sressively as follows: 85.7 per cent, 52.3 per cent and 24 per cent: too few operations were carried out at the fourth stage to provide significant percentage values. Yatsu inter- preted these results as demonstrating that there must take place some_ progressive changes in the general make-up of the ese during the period extending from the time of dissolution of the serminal vesicle to the union of the germ nuclei, in other words. that cytoplasmic localization has been pro- gressively established.+ * Similarly animal and vegetal halves, as well as meridional halves of unfertilized eggs of the ascidian Ascidiella scabra, can gastrulate and de- velop into remarkably normal tadpole larvae, althoush there are many exceptions to this state- ment (Dalcq, 738). + Centrifugation experiments on uncleaved eggs of the mollusks Physa ancillaria and Lymnaea cata- By contrast, isolated animal halves and vegetal halves of unfertilized eggs of other invertebrates develop into larvae which dif- fer from each other and from normal larvae in certain characteristic ways. For example, Hérstadius (28, ’37b) has demonstrated that isolated animal halves of unfertilized sea urchin eggs (Paracentrotus, Arbacia) de- velop into blastulae with enlarged apical tufts and that these blastulae fail to gastru- late or form a skeleton, whereas isolated vegetal halves of the same eggs gastrulate and form a skeleton and develop in some cases into quite normal plutei, although usually they develop into larvae with an ovoid body shape, no mouth, and a mal- formed skeleton. By contrast, meridional halves of unfertilized sea urchin eggs de- veloped into more or less typical plutei.t Similarly Wilson (’04a) has demonstrated characteristic differences in development of isolated animal and vegetal halves of un- fertilized eggs of the mollusk, Dentalium, whereas such differences appear to be lacking in isolated meridional halves. He states (p. 69), “Fertilized fragments of the unseg- mented unfertilized egg, obtained by hori- scopium (Conklin, 710) and Physa heterostropha (Clement, ’38) seemed to indicate that the effects of centrifuging are least injurious just after the com- pletion of maturation and most injurious just before the first cleavage. Again it was suggested that this might indicate increasing differentiation of the egg with time prior to the first cleavage. However, Raven and Bretschneider (42) obtained no definite increase in injurious effects with time in their centrifugation experiments on Limnaea stagnalis; thus their experiments lend no support to the hy- pothesis of a progressively increasing differentiation of ooplasm between maturation and first cleavage; they feel that their results are more critical since their material was examined histologically whereas that of Conklin and Clement was not. + Contrary to the above results, Taylor and Ten- nent (24, p. 205), using eggs of the sea urchin Lytechinus (Toxopneustes) variegatus, find that “From many pairs of both horizontal and vertical sections we obtained blastulae with mesenchyme, normal gastrulae with triradiate skeletal spicules, and plutei which, except for size, could not be dis- tinguished from those developing from normally developing entire eggs.” Tennent, Taylor and Whitaker (29, p. 4) stated, “. . . the results showed that both halves of the egg, no matter what the plane of section, might form normal identical larvae. The animal halves, like the vegetal halves, produced larvae with mesenchyme and an archen- teron.” Horstadius (’37b) feels strongly that such results must have been due to some error in orienta- tion of the cut or to a rotation of the egg during cut- ting, and his arguments are convincing. SELECTED INVERTEBRATES zontal or oblique section, differ in develop- ment according as they do or do not contain the lower white area. The upper fragment . . . produces a larva similar to the lobeless ones. ‘The lower one . . . may produce a normally formed dwarf trochophore. Frag- ments obtained by vertical section through the lower white area may . .. produce nearly normally formed dwarf trochophores.” Taken together these experiments on un- cleaved eggs suggest that in the develop- mental history of the egg there is probably no structural organization along an animal- vegetal axis at first, that such a pattern prob- ably arises progressively, and that it may be realized earlier in some uncleaved eggs (sea urchin, mollusk) than in others (nemer- tine, tunicate). This is the same conclusion that was reached concerning the establish- ment of visible differences along the axis of polarity. ARE EXTRINSIC OR INTRINSIC FACTORS RESPONSIBLE FOR ESTABLISHMENT OF AN AXIS OF POLARITY? As a general rule the attached end of the egg in the ovary becomes the vegetal pole, the free end the animal pole. There are those who believe that this relationship is a causal one, such that the different environments at the two ends of the egg cause those ends to develop differently. If this were true, the basic organization of the egg along the animal-vegetal axis would be imposed upon it from the outside. Child (41) has empha- sized the importance of extrinsic factors in the establishment of polarity in animal eggs. Although such a causal relationship may actually exist, it has not been proved experi- mentally for eggs of invertebrate animals, and such proof would be difficult to obtain. It is equally difficult to prove that such a causal relationship does not exist. Be this as it may, one important feature of egg organization, at least in snails, ap- pears to be controlled by intrinsic factors, 1.e., by the genotype of the unreduced egg. The resulting egg organization in some way controls (1) the position of the cleavage spindles for the second cleavage, whether dexiotropic or laeotropic; (2) the position of the primary mesoderm cell, 4d, whether to the left or the right of the first cleavage plane; (3) the direction of coiling of the visceral mass and shell, whether clockwise or counterclockwise. Sturtevant (723) has suggested that one pair of genes is involved, a dominant dextral factor and a recessive 317 sinistral factor. In terms of the usual Mende- lian inheritance a cross between two hetero- zygous parents should give 3 dextral: 1 sinistral offspring; actually all are dextral. A cross between a homozygous recessive female from the above cross and a homozy- gous dominant male should give all dextral offspring; actually all are sinistral. Thus the genotypic make-up of the egg seems to have exerted a permanent effect upon the egg organization prior to onset of the maturation divisions. Even though Sturtevant’s inter- pretation may not account for the results of all crosses (consult Morgan, ’27), this does not detract from the generalization that in- trinsic, as well as extrinsic, factors play im- portant roles in the establishment of egg organization. STRATIFICATION OF EGG CONSTITUENTS ALONG THE AXIS OF POLARITY IS THE RESULT, NOT THE CAUSE, OF POLAR ORGANIZATION It can be demonstrated experimentally by centrifugation that the specific distribution assumed by the various visible constituents of the egg does not create the basic polar organization which underlies regional dif- ferentiation, but is a consequence of that polar organization. The latter basic organiza- tion remains unchanged if a new axis of stratification of visible constituents is im- posed by centrifugation (although the posi- tion of maturation spindles and cleavage planes may be modified to conform to the axis of centrifugation). Thus in sea urchin eggs the axis of stratification imposed by centrifugation was found to bear every pos- sible relationship to the axis of differentiation (Morgan, ’27), and a similar situation characterized the eggs of the annelids Chaetopterus and Nereis (Lillie, ’09), of such mollusks as Cumingia (Morgan, ’27), Physa and Lymnaea (Conklin, ’10), and of the tunicate Styela (Conklin, ’31), etc.* It therefore appeared that polar organization of the egg was a property of the ground sub- stance, by which Lillie (’06, p. 156) meant the “fluid that contains and suspends all the eranules and droplets.” According to Lillie (09), polarity must depend on some definite architecture of the ground substance, and movements of visible constituents produced by the centrifuge cannot be mass movements of entire protoplasmic areas, but only gran- ule movements through the ground sub- * Raven and Bretschneider (’42) disagree with this conclusion. 318 EMBRYOGENESIS: PROGRESSIVE DIFFERENTIATION stance, whose structure is therefore not essentially disturbed by centrifuging. The possibility that this architecture of the ground substance might involve liquid crys- tals has been discussed (Needham, ’42; Har- rison, °45). Conklin (’24) stated that the organization of the egg resides in the internal framework of entoplasm and the cortical layer of ectoplasm. EVIDENCE THAT THE STRUCTURAL BASIS OF EGG ORGANIZATION RESIDES IN THE EGG CORTEX If some permanent internal framework exists in the entoplasm of the ground sub- stance, it should be possible to obtain evi- dence of its existence from viscosity measure- ments (Howard, ’32). The presence of cohering molecular aggregates ramifying through a liquid phase would be expected to cause plastic flow of the system as a whole, as a result of a certain amount of mechanical rigidity to shear which is characteristic of such structures. Centrifugation experiments on two sea urchins (Arbacia punctulata and Strongylocentrotus purpuratus) revealed no unequivocal plasticity and led Howard to conclude (p. 368) that “. . . no continuous structure is present which can significantly affect diffusion within the resting egg.” If so, there is no constant protoplasmic structure in the entoplasm which could account for the relatively stable organization of the egg. It has been noted that usually the ectoplasm (cortex) of the egg is relatively unaffected by centrifugation experiments, and it was nat- ural to suggest that the ultrastructural basis for egg organization must be localized in the egg cortex (Motomura, ’35; Raven and Bretschneider, ’42; and others). The impor- tance of the ectoplasm in this respect would also seem to be indicated by the work of Horstadius, Lorch and Danielli (750), who withdrew large quantities of entoplasm from the sea urchin egg (reducing the volume of the egg by more than 50 per cent) and still obtained normally proportioned plutei; never- theless, these authors seem reluctant to sug- gest that their experiments provide any conclusive evidence on the problem. If the basic polar organization of the egg does reside in the cortex, it would be essential that at least some components of the cortex maintain their relative positions throughout cleavage; otherwise cleavage would disrupt such an organization. Runnstrém (’28a, ’28b) has described a dark-field ring in the cortex of the vegetal half of the sea urchin egg which maintains its position unchanged throughout cleavage; during gastrulation the cells in which this dark-field ring is localized largely become invaginated to form the walls of the archenteron. Accordingly, part of the cortex, at least, seems not to be dislocated during formation of cleavage furrows and could serve as the locus of the basic organi- zation of the egg along the animal-vegetal axis (see Lehmann, 45). In general it may be said that the axis of differentiation rarely becomes dissociated from the original axis of polarity; conse- quently it is of basic importance for an un- derstanding of differentiation that we arrive at some concept of the ultrastructure of the egg responsible for the basic polarity made visible by morphological and physiological differences. Recent speculations by Weiss (49a, ’49b, °50) in terms of molecular ecol- ogy provide a model of the way in which an initial organization of the egg crust could impose organization upon the more fluid egg core. SPECIFICATION OF BLASTOMERES ACCORDING TO THE PORTION OF THE CYTOPLASM SEGREGATED WITHIN THEM The most striking feature about the de- scriptive embryology of certain annelids, mollusks, ctenophores and tunicates is the precocious localization of visibly different cytoplasmic areas and the segregation of these areas into specific blastomeres or groups of blastomeres. In many instances isolation and transplantation experiments prove that such blastomeres or groups of blastomeres differ markedly from other blastomeres in their developmental capacities. ANNELIDS AND MOLLUSKS Experiments on Eggs Lacking Polar Lobes and Pole Plasms. Especially clear-cut evi- dence for specification of blastomeres accord- ing to the portion of the cytoplasm segre- gated within them comes from experiments on the eggs of the mollusk Patella (Wilson, ’04b) and the annelid Nereis (Costello, ’45). The trochophore larva of Patella is illustrated in Figure 120A. If a single cell of the first quartette of micromeres is isolated (Fig. 120B), it develops into a partial larva con- sisting of four primary trochoblasts, two smaller secondary trochoblasts, two apical tuft cells, and a group of non-ciliated ecto- blast cells (Fig. 120C). The first quartette of micromeres divides into four upper cells SELECTED INVERTEBRATES (fa!-1d!) and four lower cells (1a?-1d?); see Figure 120D. If 7a‘ is isolated (Fig. 120£Z), it develops into two secondary trochoblasts, several non-ciliated cells and a small number of apical tuft cells, whereas an isolated fa? S19 are isolated or remain as part of the whole egg. Each cell or group of cells differentiates in a specific way because specific portions of a heterogeneous cytoplasm are segregated into them. Fig. 120. Specification of blastomeres in molluscan eggs. (A, C, E and F from Wilson, ’04b; G, H and J from Wilson, ’04a.) A, Normal trochophore larva of the mollusk Patella. B, Generalized scheme of cleavage in annelids and mollusks (from Wilson, 1899). The first somatoblast (2d) is indicated by sparse stippling; the second somatoblast (4d) is indicated by heavy, dense stippling. C, Differentiation of isolated cell of first quartette of micromeres of Patella. D, Generalized scheme of cleavage of first quartette of micromeres. E, Differentiation of isolated 1a? cell of Patella. F, Differentiation of isolated 1a? cell of Patella. G, Normal trochophore larva of the mollusk Dentalium. H, Type of larva which differentiates from the Dentalium egg lacking first polar lobe, or from the following cells when isolated: AB, A, B, C, 1a, 1b or 1c. Note absence of apical tuft and post-trochal region. J, Type of larva which differentiates from the Dentalium egg lacking only the second polar lobe, or from an isolated 1d cell. Note that only the post-trochal region is lacking. blastomere (Fig. 120F) divides only twice to produce four primary trochoblasts; etc. Es- sentially the same results were obtained with the Nereis egg except for minor differences which correspond to slight differences in cell lineage in these two animals. Thus with each successive cleavage, cells become more sharply specified for one fate only, and that fate is realized whether the cells involved Experiments on Eggs Possessing Polar Lobes. In the eggs of some annelids (Sabellaria, Chaetopterus) and mollusks (Dentalium, Ilyanassa) a peculiar phenomenon known as polar lobe formation occurs, the first polar lobe appearing just before the first cleavage furrow, the second polar lobe appearing just as the second cleavage is initiated. The cyto- plasmic contents of the polar lobes enter the 320 EMBRYOGENESIS: PROGRESSIVE DIFFERENTIATION CD cell when the first cleavage is completed and the D cell when the second cleavage has terminated. The trochophore larva of Dentalium is illustrated in Figure 120G. If, in Dentalium (Wilson, ’04a), the AB blasto- mere or the A, B, or C blastomeres, all of which lack the substance of the polar lobes, are isolated, they develop into larvae lacking an apical tuft and a post-trochal region (Fig. 120H). The same type of defective larva develops if the first polar lobe is extirpated while the first cleavage furrow is forming. Thus it appears that the formation of apical tuft and post-trochal region is dependent upon that part of the cytoplasm contained in the first polar lobe. If the first polar lobe is not removed, but instead the second one is extirpated, the resulting larva develops an apical tuft, but still lacks the post-trochal region (Fig. 1207). Thus it appears that the portion of the cytoplasm essential for apical tuft formation is not present in the second polar lobe, although the portion of the cyto- plasm essential for formation of the post- trochal region is contained therein. At the third cleavage the first quartette of micro- meres forms, and when the cells of this quartette (fa-1d) are isolated, only cell 7d forms an apical tuft (none of the micro- meres forms a post-trochal region). Thus it appears that the portion of the cytoplasm essential for apical tuft formation is progres- sively translocated from the first polar lobe into the CD cell and finally into the 7d blastomere. The latter cell alone is then able to develop into a larva with an apical tuft because a special portion of the heterogene- ous cytoplasm has been segregated into it. Similarly the 7D cell receives the portion of the cytoplasm contained in the second polar lobe, and it alone of the macromeres of the 8-cell stage has the capacity to form the post- trochal region. Strikingly similar results have been obtained with eggs of the annelid Sabellaria by Hatt (’32) and Novikoff (’38), although there are differences in detail. Experiments on Eggs Possessing Pole Plasms. The eggs of Oliosochaeta and Hirudinea are characterized by the accumulation of an area of distinctive protoplasm at the animal pole and another one at the vegetal pole just after the second maturation division; such local- ized protoplasmic areas are termed pole plasms. In Tubifex (Penners, ’22) these two pole plasms are progressively restricted dur- ine cleavage to the CD cell, then to the D cell where they fuse, then to the 7D cell. About half this pole plasm then enters the 2d cell (first somatoblast—see Fig. 120B) of the second quartette and finally enters one of its derivatives (2d!!1), which then consists almost exclusively of pole plasm. This cell divides into 7/7 (to the left of the dorsal midline) and Jr (to the right), and from these two cells all ectodermal structures are derived (ventral nerve cord, cerebral gang- lion, circumesophageal commissures, circular muscle, epidermis, and probably the setal sacs). The rest of the pole plasm (Penners, 24a), after passing through cells 2D and 3D, finally enters cell 4d (second somatoblast— see Fig. 120B) which divides into Myl (to the left of the dorsal midline) and Myr (to the right). From these two cells all mesodermal structures are derived (the somatic meso- derm and the longitudinal musculature de- rived from it, the splanchnic mesoderm, the septa and the nephridia). The possibility exists that the portion of the pole plasm entering cell 2d111 specifies the derivatives of that cell to become ectodermal structures, whereas the portion of the pole plasm enter- ing cell 4d specifies the derivatives of that cell to become mesodermal structures. Blastomeres of Tubifex eggs are very sensi- tive to radiation by ultraviolet light; indi- vidual radiated blastomeres are cast out from the developing system by non-radiated cells (Penners, ’26). As stated above, half of the pole plasm appears to be segregated progres- sively into cells 2D, 3D, 4d and finally Myl and Myr; mesodermal structures fail to form if any of these blastomeres are destroyed. The other blastomeres develop quite normally but parts are frequently displaced from their normal positions relative to one another and metamerism is lacking. Penners (738) con- cluded that the pole plasm passing through the blastomeres listed above specified Myl and Myr for mesodermal differentiation and further that the mesoderm must exert some organizing influence upon the ectoderm and entoderm. Similarly, half of the pole plasm appears to be segregated progressively into blastomeres, 2d, 2d1, 2d11 and 2d11; in early experiments (Penners, 726) ectodermal struc- tures failed to form if any of these blasto- meres were destroyed. The rest of the blasto- meres developed quite normally and Penners concluded that the pole plasm _ passing through these blastomeres specified 2d111 for ectodermal differentiation. However, Pen- ners (’37) discovered that if he kept these embryos alive as long as possible, consider- able regulation occurred in the absence of derivatives of 2d111, with all parts normally forming from this cell now arising from derivatives of Myl and Myr. Thus the meso- SELECTED INVERTEBRATES dermal part of the embryo can regulate to such an extent that it produces derivatives normally of ectodermal origin. Therefore the Tubifex egg is not as strik- ing an example of specification of blasto- meres by segregation of specific portions of the cytoplasm into them as it was once thought to be. Essentially similar results have been obtained with the egg of Clepsine (Leopoldseder, ’31; Mori, 32), although cer- tain differences in detail are encountered. CTENOPHORES Isolation, deformation and defect experi- ments on the ctenophore egg (Driesch and Morgan, 1895a, b; Fischel, 1897, 1898; Zieg- ler, 1898; Yatsu, ’12) give the following re- sults, which indicate a progressive restriction of the capacity to form swimming plates to the first octet of animal micremeres and which indicate that each micromere of the first octet receives the developmental factors necessary for formation of one row of swim- ming plates. Each of the first two blasto- meres, when isolated, develops into a larva with four rows of swimming plates; each of the first four blastomeres into a larva with two rows of swimming plates; each of the first eight blastomeres into a larva with one row of swimming plates. If two macromeres are removed in the 16-cell stage, such that six macromeres and eight micromeres re- mained, the resulting larva possessed all eight rows of swimming plates. From this it was concluded that the morphogenetic fac- tors essential for formation of one row of swimming plates had been segregated into each of the first octet of micromeres (Fischel, 1897). But in the case described by Fischel two of the rows of swimming plates were much shorter than the others, and the pos- sibility was not eliminated that this defi- ciency was due to the lack of the second set of micromeres in two octets, due in turn to the absence of the macromeres (see Schleip, 29). Yatsu (12), however, seems to have furnished evidence that only the first octet of micromeres is concerned with formation of swimming combs. If he separated two cells of the 8-cell stage from the remaining six, a larva with two rows of swimming plates developed from the former. If, however, he separated two macromeres together with their two micromeres from the rest of the blastomeres in the 16-cell stage, and then removed the two micromeres, the two macromeres formed a larva lacking swim- ming plates. a2 Spek (26) has demonstrated that a pro- gressive segregation of corticoplasm into the first octet of micromeres parallels this pro- gressive restriction of potency during cleav- age. This corticoplasm appears green in eggs of Beroé under dark-field illumination. In the fertilized but uncleaved egg it forms a distinct and uniform layer over the egg. As the first cleavage furrow develops the green corticoplasm accumulates at the animal pole, but as this furrow advances towards the vegetal pole, much of the corticoplasm is swept downwards towards the vegetal pole and accumulates there temporarily. It then becomes uniformly distributed again. These shiftings of corticoplasm are repeated in the second division and in the third cleavage, except that in the latter the green cortico- plasm remains as a cap at the animal pole of each cell. At the following cleavage each of these green caps is cut off almost com- pletely into one of the first octets of animal micromeres. Again a segregation of specific portions of the cytoplasm into certain blasto- meres is accompanied by a restriction of certain developmental capacities to those blastomeres. * TUNICATES The greatest variety of visibly distinct and sharply localized cytoplasmic areas ever dis- covered occurs in the egg of the tunicate Styela. The progressive segregation of these areas into the blastomeres is spectacular to watch in living eggs (Conklin, ’05a, b). Un- fertilized eggs contain three protoplasmic re- gions which are differently colored and are distributed around the animal-vegetal axis (Fig. 121A). Following fertilization, striking rearrangements of these regions occur, in- volving a rapid flowing of the yellow proto- plasm first to the vegetal pole (Fig. 121B), then to the future posterior end of the egg where it accumulates as a distinct yellow crescent (Fig. 121C). The animal half then contains clear protoplasm and the vegetal * But some of the green corticoplasm enters the second octet of micromeres and some of it later enters micromeres which arise at the vegetal pole, and these cells are not involved in formation of swimming plates. Moreover, not all derivatives of the first octet of micromeres form swimming plates; some form non-ciliated ectoderm between the eight rows of swimming plates and between swimming plates within each row. Thus only a part of the green corticoplasm appears to be utilized in the formation of swimming combs. Spek emphasized that the green color of the comb-forming cells dif- fered in no way from that in other ectodermal de- rivatives. 322 EMBRYOGENESIS: PROGRESSIVE DIFFERENTIATION NERVE CORD CEREBRAL VESICLE AND SENSORY CELLS J ADHESIVE PAPILLAE Fig. 121. Cleavage stages and prospective fates of specific regions of the egg of the tunicate Styela (Cynthia) partita. In all illustrations the anterior end is to the left, the posterior end to the right. (A, B, C, F, G and I from Conklin, ’05a; D and E from Conklin, ’05b.) A, Unfertilized egg showing germinal vesicle (GV), central gray yolk (oblique hatching), and peripheral yellow protoplasm (dotted). B, Fertilized egg showing accumulation of yellow protoplasm and clear protoplasm at the vegetal pole. C, Uncleaved, fertilized egg with yellow protoplasm accumulated as a crescent across the posterior end of the egg. D, 2-cell stage showing light gray crescent (vertical hatching) across the anterior end of the egg. E, 8-cell stage. F, Animal half of 64-cell stage viewed from the animal pole; vertically-hatched cells represent prospective cerebral vesicle. G, Vegetal half of 64-cell stage viewed from vegetal pole; vertically-hatched cells represent prospective nerve cord; cells hatched vertically and horizontally represent prospective chorda; solid dots represent prospective muscle cells; circles represent prospective mesenchyme. H, Gastrulation stage; dia- grammatic. J, Young tadpole larva. J, Definitive tadpole larva; diagrammatic. half anterior to the yellow crescent contains differences in cytoplasmic composition. dark gray protoplasm. The first cleavage fur- row passes through the mid-sagittal plane (the plane of the paper in Figs. 121C, D), bisecting the yellow crescent, and also bi- secting a light gray crescent which has now become visible around the anterior end of the ege (Fig. 121D), extending slightly above the equator. Thus there is in this egg shortly after fertilization an early, sharp, and visible localization of several indices of underlying Cleavage progressively segregates these re- gions from one another until by the 64-cell stage each cell is “pure” for cytoplasm of one type only (Figs. 121#, F, G). The structures of the tadpole larva derived from each group of distinctively colored cells can be deter- mined at a glance (Figs. 121H, I, J). These ooplasmic regions are sufficiently distinct that every important step of their localization and segregation can be followed in black and SELECTED INVERTEBRATES white photographs of living eggs (Conklin, Was ey Do differences in developmental capacities of blastomeres parallel the visible differences? Conklin (05c, ’06) sought an answer to this question by defect experiments. Eggs in the 2-, 4-, 8-cell stages, etc., contained within the chorion, were strongly spurted with a pipette or were shaken in a vial, whereby some of the blastomeres were injured and others re- mained uninjured and continued to develop. Injured blastomeres were rarely killed, but their nuclei were frequently broken and their chromosomes scattered, such that these cells could not undergo cleavage. By such crude treatment, eggs were obtained with injured and uninjured blastomeres in vari- ous combinations. These could be sorted into groups as follows: eggs with the right half uninjured, the left half injured, and vice versa; those with the anterior half unin- jured and the posterior half injured, and vice versa. In this way Conklin obtained sufficient material to study in detail the development of right and left half embryos, three-quarter embryos, anterior and posterior half embryos, quarter embryos and eighth or sixteenth em- bryos. Regulation of the fragments was limited entirely to closure of the larva (or organ) on the injured side and there was no restitution of missing parts. For example, the anterior half-egg gave rise to an embryo con- sisting of ectoderm, neural plate, sense spots, notochord and entoderm, but lacking muscles and mesenchyme; the posterior half-egg gave rise to an embryo consisting of ectoderm, muscles (at least myoblasts), mesenchyme and entoderm, but lacking neural plate, sense spots and notochord. Each fragment of the egg under these circumstances gave rise to those structures which it would have formed as part of the whole developing egg (see Fig. 121). From results of this sort Conklin concluded that development of the ascidian * No such elaborate color scheme exists in other tunicate eggs, although Berrill (29) reports an orange substance in the eggs of Boltenia hirsuta which segregates into the muscle and mesenchyme cells exactly as does the yellow substance in Styela. In Ciona intestinalis, Phallusia mamillata and Molgula manhattensis the eggs are not colored; nevertheless the same ooplasmic regions can be recognized in stained sections, and have the same prospective fates. Various ooplasmic regions in yet another ascidian egg (Ascidiella aspersa) have the same prospective fates, although the fates of these regions can be demonstrated only with the aid of vital stains (Tung, ’32). In Ascidiella scabra slight differences in the position of organ-forming regions have been demonstrated (Dalcq, ’38). 323 egg is a mosaic work since the individual blastomeres or groups of blastomeres are composed of different kinds of ooplasmic materials, and moreover since the develop- mental fate of any ascidian blastomere or group of blastomeres is primarily a function of its material content. He saw no indication of dependent differentiations or inductions. He reached similar conclusions following defect experiments on the eggs of the Euro- pean tunicate, Phallusia mamillata (Conk- lin, *11). His conclusions received strong support from the studies of von Ubisch (’39a) who removed blastomeres at later stages (32- and 64-cell stages) when each kind of ooplasm was restricted to separate blasto- meres. Systematically removing each group of blastomeres, he found only one instance where the potency of the remaining cells was greater than their prospective fate, viz., ectoblast could give rise to entoderm.+ DIFFERENTIATION WITHOUT CLEAVAGE In annelids, mollusks, ctenophores and tunicates progressive differentiation (i.e., temporal and spatial restriction of potencies) seems to involve primarily a precocious lo- calization of visibly different cytoplasmic areas and the segregation of these areas into specific blastomeres or groups of blasto- meres whose developmental capacities then prove to differ markedly from other blasto- meres. To what extent can differentiation occur if cleavage fails to segregate these localized areas? Lillie (02) was able to suppress cleavage by treating unfertilized or fertilized Chaetopterus eggs with sea water containing potassium chloride. His most striking illustration of the degree to which differentiation can progress without cleavage is illustrated in Figure 122A (which should be compared with the normal trochophore larva of Chaetopterus illustrated in Fig. 122B). His description reads as fol- lows (p. 481): “This structure possesses a certain undeniable resemblance to a trocho- phore: if the smaller hemisphere be com- } Isolation experiments (Berrill, ’32; Tung, ’34; Cohen and Berrill, ’36; Rose, ’39; von Ubisch, 40; Reverberi and Minganti, ’46a,b; Pisand, 49) gave essentially similar results. There was some disagree- ment as to whether cerebral vesicles could form in isolated animal halves, although there was agree- ment that sensory spots could not form in such halves; similarly there was disagreement as to whether entoblast could give rise to ectoderm in iso- lated vegetal halves. 324 EMBRYOGENESIS: PROGRESSIVE DIFFERENTIATION pared to the pre-trochal, and the larger to the post-trochal region, the large vacuoles occupy approximately the position of the prototroch. It is interesting to note that a similar girdle of large vacuoles is found in this position in the trochophore. Continuing the comparison, we may note that the aggre- gation of yolk is in a similar position to the gut of the trochophore.” Unusually strong and active cilia were formed and were ar- ranged with great regularity. Thus consid- erable differentiation can occur when cleay- age fails to segregate sharply localized cyto- plasmic regions. Fig. 122. Differentiation without cleavage in the annelid Chaetopterus (from Lillie, ’02). A, Experi- mental larva; optimal differentiation without cleay- age. B, Control larva. Although similar results have been ob- tained by other investigators, this still re- mains the classic example of this phe- nomenon. This experiment would seem to indicate the primary importance in differ- entiation of the precocious localization of different cytoplasmic areas and the secondary importance of the segregation of these areas by the cleavage process. * CHEMICAL CHARACTERIZATION OF LOCALIZED CYTOPLASMIC AREAS IN MOSAIC EGGS If precocious localization of cytoplasmic areas is of primary importance for the dif- ferentiation of eggs of annelids, mollusks, * Brachet (’37) has demonstrated that such dif- ferentiation without cleavage is impossible when oxidations are inhibited by cyanide, and also when metabolism is interfered with by monoiodoacetic acid, although some localization of cytoplasmic ma- terials can take place in the presence of the latter agent. ctenophores and tunicates, it is essential that the chemical and physical nature of these areas be analyzed by cytochemical methods or by any other methods which can be ap- plied to minute objects. Examples of this approach to the study of differentiation have been illustrated, summarized and evaluated by Needham (’42, pp. 131-140) and by Brachet (750, pp. 271-291). Needham states (p. 139), “. . . in many eggs showing mo- saic development a profound chemical het- erogeneity of the parts of the egg sets in very early, parallel with the early deter- mination. Histochemical and experimental methods confirm each other. The former alone will not carry us much further, but with the micro-chemical technique which should before long be available, it ought to be possible to establish by unimpeachable chemical methods the differences between the determined areas.” A step in this direc- tion has been taken by Berg and Kutsky (51), who studied differences in oxygen uptake of isolated blastomeres and polar lobes in the egg of the mollusk Mytilus edulis. They report a lower respiratory rate of the cytoplasm of the CD blastomere when compared with the AB blastomere and dem- onstrate that the lower rate of the former is due to the lower respiratory rate of the polar lobe cytoplasm which is incorporated into the CD cell. The importance of the polar lobe cytoplasm in development has been stressed above. INTERACTION BETWEEN BLASTOMERES IN DEVELOPMENT IN MOSAIC EGGS In the preceding section emphasis was placed on the independent differentiation of blastomeres in certain phyla, i.e., on the apparent lack of interaction between blasto- meres. Thus far interpretations have been based largely upon results of defect ex- periments or isolation experiments. Atten- tion may now be turned briefly to experi- ments involving transplantation, fusion of eggs, etc., to see whether they provide any evidence for interaction between _blasto- meres in nemertines, annelids, mollusks or tunicates. Horstadius (’37a) found that from the 8-cell stage onward, separation of the egg of the nemertine Cerebratulus into ani- mal and vegetal halves, into an,, ans, veg; and vegs layers, various combinations of these layers of blastorneres, as well as fu- sion of animal halves with meridional halves, provided no evidence for interaction SELECTED INVERTEBRATES of blastomeres. This is in distinct contrast to identical combinations of sea urchin blastomeres, as noted below. Each set of blastomeres simply self-differentiated regard- less of the other blastomeres present si- multaneously. Novikoff (738) combined halves and quarters of the egg of the annelid Sabellaria with another whole egg. The addi- tion of these extra egg fragments gave no inductions in the host; any added fragment simply self-differentiated into those struc- tures characteristic of its prospective fate 325 lation is possible in the eggs of annelids and mollusks under special conditions. It was noted earlier that in the absence of the nor- mal source of ectodermal structures in the Tubifex egg, such structures could arise from cells originally believed to be specified for mesodermal differentiation only (Penners, 38). Also it was noted that in the absence of mesodermal structures metamerism was lacking and ectodermal and _ entodermal structures were displaced, suggesting an or- ganizing influence of the mesoderm upon the WAR Fig. 123. Types of twins experimentally produced in annelids. A, Duplicitas cruciata twin of Chaetopterus (from Titlebaum, ’28). B. Duplicitas cruciata twin of Sabellaria (from Novikoff, ’40). C, Duplicitas cruciata twin of Nereis (from Tyler, ’30). (i.e., an extra A or B cell added extra apical cilia, an extra C cell added an extra apical tuft, and an extra D cell added an extra post-trochal region, etc.). The localization of cytoplasmic factors for development of apical tuft and post-trochal region in the polar lobes of annelids and mollusks has already been discussed; attempts by Novikoff to transplant polar lobes to entire eggs of Sabel- laria were unsuccessful in that nothing additional formed in the hosts to which such polar lobes were attached (although the union of polar lobes to eggs was sufficiently close that vital stain passed from the former into the latter). Fusion of two or more eggs (Hatt, ’31; Novikoff, ’38) did not re- sult in a single giant larva of unit structure but yielded double embryos or more complex monsters, each egg in the fused mass dif- ferentiating independently of the others. In spite of these negative results insofar as demonstrating interaction of blastomeres is concerned, a considerable amount of regu- distribution of ectodermal and entodermal derivatives. Moreover, in Tubifex an isolated CD or D cell containing the pole plasm regu- lates itself into a normally proportioned, though small, worm. But even more striking is the observation that if the vegetative pole plasm be equally distributed to the first two blastomeres (Tubifex, by heat or depriva- tion of oxygen, Penners, ’24b; Chaetop- terus, Nereis and Cumingia by compression, high or low temperatures, centrifugation or anaerobiosis, Titlebaum, ’28, Tyler, ’30; Sabellaria by treatment with potassium chloride, Novikoff, ’*40), double monsters of the duplicitas cruciata type are formed (Fig. 123), or the blastomeres may, if isolated, give rise to more or less complete larvae. Costello (45) has emphasized that the treatment must do more than distribute pole plasm equally to the first two blastomeres; it must set up an effective barrier between the two cells such that each of the first two blastomeres and their derivatives do not 326 EMBRYOGENESIS: PROGRESSIVE DIFFERENTIATION react with one another, but give rise to a separate individual. This would seem to im- ply that they do interact with one another under normal circumstances. Tung (34) has accomplished the fusion of two tunicate eggs in the 2-cell stage, ob- taining results which favor rigid mosaic development in that two complete sets of organs formed. However, von Ubisch (’38, especially ’39a) was able to produce giant tadpole larvae from fusion of two tunicate eggs oriented in such a way that their ani- mal-vegetal axes were parallel. It appeared that a point-to-point fusion of identical cyto- plasmic areas occurred, with each such area then forming an organ larger than normal. Thus no evidence for interaction of blasto- meres was furnished by either investigator. Nevertheless there is evidence from other experiments for the induction of at least two structures in the tadpole larva of tuni- cates, viz., adhesive papillae and sensory cells. Tung (’34) separated the animal half from the vegetal half in the 8-cell stage and rotated the former through 90 to 180 degrees before replacing it. Adhesive papillae were frequently absent following such rotation; when present they always occurred at the original anterior extremity of the vegetal half, although they originate from cells of the animal half. This means that they have formed from ectoderm which does not nor- mally produce them and that the stimulus for their formation originated in the an- terior vegetal cells. This appears to be an example of true induction. Similarly in cen- trifugation experiments in which various structures underwent relatively normal his- togenesis, but occurred in abnormal rela- tionships to one another, Tung, Ku, and Tung (41) observed that adhesive papillae differentiated only from ectoderm which was in contact with entoderm (no exception in 49 cases). Reverberi and Minganti (751) agree that the stimulus necessary for de- velopment of adhesive papillae emanates from anterior vegetal cells, but they find, in contrast to Tung (734), that such adhesive papillae can be stimulated to form only from anterior animal cells, i.e., from cells which normally form adhesive papillae. They pre- fer, therefore, to speak of the stimulus as an “evocation” rather than as an “induc- tion.” Similar evidence is available that inter- action of blastomeres is involved in the dif- ferentiation of sensory cells. Isolated lateral halves of the 8-cell stage form sensory cells, as do isolated anterior halves; by contrast, isolated posterior halves or isolated animal and vegetal halves do not form sensory cells (Berrill, °32; Tung, 34; von Ubisch, 40). Such cells originate from the animal half, but it appears that some stimulus from the anterovegetal half is necessary for their dif- ferentiation. This conclusion is supported by Rose (739), who found that cerebral vesicles and sensory cells formed in 15 of 46 cases when the two posteroanimal cells were com- bined with the two anterovegetal cells, whereas in reciprocal combinations (two anteroanimal plus two posterovegetal cells) neither cerebral vesicles nor sensory cells formed in 20 attempts.* Rose had some evi- dence that the inductor of the sensory cells was restricted to the prospective entodermal portion of the two anterovegetal cells; how- ever, von Ubisch (40) obtained sensory cells in defect experiments involving the complete removal of prospective entoderm. Reverberi and Minganti (751) agree that some stimulus necessary for development of brain and sensory cells emanates from the anterior vegetal cells; but, as in the case of adhesive papillae, they believe that the development of such structures can be stimu- lated only from cells which usually give rise to them, i.e., only from anterior animal cells. Again, then, they prefer to speak of an “evocation” of brain and sensory cells, rather than of their “induction.” They are convinced that this evocation is mediated by a chemical substance. In the words of Reverberi and Minganti (51), “all other presumptive territories (of muscle cells, notochord, epidermis, mesen- chyme, endoderm) present a complete ca- pacity of self-differentiation, even in the ab- sence of this ‘center,’ that is of the anterior vegetal blastomeres.”’-+ * But there must be species differences in this respect, since Tung (734) obtained differentiation of sensory cells when he rotated the animal half of the 8-cell stage through 180 degrees which would place the anterior animal cells above the two posterior vegetal cells. + Both Rose (739) and von Ubisch (40) noted that myoblasts complete their differentiation into muscle cells only when the anterior vegetal cells are pres- ent together with the posterior vegetal cells from which myoblasts originate. However, it seems likely that this stimulus necessary for complete dif- ferentiation of myoblasts may be simply a mechan- ical one imposed normally by the elongating chorda which originates from cells of the anterovegetal region, and of quite a different nature from the in- ductive stimulus necessary for formation of ad- hesive papillae and possibly of sensory cells. For more recent experiments see Reverberi and Min- ganti (’47). SELECTED INVERTEBRATES From this brief survey it appears that interaction between blastomeres is of minor importance in the development of nemer- tines, most annelids and mollusks; it appears to be essential to the development of a few structures of tunicate larvae, viz., cerebral vesicle, sensory organs, and adhesive papil- lae. By contrast, interaction between blasto- meres appears to be involved in the develop- ment of almost every larval structure of echinoderms. Such interactions have been analyzed most extensively in developing eggs of sea urchins and sand dollars. IN ECHINODERM EGGS That fragments of sea urchin eggs are capable of considerable regulation (as also INCREASING DEGREES __OF ANIMALIZATION ZO—APN-—TPAMOM< 327 ingly, the two meridional halves of the sea urchin blastula can regulate into a normal, though small, pluteus larva. Obviously the fate of individual blastomeres or even of groups of blastomeres must become rigidly fixed much later than in the eggs thus far examined; otherwise those in each half blas- tula could not respond to the altered rela- tionships produced by such experimental interference by collectively producing a nor- mal individual. Existence of Animalizing and Vegetalizing Influences. The sea urchin egg is by no means devoid of organization. Not every part of it can form a whole (as Driesch originally be- lieved); for if the egg is divided into two halves equatorially either before fertilization or at the 8-cell stage by separating the four @ Re IN ISOLATED IN WHOLE EGGS BY CHEMICAL TREATMENT ANIMAL HALVES ZO-HPN-CFPZ-ZP INCREASIN DEGREES OF VEGETALIZATION IN ISOLATED VEGETAL HALVES S IN WHOLE EGGS BY GHEMICAL TREATMENT Fig. 124. Modifications of sea urchin development by operative and chemical procedures (A through J from Ho6rstadius , 35). A—D, Series of animalizations characteristic of isolated animal halves; increasing degrees of animalization from right to left. A’—D’, Series of definitive animalizations arising from transitory types A-D. E—J, Series of vegetalizations characteristic of isolated vegetal halves; increasing degrees of vegetalization from left to right. A—S, Series of vegetalizations of whole eggs produced by LiCl treatment; increasing degrees of vegetalization from left to right (K, M and N from Herbst, 1896; others from Child, 40). T—Z, Series of animalizations of whole eggs produced by treatment with NaSCN;; increasing degrees of animalization from right to left (from Lindahl, ’36). are fused eggs) has long been known. Each of the first two blastomeres, or each of the first four blastomeres can produce, when isolated, a relatively normal, though small, pluteus larva (Driesch, 1891, 1892; Hor- stadius and Wolsky, ’36); even more strik- animal cells from the four vegetal cells, the resulting larvae usually differ considerably, not only from each other but also from the normal pluteus (Fig. 124). The most ex- treme larva derived from the animal half is characterized by (1) an apical tuft which is 328 EMBRYOGENESIS: PROGRESSIVE DIFFERENTIATION too large (Fig. 124A), often covering as much as three-fourths of the surface, al- though it later disappears and the surface becomes uniformly ciliated (Fig. 124A’); (2) the absence of both stomodaeum and ciliated band, normally derived from cells of the animal half; (3) the absence of skele- ton; (4) failure to gastrulate. Such larvae are said to be “animalized” or ‘“ectoderm- ized.” But in less extreme cases the apical tuft is more restricted (Figs. 124C, D) and in such cases a ciliated band (Fig. 124C’) and even a stomodaeum (Fig. 124D’) may de- velop (Horstadius, ’39). The vegetal halves of eggs, the isolated animal halves of which produce overdevelopment of the apical tuft, may form quite normal plutei (Figs. 124£, F), or they may form exogastrulae (Fig. 124G) with poorly developed skeleton and arms. At the other extreme the vegetal halves of eggs, the isolated animal halves of which produce a restricted apical tuft (Figs. 124C, D), may produce very abnormal larvae (Figs. 124H,/,J). The latter consist mostly of gut contained within an ecto- dermal vesicle which may develop a ciliated field at its most animal border (Fig. 124H) and a skeleton which has not advanced be- yond the spicule stage (Figs. 124H,J/) or which may be entirely absent (Fig. 124/). Such larvae are said to be “vegetalized” or ‘“entodermized.” It is not necessary to separate the animal half completely from the vegetal half in order to produce such developmental modi- fications. If the egg is simply stretched along its animal-vegetal axis (Lindahl, °36), or if a constriction is formed by a single silk fiber around the equator of the egg (Hor- stadius, ’38), the partially separated animal and vegetal halves develop much as if they were completely isolated. Thus it appears that some influence from the vegetal half (vegetalizing influence) normally passes into the animal half and exerts the following effects upon development of the latter: (1) restriction of the apical tuft; (2) stimula- tion of stomodaeum invagination; (3) stimu- lation of formation of a ciliated band on the ventral side parallel to the animal-vege- tal axis. Similarly, it appears that some in- fluence from the animal half (animalizing influence) normally passes into the vegetal half and exerts the following effects upon the latter: (1) restriction of the size of the gut; (2) control of skeletal development. Recent experiments (Horstadius, Lorch and Dani- elli, 53) seem to demonstrate that enucleated vegetal halves cannot exert vegetalizing in- fluences, but that enucleated animal halves can still exert animalizing influences. Time Required for Animalizing and Vege- talizing Influences to Exert Their Effects. /:x- periments designed to answer this question can best be summarized by paraphrasing the words of Horstadius (739, pp. 160-161). Animal halves are isolated every second hour from the 16-cell stage (4 hours post-fertiliza- tion) up to the beginning of gastrulation (16 hours post-fertilization); if isolated 4 to 6 hours after fertilization they produce a large apical tuft and develop into ciliated blastulae or blastulae with a ciliated field (Figs. 124 A, A’, B, B’), whereas if isolated later (10 to 16 hours post-fertilization) they produce a restricted apical tuft and develop a ciliated band and stomodaeum (Figs. 124D, D’), the structures which they should pro- duce as parts of the intact egg. The change in differentiation of the animal halves occur- ring about 8 or 10 hours post-fertilization (blastula before formation of primary mesen- chyme) indicates that by that stage the vegetalizing influences have so interacted with the animal half that the latter can then self-differentiate according to its prospective fate. Isolation of the most animal quarter of the egg, corresponding to am,, indicates that vegetalizing influences have modified its fate somewhat later, from 14 to 16 hours after fertilization. This indicates that de- termination, as regards the influence of vegetative material upon the animal, pro- ceeds from the vegetative towards the animal pole. The isolated vegetal halves produce ovoid plutei until 14 to 16 hours post-fertili- zation when suddenly they develop into a new type of larva, a pluteus with long anal arms, but entirely without oral lobe, 1e., they develop as if still part of the entire egg. This change in differentiation indicates that by that stage the animalizing influences have so interacted with the vegetal half that the latter can self-differentiate according to its prospective fate. Do Isolated Animal Halves Lose Their Com- petence to Respond to the Vegetalizing Influ- ences at the Same Time that Capacity for Self-differentiation Is Attained? To answer this question animal halves are isolated from vegetal halves at progressively later stages and micromeres are then added to such iso- lated halves to determine how late a nor- mal pluteus with gut can be induced from animal halves. If four micromeres are added to isolated animal halves even as late as the stage when onset of gastrulation normally occurs (16 hours post-fertilization), a gut SELECTED INVERTEBRATES can still be induced from cells of the animal half, and the necessary shiftings of the posi- tion of stomodaeum and ciliated band can still take place such that an almost normal pluteus results. Thus the animal half is competent to respond to vegetalizing influ- ences of the micromeres at a stage later than it is capable of self-differentiation. This must mean that determination is progressive, that when it has advanced far enough to enable an isolated region to self-differentiate, it is still only a labile determination and can be changed if brought under appropriate in- fluences. * Do Animal Halves and Vegetal Halves Lose Their Capacity to Exert Animalizing and Vegetalizing Influences in Advanced Stages? Horstadius (50) demonstrated that an ani- mal half from an advanced stage (blastula before formation of primary mesenchyme) can still exert animalizing influences on much younger vegetal halves, and similarly that a vegetal half of a beginning gastrula can still exert vegetalizing influences on a younger animal half sufficient to inhibit extension of the apical tuft and evoke for- mation of a ciliated band and stomodaeum. Graded Intensity of Animalizing and Vege- talizing Influences. Transplantation experi- ments (Horstadius, ’39) suggested that ani- malizing and vegetalizing influences were not distributed uniformly throughout the animal and vegetal halves respectively, but existed as overlapping gradients with po- lar concentrations. Animalizing influences seemed strongest in am,, less strong in amo, less still in veg, and least in vegy; vegetaliz- ing influences seemed to be strongest in the micromeres, less strong in veg», weaker still in vegi1 and weakest in the animal hemi- sphere. Normal development appeared to de- pend upon a proper balance between ani- malizing and vegetalizing influences, rather than upon their absolute intensities. The evidence for this has been reviewed by Horstadius (739, °49). Chemical Modifications of Development. Vegetalization (Entodermization) of Entire Eggs. If entire sea urchin eggs are fertilized in normal sea water and are then transferred *Tt should be noted that animal halves retain their competence only if they remain in contact with the vegetal halves until just before micromeres are added: if animal halves are isolated 4 hours after fertilization and the isolated animal halves reniain in isolation until 16 hours vost-fertilization, at which time micromeres are added, the animal halves are no longer competent to respond to the vegetalizing influences. They form no stomodaeum, no ciliated band, no restricted apical tuft and no gut. a29 to sea water containing lithium chloride, many of the resulting larvae resemble in a striking way those derived from isolated vegetal halves (Herbst, 1892, 1893, 1896; Runnstrém, ’28a; Child, °40). This is very evident upon comparison of Figure 124K with 124F, 1242 with 124G, 124M with 124H, 124N with 124/,J, etc. Treatment with lithium apparently shifted the bound- ary between ectodermal and entodermal de- velopment above the equator (i.e., prospec- tive ectoderm was converted to entoderm) such that an excessively large gut devel- oped.+ That this is the correct interpreta- tion is indicated by two types of informa- tion. It is known that primary mesenchyme cells aggregate into two masses at a level along the animal-vegetal axis determined by the ectoderm (von Ubisch, ’39b); a skeletal spicule forms within each such aggregation. Following lithium treatment these micro- mere derivatives aggregate much closer to the animal pole, suggesting that the lithium treatment has restricted ectodermal differen- tiation to a level closer to the animal pole. Moreover, the dark-field ring mentioned earlier as associated with the cortex of pro- spective entodermal cells likewise extends further towards the animal pole in lithium- treated eggs (Runnstrém, ’28a,b). Lithium can also modify development of isolated ani- mal halves in the same direction as micro- meres do (von Ubisch, 729); lithium-treated animal halves can develop into pluteus-like larvae just as they can when micromeres are added. Animalization (Ectodermization) of Entire Eggs. If entire sea urchin eggs are treated before fertilization with calcium-free sea water to which isotonic sodium thiocyanate has been added (Lindahl, ’36), and are then returned to sea water, are fertilized and allowed to develop, many of the resulting larvae resemble in a striking way those de- rived from isolated animal halves. With in- creasing degrees of animalization modifica- tions occur in the following sequence: the cut fails to connect with the stomodaeum + Only such development is true vegetalization. According to Lindahl (42), only a few treatments result in true vegetalization, and of these lithium is definitely most effective. Vegetalization and exo- gastrulation are not synonymous; there are many chemical and physical agents which produce simple exogastrulation of the type illustrated in Figure 124Z without increasing the ratio of entoderm/ ectoderm (see Child, 741, footnote, p. 222). Extreme reduction of ectoderm and enlargement of entoderm occurs when lithium-treated embryos are returned to sea water (Figs. 1240-S). 330 EMBRYOGENESIS: PROGRESSIVE DIFFERENTIATION (Fig. 124¥); the skeleton becomes abnormal ‘Figs. 124X,W); gut fails to form (Figs. 124W, V, etc.) and no coelom forms. With increasing animalization an apical tuft also appears at the most vegetal pole (Fig. 124V), and with continued animalization spreads in proportion as the animal apical tuft spreads (Fig. 124U) until finally the most extreme animalized form (Fig. 1247’) is completely covered with apical tuft cilia. Such modifications constitute animalization, since ectodermal differentiations occupy a greater part of the body than normally. Hypotheses to Account for Vegetalization and Animalization by Chemical Agents. RUNNSTROM-LINDAHL HYPOTHESIS. Runnstrém (33) suggested that the gradient of animal- izing influences decreasing from the animal pole might be the expression of a graded distribution of animalizing substance, and similarly that the gradient of vegetalizing influences decreasing from the vegetal pole might be the expression of a graded distri- bution of vegetalizing substance. Lindahl (36) attempted to account for the origin of these hypothetical substances by suggest- ing that there are two specific kinds of metabolism in the egg, carbohydrate metab- olism in the animal half, protein metabolism in the vegetal half, producing animalizing and vegetalizing substances, respectively. Lindahl assumes that lithium treatment blocks carbohydrate metabolism and_ re- duces the amount of animalizing substance formed, thus reducing the intensity of the animalizing influence and thereby enabling the unmodified vegetalizing influence to turn the development of a larger proportion of the egg than usual in an entodermal direc- tion. And conversely Lindahl suggests that ectodermization is the result of a block to protein metabolism, reducing the amount of vegetalizing substance formed, thereby reducing the intensity of the vegetalizing influence and enabling the unmodified ani- malizing influence to turn the development of a larger proportion of the egg than usual in an ectodermal direction. This hypothesis has stimulated an enormous amount of re- search, far too extensive to be reviewed and evaluated here; fortunately this has been done in recent years by Lindahl (’42), Need- ham (42), Lehmann (45), Horstadius and Gustafson (48), Ho6rstadius (’49) and Brachet (50). Needham evaluates this hy- pothesis as follows (pp. 494-495): ‘That there are two centres of morphogenetic in- fluence and a gradient system which reforms after certain experimental interferences must be regarded as established. It is un- deniable also that the actions of these cen- tres may be imitated by a variety of chemical agencies, and it is at least exceedingly prob- able that the two centres are characterised by two different kinds of metabolism. That they produce each a definite morphogenetic substance or substances is likely. But with the statement that the animal pole is asso- ciated with carbohydrate catabolism and the vegetal pole with that of protein, we reach a point where caution is necessary.” For more recent attempts to demonstrate dif- ferences in metabolism in animal and vegetal halves, and to relate these differences to dif- ferentiation, see Hoérstadius (49), Gustafson and Hjelte (751) and Gustafson and Lenicque C52): CHILD-RULON HYPOTHESIS. Instead of a double gradient system, Child (40) proposes a single gradient system diminishing from the animal towards the vegetal pole. Child interprets the mechanics of vegetalization as follows (p. 29): “If prospective entoderm originates from the lower levels of a primary apico-basal gradient, as dye reduction and differential lethal susceptibility indicate, en- todermization may result from depression or inhibition of prospective ectoderm below a certain ‘physiological level’; that is, the specific difference between ectoderm and entoderm may be a secondary result of a nonspecific, primarily quantitative differ- ence. According to this suggestion, ento- dermization occurs first in the most basal levels of prospective ectoderm and progresses acropetally with increasing inhibition be- cause lower levels of ectoderm require only relatively slight inhibition, higher levels, more extreme inhibition, to bring them down to the entodermal level.” Rulon (41), accepting Child’s concept of a single physio- logical gradient, accounts for ectodermiza- tion by sodium thiocyanate as follows (pp. 312-313): “Exposure of unfertilized eggs to NaSCN in Ca-free sea water apparently re- sults in a general depression or slowing down of physiological processes, the apical or more active region being retarded to a greater extent than the less active region. Such dif- ferential inhibition may be sufficient to de- crease the apical dominance to such a degree that the axiate pattern is partly obliterated. With the stimulation to increased activity, resulting from return to normal sea water and fertilization, the lower levels of the gradient, being wholly or partly physiolog- ically isolated, may recover and develop in the same, or almost the same, way as higher SELECTED INVERTEBRATES levels. In other words, the whole egg may now develop in a manner similar to the original apical portion. Since the apical region normally gives rise to ectoderm and ectodermal derivatives, the whole egg may now give rise to ectoderm alone, with neither mesenchyme nor gut developing.” The chief difference in the approach of these two hypotheses has been an emphasis on the specific effects of chemical treatment on certain aspects of metabolism on the one hand (Runnstrém-Lindahl) versus an em- phasis on non-specific inhibition of vaguely defined physiological activities on the other (Child-Rulon), i.e., an emphasis upon kinds of metabolism versus rates of activities. More recently Rulon (literature summarized 62) has attempted to characterize the ani- mal-vegetal activity gradient in terms of enzymes involved by exposing sand dollar eggs to specific enzyme inhibitors. In commenting upon attempts to correlate metabolism and enzyme activities with dif- ferentiation, Holter (’49, p. 73) stated, ““The indications are that in the stream of chemi- cal and metabolic events that constitute the life of the embryo from fertilization to hatch- ing, the true morphogenetic processes are only like ripples on the surface and their quantitative share in the chemistry of the whole is very small. It seems rather doubtful whether we can hope to reveal the mechan- isms which cause those ripples by studying overall metabolism and general enzyme dis- tribution. We are obtaining very interesting results as to the general biochemistry of the egg and embryo, but the crucial problems of morphogenesis may be beyond reach of the enzyme chemist.” In spite of this dis- couraging outlook some advance seems to have been made towards analysis of deter- mination of dorsoventrality in terms of en- zyme activities. ESTABLISHMENT OF DORSOVENTRALITY AND BILATERALITY Most animal eggs reveal earlier or later a bilaterally symmetrical organization. There is then recognizable a median plane separating the egg into right and left halves. In some echinoderms, for example the holothurians Psolus phantopus and Cucu- maria frondosa (J. and S. Runnstrém, ’21), bilaterality of the egg is revealed visibly even in oocyte stages. Much more commonly the median plane stands revealed visibly and becomes rigidly determined relatively late in development compared with the axis of 3a0 polarity. Nevertheless there is some evidence both in tunicates and in sea urchins that some steps in the determination of the plane of bilateral symmetry have occurred at very early stages. In the egg of the tuni- cate Styela (Conklin, ’05a) the sperm can enter at any point within 30 degrees of the vegetal pole; the sperm nucleus then moves towards one side of the egg and much of the yellow protoplasm located at that time at the vegetal pole (Fig. 121B) is drawn over with it, forming the yellow crescent just below the equator (Fig. 121C). The center of the yellow crescent indicates the future median plane. The sperm nucleus often travels through a considerable distance to reach this side of the egg and Conklin believes it is traveling towards a preformed area, in other words, that the meridian which becomes the median plane is already fixed in the uncleaved egg. Certain experi- ments on development of fragments of unfer- tilized tunicate eggs (Dalcq, °38) likewise suggested the possibility that bilateral sym- metry is already determined in the tunicate egg before fertilization. Foerster and Orstrém (33) have presented physiological evidence for the early existence of a dorsoventral organization in fertilized (uncleaved) sea urchin eggs. If such eggs are exposed to potassium cyanide a smaller and a larger hollow arise opposite each other; these hol- lows later disappear, but if the position of the larger one is marked by vital stain, the resulting larva is stained ventrally. Such an experiment suggests a dorsoventral dis- tribution of some condition of the cytoplasm which gives the same reaction to potassium cyanide, but a more vigorous one ventrally. Crowding of eggs, reducing availability of oxygen, has the same effect. If such hollows are suppressed by treatment with anionic detergents (Gustafson and Savhagen, ’50), the resulting larvae are radially symmetri- cal. Moreover, complementary deficiencies which arise in meridional halves of unfer- tilized sea urchin eggs also indicate that a bilateral organization is already estab- lished in this egg before the entrance of the sperm (Ho6rstadius and Wolsky, ’36). There is also evidence that external fac- tors can control the position of the median plane. Just (12) has demonstrated that the first cleavage plane consistently passes through the sperm entrance point in the egg of Nereis (an annelid); the first cleavage plane is perpendicular to the longitudinal axis of the body. Just concludes (pp. 250- 251), “Since . . . the sperm may enter at 432 EMBRYOGENESIS: PROGRESSIVE DIFFERENTIATION any point and since the first cleavage plane passes through this point, the structure ot the ovocyte or Nereis at the time or insemina- tion must be the same in all meridians.” Morgan and ‘lyler (’5U) have investigated this relationship in more detail in annelids and mollusks. Ditterences ot opinion exist as to the relation between the sperm en- trance point and the median plane in the sea urchin egg. Horstadius (’39) stained the side opposite the point of sperm entry and found this point in all positions in relation to the first cleavage furrow. Since the first cleavage furrow in sea urchin eggs has no rigid relation to the future median plane, it remains uncertain how the sperm entrance point is related to the median plane. Lillie (09) examined development ot fragments of unfertilized Chaetopterus (annelid) eggs separated by centrifugation. He believed that his study of cleavage pattern in these frag- ments indicated that bilaterality developed subsequent to their fertilization. He con- cluded (p. 05), “. . . if bilaterality can de- velop in the parts there is no reason for assuming its prior existence in the whole; and its origin must be regarded as a purely epigenetic process.” Similarly Pease (38), upon discovery that he could control the position of the ventral surface in eggs of the annelid Urechis by centrifugation, con- cluded (p. 422), “It seems best to regard this bilateral determination as induced in the egg and not as a rotation or shift of a bilateral axis or predisposition already pres- ent in the egg, although such may be present in a weakly defined form.” The most detailed studies of determina- tion of dorsoventrality are those on sea urchin and sand dollar eggs. If sea urchin eggs are stretched perpendicular to the ani- mal-vegetal axis by forcing eggs through a fine capillary tube (Lindahl, ’32a) or by constricting them meridionally by a single silk fiber (Horstadius, °38), the axis of stretching becomes the dorsoventral axis. In the experiments with the capillary tubes Lindahl found that the end of the stretched egg which usually develops into the ventral surface becomes the dorsal surface instead if it is overstained with vital stain. Simi- larly in centrifugation experiments the dorsoventral axis coincides roughly with the axis of centrifugation (Runnstrém, ’26; Lin- dahl, ’32b; Pease, ’39). Pease concluded from his experiments on sand dollar eggs that two factors are involved in ventral deter- mination: (1) a ventral determinant present in gradient form with its greatest intensity Ou the prospective ventral side; this he con- siuered tO ve located in thle egg Coctex aru 1iut smiled py Centricugauion; (2) a sup- suvate 10ocated In tne enopiasin, UduormMy uistcivuted In a normal egg, out concentrated at thle centripetat pole 10 centrituged eggs. ase suggestea tnat the ventrai determinant luay ve an enzyme adsorbed in tne Cucticat aayers, 11e Proceeded to test this Nypotnesis (41, 42a, bd) by exposing eggs in tne &-ceLL stage to gradients of Known enzyme 1n- nipitors. Vyith certain specific enzyme inhip- itors the ventral surtace tormed only on tne side away Irom the source of the inhibitors, suggesting that the enzyme systems con- cerned specifically with dorsoventral deter- mination will cause a ventral side to form wherever their activity is greatest. It has been known for a long time that many chemical treatments of echinoderm eggs com- pletely suppress bilaterality and produce radially symmetrical forms; more recent ex- periments by Rulon (749, ’51, listed in Rulon, 92) with specific enzyme inhibitors, fol- lowed by return of the treated eggs to sea water, seemed to cause a spread of ventral determination over a greater portion of the egg than usual, i.e., they caused a “‘ventrali- zation” of the egg. If sea urchin eggs are separated meridi- onally in advanced blastula stages, and the separation surfaces are vitally stained, some pairs of larvae are obtained in which one twin is stained dorsally, the other ventrally (Horstadius and Wolsky, ’36). This would be the expected result if dorsoventrality were already determined at the time of the operation and if the cut happened by chance to pass through the frontal plane. By con- trast, if the first two blastomeres are sepa- rated, followed by staining of the surface of separation, pairs of larvae are found in which both are stained dorsally. One larva in such a pair tends to develop more slowly and to form less perfect ventral structures than the other. Presumably the latter repre- sents the prospective ventral half of the egg, the former the prospective dorsal half, and the first cleavage plane happened to pass through the prospective frontal plane. If so, the original dorsoventral axis has been re- tained in the prospective ventral half, re- versed in the prospective dorsal half. Since such a reversal of dorsoventrality is possible in early stages, and not in later ones, estab- lishment of a rigid dorsoventral organization must be a progressive process. Similar re- sults have been obtained by Gustafson and Savhagen (’50) using anionic detergents (see SELECTED INVERTEBRATES above). Up to 6 hours after fertilization such detergents prevent development of the oral (ventral) side and radial symmetry results; but between 6 and 9 hours after fertilization the detergents gradually lose their effects and after 9 hours they can no longer inhibit development of bilaterality. REFERENCES Berg, W. E., and Kutsky, P. B. 1951 Physiolog- ical studies of differentiation in Mytilus edulis. I. The oxygen uptake of isolated blastomeres and polar lobes. Biol. Bull., 107:47-61. Berrill, N. J. 1929 Studies in tunicate develop- ment. I. General physiology of development of simple ascidians. Phil. Trans. Roy. Soc. London, B, 218:37-78. 1932 The mosaic development of the as- cidian egg. Biol. Bull., 63:381-386. Boveri, T. 1901 Die Polaritat von Ovocyte, Ei und Larve des Strongylocentrotus lividus. Zool. Jahrb. Abt. Anat. u. Ontogenie, 74:630-653. Brachet, J. 1937 la différenciation sans clivage dans l’oeuf de Chétoptére envisagée aux points de vue cytologique et métabolique. Arch. Biol., 48: 561-589. 1950 Chemical Embryology. ence Publishers, Inc., New York. Child, C. M. 1940 Lithium and echinoderm ex- ogastrulation: with a review of the physiological- gradient concept. Physiol. Zool., 73:4—42. 1941 Patterns and Problems of Develop- ment. University of Chicago Press, Chicago. Clement, A. C. 1938 The structure and develop- ment of centrifuged eggs and egg fragments of Physa heterostropha. J. Exp. Zool., 79:435- 460. Cohen, A., and Berrill, N. J. 1936 The develop- ment of isolated blastomeres of the ascidian egg. J. Exp. Zool., 74:91-117. Conklin, E.G. 1905a The organization and cell- lineage of the ascidian egg. J. Acad. Nat. Sci. Philadelphia, 73:1-119. 1905b Organ-forming substances in the eggs of ascidians. Biol. Bull., 8:205—230. 1905c Mosaic development in ascidian eggs. J. Exp. Zool., 2:145-223. 1906 Does half of an ascidian egg give rise to a whole larva? Roux’ Arch. Entw.-mech., 21:727-753. 1910 The effects of centrifugal force upon the organization and development of the eggs of fresh water pulmonates. J. Exp. Zool., 9: 417-454. 1911 The organization of the egg and the development of single blastomeres of Phallusia mamillata. J. Exp. Zool., 10:393-407. 1924 Cellular differentiation; Section IX in General Cytology, edited by E. V. Cowdry. University of Chicago Press, Chicago. 1931 The development of centrifuged eggs of ascidians. J. Exp. Zool., 60:1-119. Costello, D. P. 1945 Experimental studies of germinal localization in Nereis. I. The develop- Intersci- 433 ment of isolated blastomeres. J. Exp. Zool., 700: 19-66. Dalcq, A. M. 1938 Form and Causality in Early Development. Cambridge University Press, Cam- bridge, England. Driesch, H. 1891 Entwicklungsmechanische Stu- dien. I. Der Werth der beiden ersten Furchungs- zellen in der Echinodermenentwicklung. Experimentelle Erzeugung von Theil- und Dop- pelbildungen. II. Uber die Beziehungen des Lichtes zur ersten Etappe der thierischen Form- bildung. Zeit. f. wiss. Zool., 53:160-184. 1892 Entwicklungsmechanische Studien. III. Die Verminderung des Furchungsmaterials und ihre Folgen (Weiteres iiber Theilbildungen). IV. Experimentelle Veranderungen des Types der Furchung und ihre Folgen (Wirkungen yon Warmezufuhr und von Druck). V. Von der Fur- chung doppelt befruchteter Eier. VI. Uber einige allgemeine Fragen der theoretischen Morpholo- gie. Zeit. f. wiss. Zool., 55:1-62. Driesch, H., and Morgan, T.H. 1895a Zur Anal- ysis der ersten Entwickelungsstadien des Cteno- phoreneies. I. Von der Entwickelung einzelner Ctenophorenblastomeren. Roux’ Arch. Entw.- mech., 2:204-215. , and Morgan, T. H. 1895b Zur Analysis der ersten Entwickelungsstadien des Ctenopho- reneies. II. Von der Entwickelung ungefurchter Eier mit Protoplasmadefekten. Roux’ Arch. Entw.-mech., 2:216-224. Fischel, A. 1897 Experimentelle Untersuchun- gen am Ctenophorenei. I. Von der Entwickelung isolirter Eitheile. Roux’ Arch. Entw.-mech., 6: 109-130. 1898 Experimentelle Untersuchungen am Ctenophorenei. II. Von der kiinstlichen Er- zeugung (halber) Doppel- und Missbildungen. III. Uber Regulationen der Entwickelung. IV. Uber den Entwickelungsgang und die Organisa- tionsstufe des Ctenophoreneies. Roux’ Arch. Entw.-mech., 7:557-630. Foerster, M., and Orstrém, A. 1933 Observations sur la prédétermination de la partie ventrale dans Voeuf d’oursin. Trav. de la Stat. Biol., Roscoff, 77: 63-83. Gustafson, T., and Hjelte, M.-B. 1951 The amino acid metabolism of the developing sea urchin egg. Exptl. Cell Res., 2:474490. , and Lenicque, P. 1952 Studies on mito- chondria in the developing sea urchin egg. Exptl. Cell Res., 3:251-274. , and Savhagen, R. 1950 Studies on the determination of the oral side of the sea-urchin egg. I. Ark. for Zool., 42A (10) :1-6. Harrison, R. G. 1945 Relations of symmetry in the developing embryo. Trans. Acad. Arts Sci. Connecticut, 36:277-330. Hatt, P. 1931 la fusion expérimentale d’oeufs de “Sabellaria alveolata 1.” et leur développe- ment. Arch. Biol., 42:303-323. 1932 Essais expérimentaux sur les local- isations germinales dans l’oeuf d’un annélide (Sabellaria alveolata V..). Arch. d’Anat. micros., 28:81-98. Herbst, C. 1892 Experimentelle Untersuchun- 334 EMBRYOGENESIS: PROGRESSIVE DIFFERENTIATION gen uber den Einfluss der veranderten chemischen Zusammensetzung des umgebenden Mediums auf die Entwicklung der Thiere. I. Versuche an Seeigeleiern. Zeit. f. wiss. Zool., 55:446-518. Herbst, C. 1893 Experimentelle Untersuchungen uber den Einfluss der veranderten chemischen Zusammensetzung des umgebenden Mediums auf die Entwicklung der Thiere. II. Weiteres iiber die morphologische Wirkung der Lithiumsalze und ihre theoretische Bedeutung. Mittheil. aus der Zool. Stat. zu Neapel, 77:136-220. 1896 Experimentelle Untersuchungen liber den Einfluss der veranderten chemischen Zusammensetzung des umgebenden Mediums auf die Entwicklung der Thiere. III. Uber das In- eiandergreifen von normaler Gastrulation und Lithiumentwicklung (Ziichtung von Lithium- larven mit Entoderm und Mund). IV. Die forma- tive Wirkung des Lithiums auf befruchtete Eier von Asterias glacialis. V. Uber die Unterdriickung von Entwicklungsprozessen (Wirkung von Kalium rhodanatum und Natrium butyricum). VI. Uber den Einfluss einiger anderer organischer Salze. Roux’ Arch. Entw.-mech., 2:455-516. Horstadius, S. 1928 Uber die Determination des Keimes bei Echinodermen. Acta Zoologica, 9:1- 191. 1935 Uber die Determination im Ver- laufe der Eiachse bei Seeigeln. Pubbl. Staz. Zool. Napoli, 74:251-429. 1937a Experiments on determination in the early development of Cerebratulus lacteus. Biol. Bull., 73:317-342. 1937b Investigations as to the localiza- tion of the micromere-, the skeleton-, and the entoderm-forming material in the unfertilized egg of Arbacia punctulata. Biol. Bull., 73:295- 316. 1938 Schniirungsversuche an _ Seeigel- keimen. Roux’ Arch. Entw-mech., 738:197-248. 1939 The mechanics of sea-urchin devel- opment, studied by operative methods. Biol. Rev., 14:134-179. 1949 Experimental researches on the developmental physiology of the sea-urchin. Pubbl. Staz. Zool. Napoli, 27 (Suppl.):131-172. 1950 Transplantation experiments to elucidate interactions and regulations within the gradient system of the developing sea urchin egg. J. Exp. Zool., 713:245-276. , and Gustafson, T. 1948 On the develop- mental physiology of the sea urchin. Symposium Soc. Exp. Biol., 2:50-56. , Lorch, I. J., and Danielli, J. F. 1950 Dif- ferentiation of the sea urchin egg following re- duction of the interior cytoplasm in relation to the cortex. Exptl. Cell Res., 7:188-193. , Lorch, I. J., and Danielli, J. F. 1953 The effect of enucleation on the development of sea urchin eggs. II. Enucleation of animal or vegetal halves. Exptl. Cell Res., 4:263-274. , and Wolsky, A. 1936 Studien iiber die Determination der Bilateralsymmetrie des jun- gen Seeigelkeimes. Roux’ Arch. Entw.-mech., 135:69-113. Holter, H. 1949 Problems of enzyme localization in development. Pubbl. Staz. Zool. Napoli, 2/ (Suppl. ) :60-76. Howard, E. 1932 The structure of protoplasm as indicated by a study of sea-urchin eggs at various shearing forces. J. Cell. Comp. Physiol., 7:355- 369. Just, E.E. 1912 The relation of the first cleavage plane to the entrance point of the sperm. Biol. Bull., 22:239-252. Lehmann, F. E. 1945 Einfiihrung in die physi- ologische Embryologie. Verlag Birkhauser, Basel. Leopoldseder, F. 1931 Entwicklung des Eies von Clepsine nach Entfernung des vegetativen Pol- plasmas. Zeit. f. wiss. Zool., 739:201-248. Lillie, F. R. 1902 Differentiation without cleav- age in the egg of the annelid Chaetopterus perga- mentaceus. Roux’ Arch. Entw.-mech., 14:477-499. 1906 Observations and experiments con- cerning the elementary phenomena of embryonic development in Chaetopterus. J. Exp. Zool., 3: 153-268. 1909 Polarity and bilaterality of the an- nelid egg. Experiments with centrifugal force. Biol. Bull., 76:54~79. Lindahl, P. E. 1932a Zur _ experimentellen Analyse der Determination der Dorsoventralachse beim Seeigelkeim. I. Versuche mit gestreckten Eiern. Roux’ Arch. Entw.-mech., 727:300-322. 1932b Zur experimentellen Analyse der Determination der Dorsoventralachse beim Seei- gelkeim. II. Versuche mit zentrifugierten Eiern. Roux’ Arch. Entw.-mech., 127:323-338. 1936 Zur Kenntnis der physiologischen Grundlagen der Determination im Seeigelkeim. Acta Zool., 17:179-365. 1942 Contributions to the physiology of form generation in the development of the sea urchin. Quart. Rev. Biol., 17:213-227. Morgan, T. H. 1927 Experimental Embryology. Columbia University Press, New York. , and Tyler, A. 1930 The point of en- trance of the spermatozodn in relation to the orientation of the embryo in eggs with spiral cleavage. Biol. Bull., 58:59-73. Mori, Y. 1932 Entwicklung isolierter Blasto- meren und teilweise abgetoteter alterer Keime von Clepsine sexoculata. Zeit. £. wiss. Zool., 141: 399-431. Motomura, I. 1935 Determination of the em- bryonic axis in the eggs of amphibia and echino- derms. Sci. Rep. Tohoku Univ., 70:211-245. Needham, J. 1942 Biochemistry and Morpho- genesis. Cambridge University Press, Cambridge, England. Novikoff, A. B. 1938 Embryonic determination in the annelid, Sabellaria vulgaris. II. Trans- plantation of polar lobes and blastomeres as a test of their inducing capacities. Biol. Bull., 74:211- 234. 1940 Morphogenetic substances or or- ganizers in annelid development. J. Exp. Zool., 85:127-155. Pease, D. C. 1938 The influence of centrifugal force on the bilateral determination of the spi- rally-cleaving eggs of Urechis. Biol. Bull., 75:409- 424. SELECTED INVERTEBRATES Pease, D. C. 1939 An analysis of the factors of bilateral determination in centrifuged echino- derm embryos. J. Exp. Zool., 80:225-247. 1941 Echinoderm bilateral determination in chemical concentration gradients. I. The effects of cyanide, ferricyanide, picrate, dinitrophenol, urethane, iodine, malonate, etc. J. Exp. Zool., 86: 381-404. 1942a Echinoderm bilateral determina- tion in chemical concentration gradients. II. The effects of azide, pilocarpime, pyocyanine, dia- mine, cysteine, glutathione, and lithium. J. Exp. Zool., 89:329-345. 1942b Echinoderm bilateral determina- tion in chemical concentration gradients. III. The effects of carbon monoxide and other gases. J. Exp. Zool., 89:347-356. Penners, A. 1922 Die Furchung von Tubifex rivulorum Lam. Zool. Jahrb. Abt. f. Anat. u. Ontog., 43:323-368. 1924a Doppelbildungen bei Tubifex riv- ulorum Lam. Zool. Jahrb. allg. Zool., 47:91-120. 1924b Experimentelle Untersuchungen zum Determinationsproblem am Keim von Tubi- fex rivulorum Lam. I. Die Duplicitas cruciata und organbildende Keimbezirke. Arch. f. mikr. Ana- tomie, 702:51-100. 1926 Experimentelle Untersuchungen zum Determinationsproblem am Keim von Tubi- fex rivulorum Lam. II. Die Entwicklung teilweise abgetoteter Keime. Zeit. f. wiss. Zool., 727:1-140. 1937 Regulation am Keim von Tubifex rivulorum Lam. nach Ausschaltung des ektoder- malen Keimstreifs. Zeit. f. wiss. Zool., 149:86— 130. 1938 Abhangigkeit der Formbildung vom Mesoderm im Tubifex-Embryo. Zeit. f. wiss. Zool., 150:305-357. Pisano, A. 1949 Lo sviluppo dei primi due blas- tomeri separati dell’uovo di Ascidie. Pubbl. Staz. Zool. Napoli, 22:16-25. Raven, C. P., and Bretschneider, L. H. 1942 The effect of centrifugal force upon the eggs of Lim- naea stagnalis 1.. Arch. Néerlandaises de Zool., 6:255-278. Reverberi, G., and Minganti, A. 1946a Le po- tenze dei quartetti animale e vegetativo isolati di Ascidiella aspersa. Pubbl. Staz. Zool. Napoli, 20:135-151. , and Minganti, A. 1946b Fenomeni di evocazione nello sviluppo dell’uovo di Ascidie. Risultati dell’indagine sperimentale sull’uovo di Ascidiella aspersa e di Ascidia malaca allo stadio di otto blastomeri. Pubbl. Staz. Zool. Napoli, 20: 199-252. , and Minganti, A. 1947 La distribuzione delle potenze nel germe di Ascidie allo stadio di otto blastomeri, analizzata mediante le combina- zioni e i trapianti di blastomeri. Pubbl. Staz. Zool. Napoli, 27:1-35. , and Minganti, A. 1951 Concerning the interpretation of the experimental analysis of the ascidian development. Acta Biotheoretica, 9:197- 204. Rose, S. M. 1939 Embryonic induction in the ascidia. Biol. Bull., 77:216-232. 538). Rulon, O. 1941 Modifications of development in the sand dollar by NaSCN and Ca-free sea water. Physiol. Zool., 74:305-315. 1952 The modification of developmental patterns in the sand dollar by glucose. Physiol. Zool., 25:346-357. Runnstrém, J. 1926 Experimentelle Bestimmung der Dorso-Ventralachse bei dem Seeigelkeim. Ark. for Zool., 718A (4) :1-6. 1928a Zur experimentellen Analyse der Wirkung des Lithiums auf den Seeigelkeim. Acta Zool., 9:365—424. 1928b Plasmabau und Determination bei dem Ei von Paracentrotus lividus Lk. Roux’ Arch. Entw.-mech., 773:556-581. 1933 Kurze Mitteilung zur Physiologie der Determination des Seeigelkeims. Roux’ Arch. Entw.-mech., 729:442-444, , and Runnstrém, S. 1921 Uber die Ent- wicklung von Cucuraria frondosa Gunnerus and Psolus phantopus Strussenfelt. Bergens Mus. Aar- bok, 1918-1919, 2(No. 5) :1-100. Schleip, W. 1914 Die Entwicklung zentrifugier- ter Eier von Clepsine sexoculata. Verh. deutsch. Zool. Ges., pp. 236-253. 1929 Die Determination der Primitivent- wicklung. Akademische Verlags-Austalt, Leip- zig. Spek, J. 1926 Uber gesetzmassige Substanzver- teilungen bei der Furchung des Ctenophoreneies und ihre Beziehungen zu den Determinations- problemen. Roux’ Arch. Entw.-mech., 107:54-73. Sturtevant, A. H. 1923 Inheritance of direction of coiling in Limnea. Science, 58:269-270. Taylor, C. V., and Tennent, D. H. 1924 Pre- liminary report on the development of egg frag- ments. Yearbook Carnegie Institution, 23:201- 206. Tennent, D. H., Taylor, C. V., and Whitaker, D. M. 1929 An investigation on organization in a sea- urchin egg. Pap. Tortugas Lab., 26:1-104. Titlebaum, A. 1928 Artificial production of Janus embryos of Chaetopterus. Proc. Nat. Acad. Sci., 14:245-247. ; Tung, T. 1932 Experiences de coloration vitale sur loeuf d’Ascidiella aspersa. Arch. de Biol., 43: 451-469. 1934 Récherches sur les potentialités des blastoméres chez Ascidiella scabra. Experiences de translocation, de combinaison et d’isolement de blastomeres. Arch. d’Anat. micros., 30:381- 410. , Ku, S., and Tung, Y. 1941 The develop- ment of the ascidian egg centrifuged before fertil- ization. Biol. Bull., 80:153-168. Tyler, A. 1930 Experimental production of double embryos in annelids and molluscs. J. Exp. Zool., 57:347-407. Ubisch, L. von 1929 Uber die Determination der larvalen Organe und der Imaginalanlage bei Seeigeln. Roux’ Arch. Entw.-mech., 777:80-120. 1938 Uber Keimverschmelzungen an Ascidiella aspersa. Roux’ Arch. Entw.-mech., 138: 18-36. 1939a Uber die Entwicklung von Ascid- ienlarven nach friihzeitiger Entfernung der ein- 336 EMBRYOGENESIS: PROGRESSIVE DIFFERENTIATION zelnen organbildenden Keimbezirke. Roux’ Arch. Entw.-mech., 739:438-492. Ubisch, L. von 1939b Keimblattchimarenfor- schung an Seeigellarven. Biol. Rev., 74:88-103. 1940 Weitere Untersuchungen iiber Reg- ulation und Determination im Ascidienkeim. Roux’ Arch. Entw.-mech., 740:1-24. Weiss, Paul 1949a Differential growth; in Chemistry and Physiology of Growth, edited by A. K. Parpart, pp. 35-186. Princeton University Press, Princeton, New Jersey. 1949b Growth and differentiation on the cellular and molecular levels. Exptl. Cell Res., 1(Suppl.) :475-482. 1950 Perspectives in the field of morpho- genesis. Quart. Rev. Biol., 25:172-198. Wilson, E. B. 1899 Cell-lineage and ancestral reminiscence. Biol. Lect. Marine Biol. Lab., Woods Hole, 6:21-42. 1903 Experiments on cleavage and local- ization in the nemertine egg. Roux’ Arch. Entw.- mech., 76:411—460. 1904a Experimental studies on germinal localization. I. The germ-regions in the egg of Dentalium. J. Exp. Zool., 7:1-72. 1904b Experimental studies in germinal localization. II. Experiments on the cleavage mosaic in Patella and Dentalium. J. Exp. Zool., 1:197-268. Yatsu, N. 1904 Experiments on the development of egg fragments in Cerebratulus. Biol. Bull., 6: 123-136. 1910 Experiments on germinal localiza- tion in the egg of Cerebratulus. J. Coll. Sci. Tok- oyo, 27 (17) :1-37. 1912 Observations and experiments on the ctenophore egg. III. Experiments on germinal localization of the egg of Beroé ovata. Annot. Zool. Jap., 8:5-13. Ziegler, H. E. 1898 Experimentelle Studien iiber die Zelltheilung. III. Die Furchungszellen von Beroé ovata. Roux’ Arch. Entw.-mech., 7:34— 64. Section VI CHAPTER 4 Insects DIETRICH BODENSTEIN ProcressivE differentiation is the process by which the fertilized insect egg is transformed into the complete embryo through develop- mental events closely coordinated in time and space. The initiation and progress of organization in the egg depend upon the realization of certain dynamic phenomena which mold the embryonic material for par- ticular functions. It is the special physico- chemical nature of the insect egg that pro- vides the substrate and conditions for the action and interaction of these dynamic forces. THE INITIATION OF EARLY EMBRYONIC ORGANIZATION The insect egg is usually rich in yolk. Its nucleus lies in a central position and is embedded in a small cytoplasmic island. Fine cytoplasmic strands of this island ram- ify through the yolk and often condense at the periphery of the egg, forming here a cortical layer. After the nucleus has di- vided and its daughter nuclei have populated the yolk, the majority of them move toward the egg surface. Here they arrange them- selves with the formation of cell boundaries into a single cell layer, the blastoderm. The first visible differentiation of the embryo is the germ band. It appears in the blastoderm in the region of the presumptive prothorax and from here continues its differentiation anteriorly and posteriorly. These beginnings of organization in the egg are governed by two different centers, the activation center and the differentiation center (Seidel, ’29). Through an alternation of dynamic processes and material reactions, these two centers interact, thus setting into motion the whole process of embryonic organization. The activation center is located at the pos- terior pole of the egg. Its function depends on the interaction of the cleavage nuclei with some factor in the region of the center. The product of this reaction, presumablv 337) a specific material substance, spreads for- ward in the egg, evoking in its course an- other reaction which changes the structure of the yolk system. This in turn causes a contraction of the yolk system. The latter reaction provides the necessary situation for the aggregation of the blastoderm cells to form the germ band. The contraction begins at the site of the presumptive prothorax and spreads from here anteriorly and posteriorly in a wavelike fash- ion. This region is known as the differentia- tion center; in this same region the first visible differentiation of the germ band may be witnessed. The morphologically defined differentiation center hence can be visualized as a center of morphodynamic movement which provides the stimulus for the aggrega- tion of the cells that form the germ band. The chain of reactions evoked by the activa- tion center is thus essential for the function of the differentiation center, which repre- sents the focal point for all subsequent proc- esses of differentiation of the embryo. While the described phenomena do not directly concern the formation of the blasto- derm, it must be realized that the entire specialized organization of the insect egg as a dynamic system, including the yolk sys- tem, the cleavage nuclei with their cyto- plasmic connections, and the blastoderm, is the prerequisite for the normal sequence of the described reactions. DETERMINATION AND REGULATION The problem of regulation and determina- tion is closely related to, and can only be understood in the light of, the above dis- cussed general principles of insect develop- ment. It has been shown that the regulative capacity of the egg varies greatly within the different groups of insects (Seidel, Bock, and Krause, ’40). As a matter of fact, all transi- tions from eggs with great regulative powers to eggs exhibiting strictly mosaic develop- 338 EMBRYOGENESIS: PROGRESSIVE DIFFERENTIATION ment may be found. For instance, in the dragonfly Platycnemis, twin embryos of har- monious proportions can be experimentally produced by injuring of the presumptive germ band in early cleavage stages. In this form, the embryonic material maintains its regulative capacity until some time after the blastoderm is formed. In Diptera, on the other hand, the determination of the pre- sumptive embryonic parts is already ac- complished at the time of fertilization, for when egg parts at this early stage are re- moved, the remaining parts are unable to develop beyond their prospective significance (Reith, °25; Pauli, ’27). An intermediate position between these two extremes is taken by the honeybee (Schnetter, ’34) or the pea weevil Sitonia (Reith, 35); in these the determination of the embryo is completed at the early blastoderm stage. The regula- tive capacity of the embryonic material is gradually restricted and finally lost as de- velopment proceeds. In Platycnemis, where in early development the various regions of the presumptive embryo can still replace each other, these regions become more and more committed to special tasks with con- tinued development. Finally the entire em- bryonic zone reaches the mosaic stage, in which each part is irrevocably determined and destined to a specific end. This state of affairs prevails not only in the determination of the early embryonic material but also in the determination of parts within organ fields of larval or imaginal character, and extends even to the processes of regenera- tion: it is a general principle and well founded indeed. Thus the principle of progressive deter- mination applies to insects as well as to other forms. Is the difference between regu- lative and mosaic insect development to be explained by this principle? If, as we usually assume, it is the early or late determination of embryonic parts for their final destiny that distinguishes the regulative from the mosaic type, then this principle does not ap- ply. For eggs exhibiting mosaic characteris- tics in early stages of development can in the course of their development acquire the capacity to regulate (Ewest, ’37). The differ- ences between the determinate and indeter- minate egg type are not based on alterations of events in time, but are found in the spe- cialized architecture of the different insect eggs. Specializations in the configuration of the internal egg structure which restrict or enhance the freedom of movement of the embryonic material in the dynamic egg system are the factors that decide and dis- tinguish the determinate from the inde- terminate type (Seidel, Bock, and Krause, °40). EMBRYONIC INDUCTION BETWEEN GERM LAYERS Among the causal factors of embryonic determination, we must include the processes of embryonic induction. The experiments supplying this important information for insects were carried out on the eggs of the neuropteron Chrysopa perla (Bock, ’39). The decisive facts are briefly as follows: When at an early stage of germ layer formation ectoderm is removed at one side of the seg- ment, the mesoderm spreads normally be- low the ectoderm-free regions, but fails to form an epithelium and soon degenerates. Thus the epithelization of this layer is de- pendent on the presence of ectoderm. The question whether the ectoderm also decides the further differentiation of the underlying mesoderm was determined by other experi- ments. Within one half of a segment large parts of ectoderm located near the median line or more lateral to it were removed. The mesoderm below the median as well as the lateral remains of ectoderm formed coelom- epithelium of diminished proportions, corre- sponding in size to the reduced ectoderm portions above. The epithelium differenti- ated in due course into intestinal muscles and cardioblasts below the lateral ecto- derm, while these differentiations were ab- sent below the median ectodermal region. In other words, the mesoderm always differ- entiated according to position, regardless of its prospective significance. From these re- sults and from other experiments showing that the mesoderm was isopotent for the structures in question, it was concluded that the various ectodermal regions determine the type of differentiation of the mesoderm be- low them. The lateral ectoderm regions hence contain factors necessary for the formation of intestinal muscles and cardioblasts, which factors were absent in the median ectodermal portions. Thus the capacity of the ecto- derm to imprint its specific demands on the underlying mesodermal substratum must be considered as a phenomenon of real em- bryonic induction. In the light of these facts, it is interesting to venture a comparison of the types of organizing events between am- phibians and insects. The decisive factors for the organization of the embryo are lo- cated in a definite germ layer in amphibians INSECTS as well as in Chrysopa. In amphibians they are found in the mesoderm, in Chrysopa in the ectoderm. The inducing material of amphibians enjoys a great freedom of move- ment. The extent of these movements deter- mines to a large degree the size and propor- tions of the induced structures. In Chrysopa, on the other hand, the inductor, the ecto- derm, is a relatively rigid system, endowed with a mosaic of inductive potency. The mesoderm underlies this system and _ re- sponds with specific differentiations to the stimuli emanating from it. The movements of the embryonic material, different as they are in both classes of animals compared, undoubtedly have the same basic signifi- cance, namely the bringing together of de- velopmental systems for the purpose of interactions that provide the basis for the organization of the embryo (Seidel, Bock, and Krause, *40). CYTOPLASMIC DETERMINERS In the determinate egg type, the cortical cytoplasm is usually heavy and already in the fertilized egg seems to consist of a mosaic of differentials that determine the various parts of the future embryo. When in the beetles Leptinotarsa (Hegner, 7°11) or Bruchus (Brauer and Taylor, ’34) small areas of the cortical cytoplasm are destroyed before the cleavage nuclei have arrived and without apparent injury to them, those parts of the embryo are missing which would have developed from the destroyed regions. Since no degeneration of the cleavage nuclei occurred, it follows that they must have taken part in the development of some other parts of the egg. The cortical cytoplasm they happen to invade as the result of the experimentally altered conditions must have determined their fate. The fact that these nuclei were able to respond to the cyto- plasmic influences of regions normally for- eign to them indicates their totipotency. The totipotency of cleavage nuclei in early stages of development, it might here be added, was definitely proved in experiments on other forms, notably Platycnemis, where it persists at least until the seventh cleav- age division (Seidel, ’°32). These experiments show quite conclusively that the cortical cytoplasm is a differentiated continuum in which localized differentials exert specific influences on the cleavage nuclei, leading them towards special assignments. One has to assume that the cytoplasmic regions pos- sess their specific qualities only when in 3) normal topographic relationship to the corti- cal cytoplasmic layer as a whole. Their influences must be regarded as of a general directive nature in that they set up differ- entials in the cleavage nuclei, thereby creat- ing a definite pattern within the framework of the blastoderm, which forms the basis for the ensuing developmental events. If one could excise the cytoplasm of the presump- tive eye region and supply it with any number of the totipotent nuclei, the isolated bit of tissue would in all probability never give rise to an eye or to any specialized structure. Within the realm of cytoplasmic deter- miners one has to include the pole-plasm. The cortical cytoplasm at the posterior pole of the egg is in certain insects distinguished from the rest of this layer by the presence of deeply staining granules. This region is called the pole-plasm. The cleavage nuclei which penetrate this region become known as pole cells; they represent the germ-cell primordia. When in the egg of the beetle Leptinotarsa (Hegner, ’11), the polar cyto- plasm is removed before the cleavage nuclei have entered the pole, germ cells are lack- ing in the embryo. This shows that the polar cytoplasm contains some factor essential for the formation of the germ cells. The polar region of the cortical cytoplasm is thus en- dowed with a specific organ-forming prin- ciple which takes effect at an early stage in development and which determines the cleavage nuclei towards their future des- tiny. It has been assumed that the formative role of the polar influences was such as to predetermine rigidly the polar cells for their fate. That this is not the case has recently been demonstrated in Drosophila (Poulson, *47). For some time it has been known that not all the pole cells take part in the formation of the gonad, but that some of them go astray on their way to their final location in the interior of the embryo. These “lost” cells were formerly supposed to degenerate, apparently because of their failure to become germ cells. Now Poulson has made the striking discovery that these “lost” cells do not degenerate at all, but that they can become incorporated into the epithelium of the gut, of which they actually become a part. This fact deserves emphasis, for it demonstrates that the cyto- plasmic pole factors do not decide finally the ultimate fate of the pole cells. They obviously endow the pole cells with the potentialities necessary for the formation of germ cells. But whether or not these or other 340 EMBRYOGENESIS: PROGRESSIVE DIFFERENTIATION potentialities become realized is decided much later in the course of development, apparently at or near the site these cells finally occupy. PROGRESSIVE FIXATION OF CHARACTERS The entire process of embryonic deter- mination, it will be noted, is a step by step affair. First, a general body zone is mapped out (differentiation center). This zone sub- divides into smaller more specialized zones, with less power of change (germ layers). Progressive specialization continues, bring- ing about a more detailed subdivision of the various embryonic zones into organ fields. Segregation of material, and with it re- striction of potencies, is now so advanced that the different organ fields can no longer replace each other. Yet the capacity for limited regulation might still prevail in certain organ fields even at relatively late stages of development (Liischer, ’44). Soon the power of regulation subsides even in these fields; finally, at the time of visible organ segregation all power of regulation is apparently lost. The organism developing in the egg is the larval form of the species. It is the de- termination of its parts that has been dis- cussed so far. When the fully developed larva leaves the egg, it carries in its body (holometabolous insects) reservoirs of cell groups from which many structures of the adult animal take their origin. These groups of cells are called imaginal discs; they repre- sent the primitive anlagen of the future imaginal organs, such as legs, wings, eyes, etc. Not all these discs are morphologically distinct at the time the larva hatches from the egg. The time of their appearance varies greatly within different insect groups and even within the various organs of the same individual. In Lepidoptera, for instance, the imaginal wing discs are already visible in the youngest caterpillar stage, while the leg discs make their appearance only in the last larval instar (Bodenstein, 36). In Dip- tera, both leg and wing discs are visible as small cell aggregations at the time the larva hatches (Auerbach, 736). The regions in the larval body from which these various imaginal discs arise are mapped out lone before the discs become actually visible (Bodenstein, ’36). Now the evidence sug- gests that the determination of these imaginal regions already occurs in embryonic de- velopment, and that it takes place a short time after the larval structures in the egg are determined (Liischer, 44; Geigy, ’31). The insect egg, as it has been said, thus becomes the carrier of a double embryo— a larval and an imaginal embyro (Geigy, ’°31). The imaginal anlagen or their pre- sumptive regions may remain in an un- differentiated embryonic state till very late in postembryonic development and maintain throughout this period a remarkable degree of regulative capacities. The different ele- ments of the imaginal organs become deter- mined at different times. The character of the scales in adult Lepidoptera legs or the type of cuticle sheath formed by the legs at pupation is already determined in the third instar larva while at this time the presumptive materials for femur, tibia, or tarsus can still replace each other. The determination of these latter parts appar- ently takes place shortly before the leg discs become visible at the last larval instar (Bodenstein, ’37). The appearance of the imaginal discs per se does not mark the final state of determination, for in Droso- phila, where the leg discs are present in the first instar larva, isolation experiments showed that the discs of young last instar larvae are still capable of regulation (Vogt, 46), reaching their final mosaic state in a later period of this instar (Bodenstein, °41). The genital discs of last instar Droso- phila larvae also still possess considerable powers of regulation (Hadorn, Bertani, and Gallera, ’49). At this advanced stage these discs are composed of a mosaic of separate fields, each of which gives rise to a special element of the imaginal genital apparatus These fields cannot replace each other, but the various parts within a single field can regulate; they differentiate into normally shaped adult structures of characteristic size Duplications of various other Drosophila or- gans. such as wing, antenna, palpus, or scutellum, following x-ray treatment (Wad- dington, ’42) or temperature treatment (Vogt. 47a) of young last instar larvae gave further proof of the indeterminate state of these or- gans in late stages of development. The determination of all imaginal systems is not necessarily completed at the end of the larval period. The venation in wings of some Lepidoptera is not finally determined before the pupa is several days old (Henke, 33). It is also during the pupal stage that the color pattern of the wing becomes deter- mined. This wing pattern is a complicated developmental system, composed of different zones which are determined at different times. In the determination of the wing INSECTS pattern a succession of determination streams passes over the wing surface determining the various component parts of this system. The nature of the determining influences is not known. The experimental analysis of wing pattern formation has furnished a wealth of information concerning the state of determination of the various zones at dif- ferent times in development and has brought to light the interdependency of these zones in the formation of the entire pattern (Cas- pari, °41). DEVELOPMENTAL INTERACTIONS IN LATER STAGES Although the different elements comprising the component parts of an organ can still replace each other in organ anlagen of postembryonic stages, the imaginal discs or the prospective regions from which they arise are at these late stages usually com- mitted to the type of organ or body part they eventually form. There are, however, experiments indicating that the plasticity ot the formative disc material goes beyond the limit of regulations of its own parts. It has been shown that the developmental path of one type of disc can be changed into an- other. In late postembryonic stages of de- velopment, therefore, the prospective signifi- cance of these discs is not yet organ specific. For example, the presumptive antenna mate- rial of last instar Drosophila larvae can be changed into leg material by appropriate temperature treatment (Vogt, ’46b) or x-ray treatment (Waddington, °*42). Also in Drosophila the change of presumptive head chitin into eye facets (Steinberg, °41; Chevais, ’44) and wing tissue into body skin (Waddington, 42) by similar methods are other instances of the same sort. Develop- mental modifications produced by treating Drosophila larvae or pupae with ether (Gloor, °46, °47) reveal effects with lke implications, and so do many other inves- tigations dealing with phenocopies. It is of interest to note further that transforma- tions of presumptive antenna material into leg material were obtained in Drosophila by the use of two chemicals, colchicin (Vogt, ’'47b) and a nitrogen mustard compound (Bodenstein, ’47)—chemicals that are known to have a marked effect on cell division. The long-suspected importance of growth rela- tions (Goldschmidt, ’38) between the various organ-fields as decisive factors in determin- ing the type of organ differentiation has thus gained support. The phenomenon 341 known as homoeosis, i.e., the replacement by regeneration of one organ by one belong- ing to another region of the body, needs to be mentioned in this connection, In studying this type of regeneration, it was found that regional specific regeneration occurs if relatively little of the original or- gan is removed, while amputation of larger parts is usually followed by heteromorphic regeneration (Brecher, ’24). It thus seems that in the first instance the organ-field of the original organ takes the lead in determin- ing the new structure, while in the second instance this lead is taken by a neighboring field. In ontogenesis as in regeneration, there- fore, it is the interaction between zones within one organ-field, or that between neighboring fields, which determines the type of organ finally to be formed. An interesting and special case showing a change from one type of development into another has been reported in the moth Orgyia (Lep.). Here, the secondary sexual characters of the female wing discs are still not irrevocably determined in fourth instar caterpillars, for fourth imstar wing discs when transplanted into male hosts and made to regenerate in their new surroundings will form male wings. In the reverse experiments, however, transplanted male wings will re- generate male wings in the female environ- ment (Paul, ’37). It might well be pointed out that this case is a singular instance of this kind, although the problem has been studied extensively in other insects. The experiments usually quoted as proof for the complete determination of secondary sexual characters are castration experiments and transplantations of gonads between indi- viduals of the opposite sex. It may be added that the above cited case is not only singular in its results but also in its method. In Orgyia, in contrast to all other investigations, the organ exhibiting the secondary sexual character was brought into contact with the entire developing system of the oppo- site sex and hence subjected to influences additional to those of the gonads. It is per- haps the lack of appropriate experimentation rather than the special condition of the Orgyia wing that made the obtained results exceptional. Finally, mention should be made of devel- opmental interactions between systems which already possess a considerable degree of dif- ferentiation. There are species of Drosophila that have spiral, and others that have non- spiral, ellipsoid adult testes. Now it has been found (Stern, "41a, ’41b) that when vasa 342 EMBRYOGENESIS: PROGRESSIVE DIFFERENTIATION efferentia from species with spiral testes are brought experimentally into contact with normally non-spiral testes, the latter coil, i.e, become spiral shaped. And vice versa, the vasa of non-spiral species when con- nected to normally spiral testes prevent the coiling of these organs. The vasa thus de- termine the final shape of the adult testis. This effect is only produced when testes and vasa, which originate independently of each other, become attached, i.e., come into cel- lular contact. The attachment of both organs takes place during pupal life, at which time both components involved are already quite well differentiated. Another type of interaction between tis- sues far advanced in development is illus- trated in the following example. If female genital discs of Drosophila are transplanted into male hosts, the host testes suffer exten- sive degeneration when the oviducts devel- oped from the transplanted disc become at- tached to the host testes. As in the previous case, the phenomenon is only brought about if cellular contact is established between these two organs; the principle causing it is not species specific (Bodenstein, ’46). In both cases cited, the stimuli eliciting the charac- teristic response are located in the genital ducts. There seems to exist an inductor-re- actor relationship between these two organs; the inductor in the vasa calls forth a growth reaction in the testis, and the inductor in the oviduct a degenerative process. The inductors act only when in contact with the reacting material. A similar mechanism of transmitting deter- mining influences through the substance of adjacent cells has been suggested in explain- ing the determination of certain structural units in the cuticle pattern of the bug Rhodnius (Wigglesworth, ’40a). In this insect, the nymphal cuticle bears bristles arising from small plaques which are distributed regularly over the surface of the abdomen. The closeness of these plaques to each other seems to affect the determination of new plaques. Each plaque apparently exerts through the substance of the epidermis cells that surround it an inhibiting influence which prevents the development of a new plaque within a certain radius. The distance over which this influence acts is dependent upon the number of epidermal cells between the existing plaques. It is noteworthy that the number of cells intervening between the plaques decides the limitation of the inhibit- ing influence. Since the cell number is the result of previous cell division, the role of growth processes in the determination events again becomes evident. In the mechanism of determination, influences of this kind are per- haps of major importance, although they are not easily detected experimentally. They might function fundamentally in the develop- mental interactions between cell populations, prospective organ regions, discs or body parts discussed in this chapter. HORMONES IN GROWTH AND DIFFERENTIATION While the processes of determination es- tablish, by progressive restriction of develop- mental potencies, the determination of organs and body parts to attain fixed fates, it is hormones that in many instances are known to be responsible for the realization of these fates. Hormone activity thus is an integral mechanism in the development of insects. For instance, the imaginal differentiation of Drosophila leg or eye discs can only take place after a specific hormone released by the ring gland has exerted its influence (Bodenstein, *43). Similarly, the nymphal cuticle of the blood-sucking bug Rhodnius (Wigglesworth, ’34 and ’36) or of the cock- roach (Scharrer, 46) or the skin of cater- pillars (Piepho, 38a) can only develop imaginal character when stimulated by the appropriate hormones. Moulting in insects is also under humoral control. Since moult- ing is an essential feature in insect growth, it is growth that is controlled by hormones. This fact is also well illustrated by the be- havior of Drosophila discs which grow only in an organic environment especially condi- tioned by hormones (Bodenstein, ’43). The ability of the different tissues to react to hormonal stimuli varies greatly. Some tissues become competent for response early in development, others later. Thus the epidermis of first stage Rhodnius nymphs (Wiggles- worth, ’34) or of caterpillars just hatched from the egg (Piepho, ’38b) already responds at this early stage to metamorphosis hor- mones and can hence be changed experi- mentally into imaginal epidermis. On the other hand, imaginal discs of first imstar Drosophila larvae do not respond to these hormones but acquire the ability to do so at the next larval stage (Bodenstein, 44). Moreover, not all tissues respond with like ease to the same hormone level (Bodenstein, 43; Kiihn and Piepho, ’36). The factors that make the discs or body parts competent to respond, or what this competence means in physicochemical terms, is obscure, as is the INSECTS nature of the mechanism of determination in discs in general. Now it should be understood that hormones are not merely general stimulators, creating a situation through which the invisibly fixed organization of the organic pattern becomes realized. They do more, for they often decide specifically which of the de- velopmental possibilities existing in the reacting material shall become manifest. In fact, it is often only through the specific action of these hormones that we can detect the existence of latent developmental poten- tialities. Take the case of the young cater- pillar epidermis. Under normal conditions, the young epidermis at the ensuing moult would have laid down caterpillar cuticle under the influence of the juvenile hormone. Yet, under experimental conditions, pre- maturely supplied with metamorphosis hor- mone, the same epidermis lays down pupal cuticle. In the reacting material, therefore, several competences must have existed side by side and the hormone decided when and which of these became manifest. As far as the determination of definite fates is con- cerned, the young larval epidermis is but labilely determined, for by the appropriate hormone it can be switched from larval to imaginal development. Not only epidermis but also organ anlagen show this dual type of determination (Bodenstein, ’42). Appar- ently a similar state of affairs prevails in the reversal of secondary sex characters in the Lepidoptera wing cited above. Here influences presumably hormonal in nature emanating from the organic environment were able to shift the development from a female into a male direction. Even at the end of develop- ment, when the adult organism emerges, it seems that absolute stability, i.e., a definite determination of the parts to pursue fixed fates, is mever reached. This is most strik- ingly demonstrated by the fact that adult insects can be made to moult in the presence of juvenile hormone and that part of the organism now possesses nymphal or larval character (Wigglesworth, 40; Piepho, ’38a). The developmental systems necessary for this reversion of development were hence latent in the cells, but needed special condi- tions for their realization. There are other instances with like implication. One of special interest is the case of the walking-stick Dixippus in which, by extirpation of the corpora allata, the normal hormonal balance was upset; this resulted in the formation of an eye-like structure in the ectoderm of the head. This eye, though never normally ob- 343 served in this species, is characteristic for related species (Pflugfelder, *39a). Now the purpose of presenting this information is to emphasize the fact that apparently rigidly determined systems may still contain latent developmental powers of amazing plasticity which, however, become manifest only under certain conditions. The role of hormones in growth has been mentioned. It is not astonishing that we find them playing an important part in the activa- tion of growth as a component process in regeneration. The walking-stick Dixippus loses its ability to regenerate new append- ages in the adult stage, while it freely re- generates these organs in nymphal stages. If the imago is induced to moult again by the transplantation of young corpora allata and thus supplied with juvenile hormone, the animal once again regains its powers to re- generate. Conversely, extirpation of the cor- pora allata in young nymphs results in the loss of the regeneration power, while reim- plantation of these glands restores this capa- city (Pflugfelder, ’39a, ’39b). Removal of the corpora allata in young nymphal stages also causes severe changes in all tissues (Pflugfelder, ’38; Scharrer, ’46). Many tis- sues degenerate, others show abnormal un- controlled growth often resembling certain tumors familiar to vertebrate pathologists. Mesodermal organs seem to be affected first. and more drastically than skin. Normal con- ditions may be restored by providing the abnormal animals again with corpora allata. Internal secretion and tissue proliferation as well as cell pathology are closely related. In general, we can visualize the part played by hormones in insect development as follows: In addition to their more specific actions in controlling the manifestation of certain morphogenetic and histogenetic features in development, the hormones play the more general role of metabolic regulators. By virtue of their reintegrating function, they bring the various parts of the organism under a common control. Thus any dis- turbance of this prominent integrating sys- tem results in severe developmental modifi- cations and functional alterations. REFERENCES Auerbach, C. 1936 The development of the legs and the wings and halteres in wild type and some mutant strains of Drosophila melanogaster. Trans. Royal Soc. Edinburgh, 53:787-815. Bock, E. 1939 Bildung und Differenzierung der Keimblatter bei Chrysopa perla (1.). Z. Morph. u. Okol. Tiere, 35:615-700. 344 EMBRYOGENESIS: PROGRESSIVE DIFFERENTIATION Bodenstein, D. 1936 Das Determinationsgesche- hen bei Insekten mit Ausschluss der friihembry- onalen Determination. Ergeb. d. Biol., 73:174— 234. 1937 Beintransplantationen an Lepid- opterenraupen, IV. Roux’ Arch. Entw.-mech., 136:745-785. 1941 Investigations on the problem of metamorphosis. VIII. Studies on leg determina- tion in insects. J. Exp. Zool., 87:31-53. 1942 Hormone controlled processes in in- sect development. Cold Spring Harbor Symp. on Quant. Biol., 70:17-26. 1943 Hormones and tissue competence in the development of Drosophila. Biol. Bull., 84: 34-58. 1944 The induction of larval molts in Drosophila. Biol. Bull., 86:113-124. 1946 Developmental relations between genital ducts and gonads in Drosophila. Biol. Bull., 97:288-294. 1947 Chemical alteration of development in Drosophila. Anat. Rec., 99:No. 4, p. 34. Brauer, A., and Taylor, A.C. 1934 Experiments to determine the presence, location and effects of an organization center in Bruchid (Coleoptera) eggs. Anat. Rec., 60 (Suppl.) :61. Brecher, L. 1924 Die Bedingungen fiir Fihler- fiisse bei Dixippus morosus. Roux’ Arch. Entw.- mech., 702:549-572. Caspari, E. 1941 The morphology and develop- ment of the wing pattern of Lepidoptera. Quart. Rey. Biol., 76:249-273. Chevais. S. 1944 Determinisme de la taille de V’oeil chez le mutant Bar de la Drosophile. Inter- vention d’une substance diffusible specifique. Bull. Biol. Fr. Belg., 78:1-39. Ewest, A. 1937 Struktur und erste Differenzie- rung im Ei des Mehlkaéfers Tenebrio molitor. Roux’ Arch. Entw.-mech., 735:689-752. Geigy, R. 1931 Erzeugung rein imaginaler De- fekte durch ultraviolette Eibestrahlung bei Dro- sophila melanogaster. Roux’ Arch. Entw.-mech., 125:406-447. Gloor, H. 1946 Die experimentelle Erzeugung von homoeotischen Modifikationen im Meta- thorax der Drosophila melanogaster. Arch. Jul. Klaus-stift., 27:308-311. 1947 Phanokopie-Versuche mit Ather an Drosophila. Rev. suisse Zool., 54:637-712. Goldschmidt, R. 1938 Physiological Genetics. Mc- Graw-Hill Book Co., New York. Hadorn, E., Bertani, G., and Gallera, J. 1949 Regulationsfahigkeit und Feldorganisation der mannlichen Genital-imaginalscheibe von Droso- phila melanogaster. Roux’ Arch. Entw.-mech., 144:31-70. Hegner, R. W. 1911 Experiments with Chry- somelid beetles. III. The effects of killing parts of the eggs of Leptinotarsa decemlineata. Biol. Bull., 20:237-251. Henke, K. 1933 Untersuchungen an Philosamia cynthia Drury zur Entwicklungsphysiologie des Zeichnungsmusters auf dem Schmetterlingsfliigel. Roux’ Arch. Entw.-mech., 728:15-107. Kiihn, A., and Piepho, H. 1936 Uber hormonale Wirkungen bei der Verpuppung der Schmetter- linge. Abh. Ges. Wiss. Gottingen, Math. physik. Kl., N.F., 2:No. 9, pp. 141-154. Liischer, M. 1944 Experimentelle Untersuchun- gen lber die larvale und die imaginale Determin- ation im Ei der Kleidermotte (Tineola biselliella Hum.). Rev. suisse Zool., 57:531-627. Paul, H. 1937 Transplantation und Regenera- tion der Fliigel zur Untersuchung ihrer Form- bildung bei einem Schmetterling mit Geschlechts- dimorphismus, Orgyia antiqua I.. Roux’ Arch. Entw.-mech., 736:64-111. Pauli, M. E. 1927 Die Entwicklung geschniirter und centrifugierter Eier von Calliphora erythro- cephala und Musca domestica. 7.. wiss. Zool., 129:483-540. Pflugfelder,O. 1938 Weitere experimentelle Un- tersuchungen iiber die Funktion der Corpora allata von Dizippus morosus. Z.. wiss. Zool., 157: 149-291. 1939a_ Beeinflussung von Regenerations- vorgangen bei Dizxippus morosus Br. durch Ex- stirpation und Transplantation der Corpora al- lata. Z. wiss. Zool., 152:159-184:. 1939b Wechselwirkungen von Driisen innerer Sekretion bei Dizxippus morosus Br. Z. wiss. Zool., 752:384408. Piepho, H. 1938a Uber die Auslésung der Rau- penhautung, Verpuppung und Imaginalentwick- lung an Hautimplantaten von Schmetterlingen. Biol. Zbl., 58:481-495. 1938b Uber die experimentelle Auslés- barkeit iiberzahliger Hautungen und vorzeitiger Verpuppung an Hautstiicken bei Kleinschmetter- lingen. Naturwissenschaften, 26:841-842. Poulson, D. F. 1947 The pole cells of Diptera, their fate and significance. Proc. Nat. Acad. Sci., 33:182-184. Reith, F. 1925 Die Entwicklung des Musca-Eies nach Ausschaltung verschiedener Eibereiche. Z. wiss. Zool., 726:181-238. 1935 Uber die Determination der Keim- esanlage bei Insekten (Ausschaltungsversuche am Ei des Riisselkafers Sitona lineata). 7Z.. wiss. Zool., 147:77-100. Scharrer, B. 1945 Experimental tumors after nerve section in an insect. Proc. Soc. Exp. Biol. & Med., 60:181-186. 1946 The role of corpora allata in the development of Leucophaea maderae (Orthop- tera). Endocrinology, 738:35—-45. Schnetter, M. 1934 Physiologische Untersu- chungen iiber das Differenzierungszentrum in der Embryonalentwicklung der MHonigbiene. Roux’ Arch. Entw.-mech., 737:285-323. Seidel, F. 1929 Untersuchung iiber das Bildungs- prinzip der Keimanlage im Ei der Libelle Pla- tycnemis pennipes, I-IV. Roux’ Arch. Entw.- mech., 779:322-440. 1932 Die Potenzen der Furchungskerne im Libellenei und ihre Rolle bei der Aktivierung des Bildungszentrums. Roux’ Arch. Entw.-mech., 126:213-276. Seidel, F., Bock, E., and Krause, G. 1940 Die Organisation des Insekteneies. Naturwissenschaf- ten, 28:433-446. INSECTS Steinberg, A. G. 1941 A reconsideration of the mode of development of the bar eye of Drosophila melanogaster. Genetics, 26:325-346. Stern, C. 1941a The growth of testes in Droso- phila. I. The relation between vas deferens and testis within various species. J. Exp. Zool., 87: 113-158. 1941b The growth of testes in Drosophila. II. The nature of interspecific differences. J. Exp. Zool., 87:159-180. Vogt, M. 1946a Zur labilen Determination der Imaginalscheiben von Drosophila, I. Biol. Ztrbl., 65:223-238. 1946b Zur labilen Determination der Imaginalscheiben von Drosophila, II. Biol. Ztrbl., 65:238-254. 1947a Zur labilen Determination der Imaginalscheiben von Drosophila, III. Biol. Ztrbl., 66:81-105. 345 1947b Beeinflussung der Antennendif- ferenzierung durch Colchicin bei der Drosophila- mutante Aristopedia. Experientia, 3:156. Waddington, C.H. 1942 Growth and determina- tion in the development of Drosophila. Nature, 149:264:. Wigglesworth, V. B. 1934 The physiology of ecdysis in Rhodnius prolixus (Hemiptera). II. Factors controlling moulting and metamorphosis. Q. Jour. Micro. Sci., 77:191-222. 1936 The function of the corpus allatum in the growth and reproduction of Rhodnius pro- lixus (Hemiptera). Q. Jour. Micr. Sci., 79:91-121. 1940a Local and general factors in the development of “pattern” in Rhodnius prolixus (Hemiptera). J. Exp. Biol., 77:180-200. 1940b The determination of character at metamorphosis in Rhodnius prolixus (Hemip- tera). J. Exp. Biol., 77:201-222. Section VII SPECIAL VERTEBRATE ORGANOGENESIS CHAP TE Raa Nervous System (Neurogenesis) PAUL WEISS THE OBJECT AND THE PROBLEMS WE az to deal here with the causal analysis of ‘the development of the nervous system.” In that generality, the task is simply un- manageable. Of the innumerable aspects the mature nervous system offers to the observer, each one has had its characteristic ontoge- netic history, hence raises separate questions as to time and manner of its origin and as to mode and means of its ontogenetic trans- formations. This points us to the only prac- tical approach, which lies in resolving the confusing complexity of the system into simpler components and addressing our questions to the more elementary events thus singled out. Most of the following account will be essentially a sample exercise in phrasing and sorting such questions of suffi- cient concreteness as to offer hope for precise answers. The answers themselves are mostly still in a very fragmentary state and will be presented without glossing over their often provisional character. I have chosen topics and examples chiefly in the spirit of the guiding theme of this book, which is to il- luminate, rather than cover, the processes of development. This also explains the argu- mentative, rather than reportorial, manner of presentation. It reflects the effort to give a coherent and consistent picture, in which facts and data are rated not as isolated items, but as tools for the clarification and solution of problems—not as sheer statements, but as answers to questions; which makes the text useful as a guide more to the understanding, than to the literature, of the field. 346 Some familiarity with the main morpho- logical, physiological, and embryological features of the nervous system will be taken for granted. Yet a brief listing of the most prominent ones may help to keep our analytical questions properly focussed from the start. Somewhat arbitrarily we shall separate the discussion of the central nervous system, which serves intercommunication among its constituent units, from that of the peripheral nerves, which serve communi- cation between the former and the non- nervous tissues of the body. THE PERIPHERAL NERVE Nerves are composite structures, contain- ing bundles of nerve fibers of different classes, associated in variable numbers, proportions, and groupings, and held together and sheathed by connective tissue, in which course blood and lymph vessels and endo- neurial fluid. In the so-called plexuses, nerves regroup or exchange some of their fibers. Farther peripherally, they branch by succes- sive dichotomies and distribute their branches over the periphery according to patterns characteristic of the given peripheral sector or organ, with considerable latitude for individual variation. The component nerve fibers themselves are composite (Fig. 125), with the axis cylinder (axon or neurite, @)—a protoplasmic exten- sion of the centrally located cell body (peri- karyon)—at the core; covered by a membrane or medullated sheath (77) consisting of alter- NERVOUS SYSTEM nating layers of myelin and protein; sur- rounded by the thin protoplasm of the sheath cells of Schwann (s), arrayed in tandem and (in medullated fibers) corresponding each to an internodal segment (between two nodes of Ranvier); the whole enclosed in a collage- nous tube (¢). Axons vary in diameter (cali- ber) according to the classes to which they belong (sympathetic, somatic motor, tactile sensory, etc.) and, within each class, accord- ing to length and peripheral distribution; the thickness of the myelin sheath and usu- 347 different types of neurons differ substantially in their constitution (molecular content), as evidenced by selective reactions to poisons, drugs, histological stains, hormones, and the production of specific secretions (neurosecre- tions). This specificity of neurons may be assumed to extend into the axis cylinders, which are true protoplasmic extensions of the cell. Many more items could be added to this list. In their totality, they make up what may be described as the “functional architec- ture” of the CNS. Fig. 125. Diagram of composition of single nerve fiber (explanation in text). ally also the length of internodes vary pro- portionately. These parameters are impor- tantly related to functional properties (con- duction velocity, threshold, susceptibility, etc.). Fibers remain single or branch out, either by terminal flaring or by side shoots (collaterals), in accordance with functional needs. Each branch ends either blindly or on a specific end-organ with which it forms transmissive and trophic connections (“in- nervation”). Nerve fibers of a given class generally are found connected with the appropriately matching type of end-organ. The density of innervation varies regionally and with the type of innervated organ. THE CENTRAL NERVOUS SYSTEM (CNS) The CNS and its outpost ganglia contain the cell bodies (perikarya) of the peripheral neurites, and the former also a host of intra- central fibers, either in orderly bundles (fiber tracts, funiculi, commissures) or ir- regularly dispersed as “neuropil.” In addi- tion, these neural masses are interlaced with various types of glia cells, blood vessels, and, according to some, ground substance of ill- defined nature. Morphologically, the cell bodies vary characteristically in size, loca- tion, grouping, number, distribution of proc- esses, and mode of interconnections, all of which may be called “geometrical” criteria. In addition to these distinctions, however, NEUROGENESIS Every feature thus singled out raises a separate question as to its origin. We have noted the finished products; but how have they come about? This circumscribes the task of neurogenesis. Our goal is to recon- struct the whole causal chain of events that leads from the properties of the egg cell to each particular item on our list. These events include complex molecular interactions with the emergence of new molecular species and the loss of others; displacements and rear- rangements of substance on a molecular, micellar, cellular, and supracellular scale; metabolic energy production and consump- tion; electric, thermal, hydrodynamic, and mechanostatic (pressure-tension) phenomena, etc.; at any rate, processes that are in prin- ciple observable, measurable, and describ- able in terms of the substances, forces, inter- actions, and conditions actually present at each step. This is the object of ontogenetic analysis. Phylogenetic (evolutionary) con- siderations, introducing past history, do not enter into this causal (or operational) analy- sis at all; they only explain why of the in- finite conceivable number of possible causal chains of events, only a very limited selec- tion has found materialization. Causal analysis must be preceded by a complete description of the events that are to be explained; description, of course, in objective and, wherever possible, quantitative 348 terms. Much of the current research in neuro- genesis is still in this descriptive phase, even though, in order to get the necessary data, it makes extensive use of experiments. The mode of the formation of a nerve fiber is a case in point. At the turn of the century, there were two opposing schools of thought given to different interpretations of incon- clusive static observations. The one main- tained that the axis cylinder is produced in fractions by tandem chains of peripheral cells, which are then secondarily joined to- gether into a single strand, drainpipe fashion; while the other contended the axon to be a protoplasmic sprout of a single central neuro- blast cell. The ingenious experimental feat of Harrison (’07a, ’10) in isolating the sup- posed neuroblasts in extraneous media devoid of peripheral cells to test whether or not they could still form axons, settled the issue: they could. A descriptive datum had been ascertained by an experimental method. Then, passing on to explore the reaction of cells in vitro to solid fiber substrata, Harrison (14) carried his research into the strictly analytical sphere, where one examines why things happen as they do—in the given case, why the nerve fiber follows one course rather than another. The Hows and Whys of our questions are thus intimately related and often enough blend into one. With this in mind we may now attempt to carve out some specific neurogenetic questions from the body of neurological data presented above. Evi- dently, this can only be a crude and frag- mentary sample. Why and where does the axon arise from the neuroblast? What causes its elongation? What gives it its course? Do the trunks and branches of the mature nerves reflect the orientation of early outgrowth? Is that out- srowth strictly oriented or is it haphazard, followed by selective abolition of unsuccess- ful connections? What determines deflections or other changes of course? What causes branches to arise, and where? Are tissues flooded with nerve fibers, or is admission selective? If the latter, how is invasion held in check? And is penetration tan- tamount to functionally effective innervation? What causes the association of sheath cells and nerve fibers, and what is the mechanism of myelin formation? How do fibers group into bundles—by active aggregation or by the enveloping action of connective tissue? And what determines the places and proportions in which the various tissue elements combine to form nerves? How does it happen that SPECIAL VERTEBRATE ORGANOGENESIS fibers of similar function are often grouped together, and how do they each reach their appropriate destinations? Or do they? And, if not, how can central functions fail to be confused? How does a nerve fiber gain in width, and what decides its final caliber? And does it change with body growth? What controls the number of fibers available for a given area—size of the source, frequency of branching, overproduction followed by ter- minal screening, or all of these? And if the size of the source is a factor, what determines it? This points us to the centers. How does the neural plate transform into primordia of brain, spinal cord, and ganglia? How does it grow? How do its cell groups specialize for their respective formative tasks, how early, in what places and what se- quence? What makes them divide or cease to divide? What causes them to migrate and in what directions, and what to assemble in defined locations? What sets the numbers and quotas of the different neuron types, and adjusts them to the functional needs of the individual? How do they achieve selective interconnections on which their later func- tioning will depend? And which ones of these are really relevant to the specific patterns, rather than just the general execution, of central functions? What provides the neuron population with the proper contingent of supportive, protective, and nutrient cells and structures of other origins, in varying com- binations according to the local needs? And how much interdependence and interaction in growth and differentiation is there between different central regions before and after they have become segregated? If there are inter- actions, what is their nature and how are they transmitted? Does exercise and practice have a constructive, or at least modifying. effect on central pathways or central size? Are fluctuating neripheral demands taken in- to account in the development of centers. and, if so. by what means? Can growing centers adjust to lesions or deformation, and how—by regeneration, comnensatoryv srowth. or substitutive functional correc- tions? And can the develonment of overt behavior be correlated with, or even ex- plained by, the stepwise emergence of neural apparatuses? Specific questions like these, rather than noncommittal generalities about ‘the de- velopment of the nervous system being a matter of metabolic processes, gradient fields and enzymatic reactions,” are effective guides to useful research. NERVOUS SYSTEM NERVE REGENERATION Since our insight into nerve development and growth has been greatly aided by studies on nerve regeneration, the essentials of this phenomenon may be briefly recapitulated here for later reference; for fuller reviews, see Cajal (’28), Nageotte (22), and Boeke (35): 349 freezing, chemical damage, etc.), the seg- ment lying distally to the lesion (the “distal” or “peripheral” stump), within a few days loses conductivity, and the individual nerve fibers in it become converted into non-con- ducting plasmatic strands (“Schwann cords” or ‘“Buengner’s cords”); myelin and axis cylinder remnants break down (o, Fig. 126B). and as they are being resorbed, their place PROXIMAL Nerve regeneration is the restoration of morphological and physiological continuity in a transected nerve. The older supposition that this may take place per primam fu- sionem of the severed ends has proved un- tenable; the observation that cut ends of axons in tissue culture that lie within a dis- tance of a few micra may merge during the first hours after severance (Levi, ’34), does not apply to nerves in the body, where this condition is practically never realized. In the body, the nerve is restituted by renewed outgrowth of fibers from the proximal stump, repeating with some modifications the process of embryonic outgrowth. Briefly, the events are as follows (Fig. 126). After a nerve has been severed or other- wise locally disrupted (e.g., by pressure, oe DISTAL Fig. 126. Diagram of regeneration of single nerve fiber after transection (explanation in text). is taken by the hypertrophying and multiply- ing sheath cells of Schwann (s, Fig. 126C). This combination of regressive and prolifera- tive processes is generally referred to as “Wallerian degeneration.” In the “proximal” or “central” stump, it remains confined to the immediate vicinity of the lesion, and although the whole injured neuron, includ- ing the central perikaryon, shows some trau- matic reaction, the part that has retained its continuity with the central cell body soon becomes the source of the regenerative proc- ess. The free tip of each proximal axon stump assumes amoeboid activity and extends into the surroundings much the same as in the first development (sp, Fig. 126B). Branching is frequent, but many of the branches are ar- 350 rested in their course and remain abortive. While the outgrowing axon branches roam about the wound area, Schwann cells spill from the cut nerve ends, notably from the “degenerating” peripheral stump. When an advancing axon tip meets such a Schwann cord, it follows it and is thus guided into the peripheral stump (Fig. 126C), and through it to the peripheral tissues, where new trans- missive connections can be established if the arriving nerve branch is of the proper type. The highly irregular connective tissue that seals the cut nerve ends, commonly referred to as “scar,” causes the dissipation of a large proportion of the newly formed branches, which may form dense tangles called “neuromas.” Nerve regeneration thus involves a great deal of overproduction and wastage of sprouts. Eventually, a near-nor- mal number of fibers may become collected in the deserted channels of the distal stump. The new branches, small at first (14 or less), gradually gain in width and develop on their surface a new myelin sheath (m, Fig. 126D), which thickens proportionately. In this man- ner, lines for the conduction of excitation between centers and periphery are rees- tablished. However, owing to the misdirection of many fibers into wrong channels, the physiological control restored by regenerated nerves does not usually attain the original perfection. In contrast to the practically unlimited power of regeneration observed in peripheral nerves, regenerative growth of intracentral nerve fibers declines with age and phyloge- netic rank so as to be little more than abortive in brain and spinal cord of adult mammals under ordinary circumstances. Whether this is due to an intrinsically lower growth po- tential of the neurons or to less favorable growth support, perhaps even greater active obstruction, by the central, as compared to the peripheral, environment, is still a matter of debate. ANALYSIS OF THE DEVELOPMENT OF A NEURON OUTGROWTH OF AXIS CYLINDER Mechanism of Elongation. As was indicated above, it is now indisputably established that the neurite (axon) develops as a direct proto- plasmic extension of the nerve cell. At a given point along the circumference of the neuroblast, cytoplasm is protruded to form a short thread with a highly mobile tip. Presumably any breach in the cell surface may serve as outlet. The fact that in the SPECIAL VERTEBRATE ORGANOGENESIS embryo the sprouts tend to emerge from the same sides in all neuroblasts of a given group must be ascribed to certain polarizing fac- tors in the cellular environment, analogous to the determination of rootlet formation in Fucus eggs by the polar action of electric fields, ultraviolet light, pH gradients, etc. (Whitaker, *40). The young sprout of axoplasm has no rigid axis skeleton, no firm sheath, nothing to propel it in a predetermined direction. The sprout continues to elongate by virtue of forces residing chiefly within the cell of origin, but the course of the elongation is determined by extraneous factors. In his classic observations on axon outgrowth in tissue culture, Harrison (710) correctly iden- tified the mode of advance as of the amoeboid type, which view has been fully borne out by later observations in explants (Lewis and Lewis, ’12; Levi, 34) and in the living tad- pole (Speidel, °33). Adopting and partly amplifying Lewis’ (50) interpretation, we may conceive of the sprout as a cylinder of firmly gelated ectoplasm surrounding a core of more fluid entoplasm streaming from the cell body distad. At the free tip, this central stream would erupt in numerous pseudo- podial processes, which then compete among one another hydrodynamically for the com- mon axial current (Fig. 127). The branch that succeeds in draining the inflow into its own channel thus automatically obliterates the weaker pseudopods, and as its surface becomes gelated, it adds its length to the already consolidated older parts of the fiber lying behind it. Meanwhile, the tip bursts forth anew, and thus the fiber advances in a continuous series of steps of protrusion of pseudopods, competition, and consolidation. Evidently, the protoplasm for the fiber is produced in the cell body, but it is added at the tip to which it is conveyed by the central stream. The motive mechanism of this convection is still obscure, but it may con- sist of peristaltic contraction-relaxation waves of the fiber surface. It provides some sort of “pumping” action, which after the fiber ha: ceased to elongate, continues to supply proto- plasm for its further growth in width (see p. 363). In contrast to true amoeboid loco- motion, however, the rear end of the nerve cell remains anchored to its surroundings so that instead of dragging the bulk of the cel! after it, the advancing tip merely spins ou! a thread of increasing length between itself and the cell body. The described active advance ends as soon as the free tip of a fiber attaches itself perma- NERVOUS SYSTEM nently to a peripheral receptor or effector cell. Further elongation becomes essentially a matter of passive extension, the fiber being in tow by the terminal tissues, which are subject to considerable migrations (e.g., muscle buds) and displacements during the subsequent phases of growth (Fig. 128). Because of this ““towing”’ process, the primary erowth pattern of nerves becomes greatly a1 problem, and a variety of “tropisms” and “attractions” of chemical, electrical, me- chanical or undefined physiological nature have been suggested as the orienting agents (see Harrison, ’35a,b). At present, we are ap- proaching a rather unified concept of the mechanism of nerve fiber orientation, which is summarized in the following account (condensed from Weiss, “41c, °44, ’50c). b c d e Fig. 127. Five consecutive phases in the advance of an axon tip (semidiagrammatic). Arrows indicate directions of flow, thrust and drain of neuroplasm. In e, dichotomous branching of fiber has been initiated. Dotted portions represent the location of earlier protrusions that have been sucked back by the draining force of the axial stream. distorted. This explains why ontogenetic shifts, hence phylogenetic relations, of in- dividual muscles can often be traced through their nerve supply, as in the pelvic fins of fishes that have migrated far forward until they have come to lie ahead of the pectoral fins. During its period of free advance, the orientation of the nerve fiber is of course determined by the course which its roving tip takes. The early “pioneer” or pathfinder fibers thus lay down the primary nerve con- nections to the nerveless tissues which they invade. Since later fibers simply follow the course of the earlier ones, the problem of nerve orientation concerns primarily the pioneers. Much experimental work, and even more speculation, has been devoted to this Mechanism of Orientation. Without recount- ing the trials and errors of the past, it is yet instructive to point to one basic fallacy of earlier concepts, namely, the tacit assump- tion that nerve fibers can penetrate structure- less space in the manner in which plants can grow into air or water—an impression strengthened by the selective nerve stains, which impregnate nerve fibers to the exclu- sion of their surroundings. The suggestiveness of plant growth as a model of nerve growth is clearly reflected in the widespread use of botanical similes in neurological terminol- ogy; e.g., “dendrites,” “rami,” “roots,” ‘“ar- borization,” “sprouts.” Actually, however, according to the best available experimental evidence, nerve processes, like most animal tissue cells, are unable to push freely into oon a liquid, but can only proceed along inter- faces, either between a solid and a liquid, or between two immiscible liquids, or between a liquid and a gas. The nearest analogon among plants would be the clinging vine. A nerve tip can traverse not even a small liquid gap without an interfacial bridge. Interfaces capable of serving as the requi- site substrata are furnished in the body by TOWING Fig. 128. Three phases in development of nerves (diagrammatic). Top: Pioneering phase (free fiber tip advances into surroundings). Middle: Applica- tion phase (pioneering tip has become attached to peripheral cell, younger tips apply themselves to course of older ones). Bottom: Towing phase (shift of peripheral cell produces corresponding displace- ment of attached nerve fiber). all the fibrous units (fibrils, fibers, filaments) that pervade the liquid spaces in and between cells and tissues and constitute the solid framework of the “ground substances” (see Section III, Chapter 1, by Schmitt). They consist mostly of chains of filamentous pro- tein molecules combined into bundles and networks of submicroscopic and microscopic dimensions. Along such filaments the ter- minal filopodia of the nerve fiber are drawn out by interfacial forces of still unresolved nature which cause protoplasm to spread out along the interface, grossly comparable to a “wetting” process. The linearity of the fibrous units along which they extend is a major factor in guiding the extending nerve fibers. In a planar interface, they would fuse to a sort of “terminal web.” However, the linear guide structures are the ones of greatest practical importance, since even planar surfaces commonly contain inhomo- geneities that describe linear tracts within the common plane (e.g., the fibrous constitu- SPECIAL VERTEBRATE ORGANOGENESIS ents of coats or membranes). The principle according to which nerve fiber tips are guided in their course by contact with sur- rounding structures has been designated as “contact guidance” (Weiss, ’41c). In an irregular network, fibrils intersect at countless places and angles. Nerve tips advancing on such a trellis will be split at each intersection, but, as was explained above, competition will usually obliterate all but one of these terminal branches, and only this one will proceed. The decision of which one of the multiple projections will endure in any given instance may be essentially a matter of the accidents of the local situation. If so, the resulting nerve course will be irregular and tortuous (Fig. 129b, e), as is the case in the neuropil of the nerve centers, in scar tissue (e.g., between severed nerve stumps), and in the plasma clots of ordinary tissue cultures. On the other hand, the more the meshes of the fibrillar network are ori- ented in a given prevailing direction, the more the resulting nerve fiber course, tracing the common directional component, will like- wise become definitely oriented (Fig. 129a). The extreme of this condition is attained when the fibrous matrix consists of parallel guide rails which leave the single-tracked nerve fibers no alternative course (Fig. 129c,d). In this case, nerve orientation resolves itself completely into a matter of the orientation of the underlying substratum and can there- fore be controlled by way of the latter, as has been proved by a variety of observations and experiments both in the living animal and in vitro. The following examples may serve as illustrations. When tension is applied in tissue culture to a blood plasma clot, either during or after coagulation, the meshes of the random net- work of fibrin threads are drawn out in the general direction of the lines of stress. Nerve fibers allowed to grow out in such a medium then move in the same prevailing direction (Weiss, ’34a). Thus, by orienting the col- loidal matrix, tension can indirectly orient nerve growth. The immediate factor is the orientation of the matrix, irrespective of how it has been obtained. Fibrous tissue exudates spreading along surfaces and being drawn out in the direction of flow, for instance, act in like manner (Weiss, ’45). That this principle of contact guidance is equally valid within the living body has been substantiated in numerous instances, most strikingly by the directional control of nerve regeneration. Without intervention, regenerating nerve fibers commonly take random courses. But NERVOUS SYSTEM it has been possible to direct them into a straight oriented course by forcing the un- derlying matrix into parallel alignment. This has been achieved both in the gelati- nous fin tissue of larval amphibians (Weiss, 50a) and in the blood clots binding severed nerve stumps of adult mammals (Weiss, 44; Weiss and Taylor, ’43). In the latter case Sema tatnk meus 353 tion. If the nerve course depends on preneural guide structures in the colloidal matrices, our attention must therefore turn to the factors producing structural orientation. Tension being presumably the commonest orienting agent, let us examine first the potential sources of tensional stresses in the body. Oriented tensions arise from external stretch- SS oe — ON ai | 6b cd ) Fig. 129. Advance of nerve fibers in fibrous media of different degrees of ultrastructural organization (randomness in center turning into prevailing horizontal orientation in left part, and strict vertical orienta- tion in right part of diagram). Along random meshes of center strip, the course of fibers a, b and e is tortuous, with frequent branching; in the more orderly parts of the medium, fiber courses become corre- spondingly aligned; in a rigorously oriented medium, fibers (c, d) run straight and remain undivided. there is a primary phase, during which longi- tudinal tensions orient the fibrin of the blood clot in a prevailing direction from stump to stump, followed by a secondary phase, dur- ing which fibrinolytic agents discharged in the wound dissolve all remaining disoriented crosslinks between the longitudinal fibrin strands. This gives a good illustration of the multiplicity of factors involved in orienta- tion. In view of the ubiquitous presence of fibrous elements in the tissue spaces, these examples may be considered to be fair models of the normal mechanism of nerve orienta- ing or internal shrinkage of a cohesive sys- tem. Differential growth, resulting in exten- sive displacement of body parts relative to one another, is an ample source of stretch effects. Nerve growth may thus be expected to trail actively advancing organs even prior to being taken in tow by them. Localized shrinkage is perhaps even more important as a source of stress. Such shrinkage occurs, for instance, around any intensely proliferat- ing area as a result of a peculiar dehydrating effect which proliferating cells exert on sur- rounding colloids (Weiss, ’29, ’34a; Grossfeld, 34). The resulting local contraction of the 354 fibrillar network automatically distorts the meshes into a radial pattern converging upon the proliferating center (Fig. 130, top). Sub- sequent nerve growth, being guided over these radial pathways toward the center, naturally will give the illusion of having been “attracted” by it. We may call this the SAA VTL ETO ERAS SNE REAR ASE Ru : Vi —— jase a UA LY ; SPECIAL VERTEBRATE ORGANOGENESIS structures, cogently explicable in terms of demonstrable chains of physico-chemical events. This may appropriately be called the “two-center effect.” Although tension has been revealed as the most common effector mechanism in the production of guide structures, it is concelv- iN VY, —_ SSS Fig. 130. Effect of local contraction on ultrastructure of a fibrous medium. Top: “‘One-center effect.” Shrinkage of area indicated by black circle from the dimensions of the left panel to those of the right panel produces radial distortion of contiguous network. Bottom: ““Two-center effect.” Two “one-center effects” in a common network yield resultant preferential orientation along connecting line between the two centers. “one-center effect.” It is a concrete example of one way in which localized chemical activ- ity can translate itself into structural pat- terns. In the presence of two separate centers of proliferation (hence, two contracting foci), the intermediate fibrous matrix is being stretched, hence becomes aligned, along the connecting line. There is thus established a fibrillar bridge which any nerve fibers in that area are bound to follow (Fig. 130, bot- tom). Figure 131, top, shows, for example, the straight tract of nerve fibers grown re- ciprocally between two proliferating spinal ganglia in a thin plasma lamella, guided not by spurious “attractions” from the dis- tance, but by contact with tangible guide able that other vectorial agents besides ten- sion, such as hydrodynamic currents, high electrostatic potentials, electrophoresis, or perhaps still wholly unsuspected processes, could effect fibrillar orientation of the req- uisite kind. On a strictly oriented substratum, nerve fiber growth is thus fully determined by con- tact guidance. On a substratum of random configuration, that is, one not previously sub- jected to orienting factors, nerve growth would remain correspondingly random unless additional factors became operative. Since each one of the countless intersections of an irregular pathway system presents the nerve tip with alternative directions, any factor that systematically favors one general Fig. 131. Two-center effect. Top: “Bridge” of cells and nerve fibers that has formed between two em- bryonic spinal ganglia of chick (dark areas) cultured in vitro in thin blood plasma membrane. X 48. (From Weiss, 34.) Middle: Regenerated nerve fibers forming “bridges” between “‘proximal” stumps (a, c) and dislocated “peripheral” stumps (recurrent, b; laterally displaced, d) (from Cajal, ’28). Bottom: “Bridges” of Schwann cells that have grown out between the open ends of two fragments of adult rat nerve explanted in a thin blood plasma clot (from Weiss, ’52b). 355 356 direction over another would entail a statis- tical deviation of the hibers in the favoreu airection—an over-all trend rather than a common course. Many dendritic melds, tov instance, show such a trend. Lhe graduas detlection, rather than detinite orientation, toward the cathodal site of neurons in tissue cultures exposed to electric tields of proper density (Marsh and Beams, ’40) presents all the aspects of this picture, indicating simply relative inhibition of filopodial protrusion on the anodal side with no cathodal “stimu- lation” or “attraction,” a view confirmed by direct observations on slime moulds (Ander- son, 51). These electric effects evidently operate not by laying down pathways, but by prohibiting some ot the existing ones (see rig. 48 in Weiss, ’50a). It must be further postulated that the chemical characteristics of the pathway sys- tems endow contact guidance with an ele- ment of selectivity. It would seem impossible otherwise to explain the fact reported below that ditferent kinds of nerve fibers tend to choose different pathway systems when faced with a choice. Only a faint trace of such selectivity has thus far been observed in tissue culture in the preference of nerve tips for interfaces of tissue exudate rather than of fibrin (Weiss, 45). At any rate, such discriminatory ability is based on affinity for the chemical constitution of the contact surface rather than on the perception of concentration gradients of diffusing sub- stances as surmised in the theory of chemo- tropism. A singularly strong affinity of this kind seems to exist between axoplasm and the protoplasm of the sheath cells of Schwann (see p. 367). The described mode of advance of the nerve tip makes it clear that the “rate of free outgrowth” is a function of both the neuron itself and the configuration of the pathway system. The rates are of similar magnitude whether determined in the em- bryo, in nerve regeneration or in tissue cul- ture (Harrison, ’35b). It must be borne in mind, however, that these are over-all values of length over time without implying uni- form speed. Actually, the advance consists of alternating spurts and delays, the fre- quency of the latter mounting with increas- ing irregularity (‘intersectedness”) of the substratum (see Weiss and Garber, ’52). Con- sequently, the total rate is fastest along straight oriented pathways (e.g., during nerve regeneration inside old Schwann tubes; Gut- mann, Guttmann, Medawar and Young, ’42) and slowest in the dense and confused fiber SPECIAL VERTEBRATE ORGANOGENESIS tangle of a scar. The maximum rate of ad- vance under optimal conditions is of the order of a few millimeters per day (at 37° C.), which is close to the autonomous rate of proximodistal movement of axoplasm ob- served in the perpetual growth of neurons, as described below (p. 364). Any faster elonga- tion of nerve fibers (e.g., Waislocki and Singer, *46) suggests passive elongation by towing. Neurotropism. “Contact guidance,” as here described, is but a modified and more de- tailed version of such concepts of nerve orientation as have been proposed by His (1887), Harrison (14), and Dustin (10). They all imply that the nerve fiber is con- ducted on its way by markings of its im- mediate contact surroundings, rather than directed from a distance by the tissue of destination issuing “attractive” forces or merely acting in the manner of a beacon. Such “distance action,” commonly referred to as “neurotropism” and assumed to be a form of either galvanotropism or chemo- tropism, has been invoked to explain oriented nerve growth in the embryo (e.g., Kappers, 17), as well as during later nerve regenera- tion (foremost: Cajal, ’28; Forssman, 1900). This concept dates from a period in which the mechanism of nerve growth was stili poorly understood; before it was realized, for instance, that nerve fibers cannot pene- trate into the interior of a structureless fluid in the manner of plants. It was assumed that remote tissues, by virtue of their electric charges or of specific chemical emanations, could “attract” nerve growth from a distance. This assumption implies (1) that the sup- posed gradients be steady and durable, and undisturbed by any activities within the in- tervening distance, and (2) that the nerve tip has means not only for perceiving the required minute differentials of potential or concentration, but also for translating them into corresponding steering actions. Neither these premises nor the basic thesis of distance attraction has ever been critically demon- strated. On the contrary, overwhelming evi- dence has accumulated over the years to dis- prove them. Repeated attempts to obtain directed nerve growth along the stream lines of an electric field have remained unsuccessful. Indeed the very possibility of an electric guidance is ruled out by the fact that nerve growth often proceeds simultaneously in diametrically op- posite directions (e.g., the ascending and descending branches of dorsal root fibers; recurrent fibers in nerve regeneration; re- NERVOUS SYSTEM SPM] shai Fig. 132. Nerve growth in “alternate-choice” experiment (from Weiss and Taylor, ’44). Top: Diagram of operation. Proximal nerve stump (left) is introduced into stem of Y-shaped artery; one branch of artery is sealed off distally, the other branch contains degenerating nerve as supposed “lure.” Middle: Experimental case, 20 weeks after operation, showing both the blind and the connected branches filled with regenerated nerve fibers. Bottom: Detail from same case at bifurcation at higher magnification. ciprocal fiber tracts in the brain, etc.). Thus, the only known electric effects are the inhibi- tory ones mentioned above (p. 356). Chemical “attractions” have been postu- lated largely on the strength of Cajal’s (’28) varied experiments demonstrating a tendency of regenerating nerve fibers to converge upon the open end of any degenerated nerve stump (Fig. 131, middle), as if the latter were a source of ‘“‘neurotropically” active substances. While the observations were correct, the in- terpretation was not. The nerve fibers are 358 guided toward the peripheral stump, not by a chemical concentration gradient, but by a structural bridge of Schwann cells which has previously spanned the gap as a result of the orienting “two-center effect” (see p. 354) which the two proliferating cut surfaces exert upon the intervening blood clot. It is easy to demonstrate this effect directly in tissue culture (Fig. 131, bottom) by placing two fragments of degenerated (axon-free) peripheral nerve into a thin plasma clot (Weiss, ’52b). Evidently, if axons were to Fig. 133. Deflection of peripheral limb nerve plexus toward transplanted limb buds (combined from Detwiler, ’36b). The left half shows the plexus of a normal forelimb (contribution from segments 3, 4, 5), the right half nerve supply in two experi- mental cases in which limb buds had been trans- planted from their normal] site (7) to anterior or posterior levels, respectively, as indicated by arrows. grow from one of the stumps, the connecting strand of Schwann cells would automatically Jead them over into the other stump. The chemical activity of the degenerated stump thus plays no part other than that of an accessory aid to structural orientation. In confirmation of this fact, degenerated nerve in a liquid medium leaves nerve growth wholly unaffected despite enhanced diffusion (Weiss and Taylor, °44), and conversely, oriented structural pathways are followed by nerve fibers regardless of whether or not they lead to supposedly “attractive” destina- tions. For instance, when a proximal nerve stump as fiber source is introduced into the stem of a bifurcated blood-filled tube, one branch of which contains degenerated nerve while the other ends blindly (Fig. 132), regenerating nerve fibers fill both branches equally well and abundantly (Weiss and Taylor, °44). SPECIAL VERTEBRATE ORGANOGENESIS In conclusion, the idea that remote tissues of destination can attract nerve fibers directly may be safely discounted; such tissues do, however, contribute to the formation of nerve patterns indirectly by the creation of pathways, as here outlined, as well as by various secondary effects on later neuron development to be detailed below. Ontogeny of Nerve Patterns. Our task is now to explore whether what we have outlined in the foregoing pages for nerve orientation in general is sufficient to account adequately for the specific nerve patterns observed in the organism. Nerve Deflection Toward Growing Organs. Evidence that embryonic nerve growth is often actively routed toward rapidly growing peripheral organs rests largely on the ex- perimental work of Detwiler (summarized in Detwiler, ’36b). Transplanting urodele limb buds, prior to the outgrowth of segmental nerves, to farther anterior or posterior sites entailed a certain shift of their nerve supply, as is illustrated in Figure 133, which is a composite of two typical cases. It can be seen that there is a tendency for the limb plexus to originate in more anterior or more poste- rior segments than normally. Yet, since this shift of the nerve source is less extensive than the displacement of the limb, the nerve trunks appear to slant forward or backward, as if “attracted” by the actual limb site (see also Lovell, 31). Such deflection toward the actively growing limb can readily be under- stood as an instance of the “two-center effect” outlined above. That the effect is quite un- specific is demonstrated by the fact that the limb nerves are similarly deflected toward transplanted eyes and nasal placodes (Det- wiler and Van Dyke, ’34), although in the latter case it has not been made clear how much of the observed cord-nasal connection originated in the cord and how much in the olfactory epithelium. Brain grafts on the other hand exert no such effect (Detwiler, ’36a), perhaps because the proliferating cell layer is shut in and not exposed to the sur- rounding matrix. Frog limb buds deprived of their ipsilateral nerve source often secure vicarious supply from the opposite side (Hamburger, ’29). It must be considered, however, that this is initiated during an early stage, when both hind limbs are still close together, and that subsequent dislocations of the plexus and fasciculation of the successful branches (see below, p. 366) tend to create an exaggerated idea of the power of nerves to reach their destination by detours. NERVOUS SYSTEM Despite this qualification, it seems fairly obvious that pioneer fibers often do take di- rective courses toward growing organs, which by their very growth activity have become hubs of pathway systems. It is in line with this view that nerves can be made to con- verge upon a transplanted limb bud only if the operation is performed prior to their first outgrowth; once established, their course can no longer be redirected (Detwiler, ’24b). Intracentral Connections. There are hardly any systematic investigations on the manner in which the different intracentral nerve tracts are laid down, the routing of which is of such paramount importance to orderly function. In attempting a causal analysis, it is well to keep in mind two basic factors. First, there is no microprecision, in the sense of rigidly determined connection patterns, on the level of the individual neuron. Only the gross group characters of the various tracts are determined, while the details are merely statistically defined, the individual elements conforming to some “norm” but being otherwise indeterminate. This consider- ably reduces the number of relevant patterns to be accounted for. Second, the incredibly complex structure of the adult brain owes its intricacy to the fact that it is the com- pound result of innumerable elementary pat- terns laid down one after another in a long succession of separate ontogenetic steps, each one in itself rather simple. If we assume that intracentral fibers, like peripheral fibers, trace oriented pathways in their colloidal surroundings, then each pathway system that temporarily dominates a given embryonic period will leave a permanent record behind in those neuron systems which happen to grow out during that period. As conditions change, the colloidal matrix will adapt to the change, assume new orientation and thus establish new nerve courses, often unrelated geometrically to the earlier ones. However, actually to resolve central nerve patterns into such simple constituent steps is still largely a task for the future. That it promises success is indicated, for instance, by the observation that fiber tracts tend to develop between central neuron groups that develop contemporaneously (Cog- hill, ’29). This looks like the ‘“two-center effect” at work again. Reasonable guesses are also possible regarding the pathway struc- tures underlying the longitudinal fiber tracts, the commissural fiber systems and the early internuncial connections of the cord. The longitudinal tracts appear to be determined by the longitudinal stretch to which the 359 neural tube is subjected by the growth in length of the surrounding body (see below). This assumption finds support in the early appearance of axial birefringence in the neural tube (Hobson, *41), indicating polar orientation of the substratum. The arched transverse pattern of the ventral commissures might be attributable to transverse stresses which the median strip of the medullary plate suffers during the bending of the plate into the tube. The early internuncial fibers travel along an interface clearly demarcated WAT HI | i Mh HH Ih Fig. 134. Deflection and recombination of nerve fibers (plexus formation) along intersecting sys- tems of ultrastructural pathways. between the dense neural epithelium sur- rounding the central canal and the more loosely packed mantle zone. All these preneural guide structures, how- ever, can only account for the initial orienta- tion of nerve patterns. Other sets of factors determine their further elaboration with re- gard to numbers, size, and connections (see below). Deflection and Plexus Formation. Many nerve courses show a sudden angular de- flection from one direction into another. This may be the result of a passive distortion— for instance, by sudden change of course of a towing organ or by wedging in of another organ—or it may be a sign of actual angular outgrowth. The latter condition is realized whenever pioneering fiber tips proceeding along one pathway system come upon another one running crosswise (Fig. 134). Depending on how completely the intersecting system obliterates the original system, fewer or more of the tips will be diverted into the new direction; a tight barrier (e.g., membrane) will produce total deflection. Intersecting structures of this kind arise, for instance, in 360 the border zone between fibrous colloids of ditferent concentrations, and model experi- ments in tissue culture have veritied the tangential deltlection of radial nerve fiber growth along such borders (Weiss, ’54a). Similar cross patterns may arise in the body at the boundary between masses or layers of cells of different kinds which exert ditfer- ent effects on the surrounding colloids. Many other processes that would lead to the same end are imaginable, all of them thus tar untested. Striking examples of angular deflection in the embryo are the dorsal roots of the cord, the various central and peripheral plexuses, and the partially decussating systems in the brain. Dorsal root fibers, after entering the cord laterally, turn abruptly into a longi- tudinal course (with or without branching), forming thus the dorsal funiculi. Evidently the switch is produced by the encounter with longitudinal pathways in the marginal veil oriented lengthwise by the passive elongation of the tube mentioned before (p. 359). Plexus formation is to be expected wherever layers with predominantly radial structure alter- nate with tangentially oriented ones, as in the retina or in the various strata of the cortex (see Fig. 134). Thus the horizontal interconnections among the vertical projec- tion systems, which are such an important functional feature, are presumably antici- pated by lamination in the texture of the ground substance, which in turn might be due either to the differential growth expan- sion of the various cell strata or to differential impregnation of the ground substance from different concentric cell layers (see Weiss, ’39, p. 509). Peripheral nerve plexuses are probably likewise caused by “crossroads”; the brachial and pelvic plexuses, for instance, by the tangentially disposed girdle mesen- chyme which lies across the nerve paths radiating toward the limb base. A cross structure in the optic chiasma, whose angle of intersection changes with the relative shifts between eyes and brain, could account for the ipsilateral deflection of optic fibers in forms with partial decussation, with the probability of diversion, hence the proportion of uncrossed fibers (important in binocular vision) being perhaps a function of the chiasmatic angle during the growth phase. Hypothetical though many of these de- tailed applications of the principle of contact guidance to concrete embryological problems may be, they at least formulate the problems involved for practical experimental attack. For final judgment, the results of the latter SPECIAL VERTEBRATE ORGANOGENESIS must be awaited. At any rate, it appears clear that plexus formation as such, with fibers turning off their former course at a sharp angle and intermingling in a common plane, otten to emerge again later as inde- pendent bundles, defeats any but a structural concept of guidance. Branching. Individual neurons may remain essentially unbranched, as in many sensory types, or they may branch more or less pro- fusely, as do the motor neurons. Since, ac- cording to the all-or-none principle, the neuron can only act as a unit, the extent of branching has great functional significance and must be either preadapted to, or actively regulated by, functional needs. Extensive branching, economical in the motor field where it enables a single neuron to engage several hundred muscle fibers at a time, would be undesirable in the sensory field, where it would blur discrimination. Despite its biological importance, however, the prob- lem of branching has not yet been system- atically studied (Sunderland and Lavarack, "53)': Branching occurs either at the tip of a growing fiber by dichotomy (terminal branching) or along the stem some distance behind the tip which is either still free or already connected (collateral branching). Terminal branching results whenever two simultaneous terminal pseudopods (see Fig. 127e) are of equal strength so that they can divide the inflow of protoplasm between themselves and continue to advance with in- dependent tips (see Speidel, ’33). The fre- quency of this occurrence depends on the structure of the pathway system; the more intersected the latter, the higher the inci- dence of branching (Fig. 129). Accordingly, terminal branching is profuse in the maze of central neuropil, as well as in peripheral scar tissue (e.g., after nerve severance), but is infrequent along well-oriented pathways. Collateral branches arise as side sprouts from already consolidated fiber stems, pre- sumably in response to local irritations, mechanical, chemical or electrical (Peterfi and Kapel, ’28; Speidel, ’33; Edds, ’53). The repeated branching of motor fibers, for in- stance, could be ascribed to a seriation of such irritations as would attend consecutive divi- sions of young muscle fibers. The size of a “motor unit” (number of muscle fibers at- tached to a single neuron) would then simply reflect the degree of ulterior growth of that muscle after receiving its primary quota of fibers. The systematic occurrence of similar irritations near certain layers or nuclei of the NERVOUS SYSTEM brain and spinal cord could account for the regular emergence of collaterals at those sites. Possibly different fiber types might even react in different degrees to the same irritant so that one type would give off a branch while another type would not. These are merely some pointers to future work. Concrete information is scanty. In view of the labile state of the neuron (see later), we must expect it to spring minor leaks in its surface all the time, es- pecially in its unsheathed terminal branches and at the nodes of Ranvier. Whether such weak spots will be repaired or become the source of a collateral branch will presum- ably depend on the vigor and rate of growth of the neuron and the competitive strength of the main stem of the fiber, which counter- acts accessory outgrowths. Agents capable of either loosening the axonal surface or in- vigorating neuronal growth or merely re- ducing the drain into the main axis of the fiber (as after amputation) should, there- fore, automatically increase the frequency of collateral branching. The compensatory sprouting of peripheral collaterals after par- tial denervation of muscles or after injection of substances from degenerated muscles (Hoff- man, 50; Edds, 53), as well as the “para- sitic’ branches forming from severed or otherwise truncated neurons (Nageotte, ’22), seem to bear out this expectation. Nerve Patterns Within Peripheral Organs. Usually, the factors that guide nerves to a given organ are not the same that will deter- mine the distribution pattern within the or- gan (Hamburger, ’29). For instance, when limb buds are transplanted to the head region and innervated by foreign cranial nerves, the latter assume a distribution pat- tern which is essentially a typical limb pattern. It is from this very observation that Harrison (’07b) first deduced the structural guidance of nerve fibers, a conclusion which was soon also adopted by Braus (711), who had previously interpreted his own similar observations as evidence of a_ peripheral (autonomous) origin of nerves. Evidently, the growing limb can impose a limb-specific arrangement upon nerve fibers coming from whatever source (see also Hamburger, ’29; Piatt, ’41; Weiss, ’37a). This plainly contra- dicts the gratuitous contention (Ruud, 729) that nerves from a given source contain the geometry of their future distribution in themselves. Rather, the “limb pattern” of distribution is determined by a _ complex, multifactorial chain of events, roughly divis- ible into two phases—a primary one governed 361 by the structural and chemical properties of the preneural pathways in the limb blastema, and a secondary one of elaboration of the primary pattern by towing, fasciculation (see below, p. 366), and the differential sur- vival, growth, and resorption of fibers, de- pending on the physiological adequacy of their terminal connections. The decisive patterning effect of the primary phase is revealed by the observation that virtually all nerve branches of the mature limb (in the frog) are already recognizable as such in the early limb bud at a very primitive stage of morphogenesis (Taylor, 43). The principal nerve paths are thus laid down by factors in the early limb blastema. If, on the other hand, a limb is kept nerve- less (“aneurogenic”) during its differentia- tion and is then grafted to a normal host body from which it can derive belated in- nervation, the invading nerves follow quite irregular and aberrant courses (Piatt, ’52). A certain predilection for some major in- vasion routes at times creates some gross resemblance to a limb pattern (Piatt, ’42), but this could be due simply to trivial ana- tomical features, offering only limited spaces between skin, muscles and skeleton for the massive advance of nerve fibers. In this in- stance, major blood vessels may also play a leading role (Hamburger, ’29), although in normal development the noted parallelism between vascular and nerve trunks is more likely to be a sign of common guidance of both systems by the same _ ultrastructural pattern in the common matrix. Whether the nerve distribution pattern within the limb follows normal or aberrant lines is of no consequence, however, as far as the later functional activity is concerned. As will be described below (p. 384), functional coordina- tion between the central nervous system and receptor and effector organs remains orderly even if the anatomical nerve connections are utterly confused. The relative stereotypism of peripheral nerves is presumably significant only as a means of insuring ubiquitous in- nervation of adequate quantity. PERIPHERAL CONNECTIONS Specificity of Preneural Pathways. Nerve fibers of a given kind can penetrate foreign organs with ease. After heterotopic trans- plantations or other deviations, cranial nerves have been followed into limbs (see above; Harrison, ’07b; Braus, ’11; Nicholas, ’33; Piatt, 41), midbrain fibers into trunk muscles (Hoadley, ’25) in the chick (not observed 362 in comparable experiments in urodeles: Det- wiler, ’36a), spinal fiber tracts into limbs (Nicholas, ’29; Weiss, 50a), limb nerves into tumors (Bueker, *48; Levi-Montalcini and Hamburger, °51), optic nerves into the nose (Weiss, *41c) or the pharynx (Ferreira- Berutti, °51). Yet, the peripheral nervous system of normal individuals is relatively stereotyped, not only in the mode of its arborization but also in its terminal connections. By and large, ventral root fibers end in skeletal muscles, spinal ganglion fibers in sensory end organs, sympathetic fibers in glands or smooth muscles. Since the attempts to refer this specificity of connections to selective neurotropic attractions have proved unten- able, other explanations must be sought. Mindful of the fact that the first outgrowth of motor fibers antedates that of sensory fibers (Coghill, ’29), it has been suggested (Harrison, ’35b; Weiss, ’39) that a systematic change in the peripheral pathway structure, with the earlier pathways leading to muscles, the later ones to skin, would automatically account for the correct routing. However, this time-lag explanation is ruled out by the observation that motor and sensory fibers take each their typical courses even when both grow out simultaneously, as for in- stance, in the innervation of the hind limb of the anuran tadpole. When the bud makes its late appearance, both motor and sensory nerve masses are already waiting at its base ready to invade it (Taylor, 43). As they penetrate the limb bud, they assort them- selves according to kinds into specifically muscular and cutaneous branches, respec- tively, coursing sometimes jointly, but often also independently. This has been revealed by withholding either the sensory or the motor nerve quota from the limb (Ham- burger, °29), and most conclusively by ex- tirpating the appropriate spinal ganglia or spinal cord segments (Taylor, °44); the developed limbs then lacked the correspond- ing kind of nerve branches. It seems difficult to account for these facts otherwise than by the assumption of selec- tive contact affinities of given nerve fiber types for matching types of preneural path- ways. This view is strengthened by the pre- dilection which nerves with aberrant origins or courses show for their typical sites or channels, as has been described for the lateral line nerve (Harrison, ’03), the dorsal roots (Detwiler and Maclean, ’40; Holtzer, ’62b), and Mauthner’s fibers (Oppenheimer, "41; Piatt, 44; Stefanelli, 50; Holtzer, ’52b). Such selective application of one tissue to SPECIAL VERTEBRATE ORGANOGENESIS another is not uncommon in development (see, for instance, the guided growth of the pronephric duct; Holtfreter, ’44); but, save for a hypothetical reference to its possible stereochemical basis (Weiss, ’47), the under- lying mechanism is still obscure. Specificity in Regeneration. The ability of given kinds of nerve fibers to select con- forming pathways seems to last beyond the pioneering phase. After transection of the mixed nerves to a young differentiating limb, the regenerating motor fibers retrace essen- tially the original muscular branches, and the regenerating sensory fibers the cutaneous branches (Taylor, ’44). Similarly, regenerat- ing lateral line nerves have been reported to give preference to an old lateral line branch over a nearby cutaneous branch (Speidel, ’48), which even suggests finer subspecificities within the sensory class. Yet, with the prog- ress of maturation, this selectivity of out- growth is lost. Adult nerves of different qualities, when cross-connected, regenerate into each other’s channels indiscriminately and without difficulty. For example, sensory fibers regenerate into motor stumps (Boeke, 17; Gutmann, °45; Weiss and Edds, ’45) and vice versa (Weiss and Cummings, ’43), somatic nerves into sympathetic stumps and vice versa (Simpson and Young, °45; Ham- mond and Hinsey, *45), etc. Apparently, the residual Schwann cords of degenerated stumps, which serve as pathways to the re- generating fibers, are of the same quality in motor, sensory, somatic and autonomic nerves, hence are indistinguishable to the regenerating fibers, which, as will be shown below, have not lost their constitutional dif- ferentials. Terminal Connections. Nerve fibers may reach the peripheral tissues either preassorted over proper pathways or intermingled over aberrant routes (see above), but neither mode of approach is decisive for whether or not they will make transmissive connections, that is, connections which will permit im- pulses to pass between nerve fiber and end organ. A clear distinction must be made be- tween (a) penetration of a tissue by nerve fibers (‘“‘neurotization”), (b) microscopic contiguity between nerve ending and effector or receptor cell, and (c) physiologically effec- tive junction. No absolute specificity prevails in (a) and (b), but (c) occurs only if end- organ and nerve fiber are generally related. For example, when sensory nerve fibers are led into muscles, they terminate on the muscle fibers in what histologically appear to be intimate motor connections (Boeke, 17); yet electric stimulation of such nerves NERVOUS SYSTEM never yields muscular contraction (Gutmann, °45; Weiss and Edds, °45). Cross unions between somatic, sympathetic and parasym- pathetic nerves that have been tried in various combinations likewise are physio- logically sterile. Cholinergic and adrenergic nerves fail to achieve physiological innerva- tion of each other’s peripheries (Langley and Anderson, °04; Dale, ’35), not because of lack of regenerative penetration, but because of transmissive failure of the terminal junc- tion. The fact of neuro-terminal selectivity is proof of the existence of specific protoplasmic differences both among the major classes of neurons (sensory, motor, etc.) and among the corresponding terminal tissues. Within each class, however, transmissive junctions can be made indiscriminately. Any motor nerve shows functional affinity to any skeletal muscle (Weiss, ’37a; Weiss and Hoag, ’46) ; cross connections of different kinds of sensory nerves have likewise been effected success- fully (Anokhin, *35), and the paradoxical sensations noted after irregular sensory nerve regeneration in man (Stopford, °30) also indicate interchangeability within the sen- sory field. There are additional finer func- tional selectivities, beyond those controlling junction, but these are imposed upon the connected neurons from their endings and will be discussed later (see p. 384). Synaptic Connections. Naturally, the ques- tion arises whether specificities similar to those observed in peripheral connections govern the establishment of central synaptic junctions. In the few instances thus far ex- amined, intracentral neurons have shown a remarkable lack of discrimination in making terminal connections. Limb buds inserted into gaps of the embryonic neural tube receive effective motor innervation from central fiber tracts (Nicholas, ’33), and the central gray matter of isolated fragments of larval spinal cord or medulla oblongata establishes fully functional connections with both muscles and skin in the complete ab- sence of primary motor and sensory neurons (Weiss, 50a). The promiscuity of junctional relations manifested in these cases contrasts sharply with the acute selectivity of func- tional response relations (see below, p. 384), which throws doubt upon any theory ex- plaining the latter purely in terms of specific anatomical connections. The fact that a junc- tion capable of transmitting an impulse has been established does not explain when and how it will be actuated in the coordinated group activities of the centers, nor indeed whether or not it will be used at all. Whether 363 disuse entails eventual rupture of junctions has not yet been clearly decided. AXON GROWTH Growth of Axon Caliber. While the elonga- tion of the nerve fiber is essentially a phe- nomenon of protoplasmic convection, it proceeds pari passu with real growth, that is, increase of the total protoplasmic mass of the neuron, and is actually sustained by the latter. This growth process continues after the fiber has reached its final length and can, in fact, best be studied during that later period, when all further protoplasmic gain accrues solely to the width of the fiber. Since the eventual caliber of the axon, usu- ally referred to as “fiber size,” is of con- siderable functional significance, as it deter- mines velocity of impulse conduction, thres- holds of excitability and susceptibility to noxious agents, etc. (Erlanger and Gasser, 37), a study of axonal growth offers both physiological and developmental interest. Nerve fiber caliber increases as animals grow to mature size (Hursh, ’39). Analytical information on axon growth is mostly derived from recent experiments dealing with the restoration of fiber diameter in regenerated nerve fibers (Weiss and Hiscoe, 48). These experiments are schemati- cally summarized in Figure 135, which shows a series of mature neurons in various stages of normal (A-E) and modified (F-I) re- generation. The nucleated cell body is at the left, the peripheral end-organ at the right, both connected by the neurilemmal tube that envelops the fiber. From the proximal stump of the severed fiber (B), a thin axonal sprout advances toward the periphery (C), effects peripheral connection (D) and gradually grows in width until it approximates its old caliber and the width of the tube (£). This recovery, however, can be significantly impeded if one constricts the distal stump and thereby reduces the diameter of all tubes at a given spot. At first, regeneration proceeds normally (compare D and F), but as soon as the axon has reached the girth at which it fills the narrow (constricted) part of the tube, the portion lying distally to the con- striction ceases to gain in width, while at the proximal side of the constriction excess axoplasm begins to pile up in configurations such as are ordinarily assumed by a steadily propelled column of plastic material sud- denly faced with an obstruction ( G, H). If, later, the constriction is released allowing the tube to re-expand, the dammed-up mate- 364 rial moves on peripherad, thus widening the formerly stunted distal portion (J). The rate of this movement was estimated to be of the order of a few millimeters per day, which corresponds closely to the optimal rate of free advance in regenerating fibers (see p. 356). These results have led to the conclusions that (1) axoplasm is synthesized solely in the SPECIAL VERTEBRATE ORGANOGENESIS terminal swellings of blocked regenerating nerve fibers (see Cajal, 728; Nageotte, ’22). Although we know nothing about the nature of the axonal pumping mechanism,* it is reasonable to assume that it is the same for first outgrowth (see p. 350) and regenera- tion. While the fiber tip advances, the ma- terial is used for elongation; after the fiber has ceased to elongate, the continuing sup- Fig. 135. Damming of axoplasm in constricted nerve fibers. A-E, consecutive stages of unimpeded regener- ation; F¥—H, consecutive stages of regeneration with “bottleneck”; 7, following H after release of constriction. (From Weiss and Hiscoe, *48.) central cell body near the nucleus; (2) axo- plasm is conveyed peripherad in a steady movement accommodated to the width of the tube which serves as channel; (3) axo- plasm is subject to continuous catabolic deg- radation all along the fiber. Accordingly, any local reduction of the width of the chan- nel throttles downward flow, hence reduces the rate of replacement of the “downstream” portion, while excess material accumulates on the “upstream” side. Thus is visualized directly what used to be postulated by earlier students of nerve growth as “vis a tergo” (Held, ’09) or “formative turgor” (Cajal, 28). Damming of axoplasm can now be taken as a direct sign of obstructed axonal transport. As such it is seen, for instance, in the ply adds to its width until a steady state is reached between rate of supply and rate of catabolic consumption. Since there is evi- dence (Weiss and Hiscoe, *48) that this centrifugal supply stream continues through- out the life of the mature neuron, nerve re- generation turns out to be but a special manifestation of a perpetual growth process. This explains why nerves can regenerate repeatedly in succession with undiminished vigor (Duncan and Jarvis, ’43). *If the rhythmic pulsations demonstrated for central glia cells (Pomerat, 51) were also a prop- erty of peripheral Schwann cells and if these were coordinated in the manner of heart muscle contrac- tions or ciliary beats, this might offer a mechanism for the massaging of axoplasm downward within its sheath. NERVOUS SYSTEM The caliber of a nerve fiber is thus deter- mined essentially by three factors: (1) the amount of synthesis of new axoplasm in the cell body; (2) the rate of its centrifugal movement; and (3) the rate of its peripheral breakdown. Since‘ the rate of movement is limited by the width of the channel in which it occurs, large nerve fibers regenerating into narrower tubes fail to gain full normal width (Holmes and Young, 7°42; Sanders and Young, ’44; Simpson and Young, ’45; Ham- mond and Hinsey, °45). However, no such limitation is to be expected during embryonic growth, before firm neurilemmal tubes have formed. Assuming, furthermore, rather uni- form rates of catabolism, we are left with the rate of central synthesis as the main variable in the determination of fiber caliber. which, in turn, is rather closely correlated with the size of the nerve cell body. Factors Controlling Neuron Growth. Early in development different neuron groups seem to acquire constitutional growth differentials which place them in different size classes. Within each class itself, however, growth rate and size are subject to further modifica- tions which are due to extraneous conditions, as illustrated in the following. When a nerve fiber is severed, hence dis- connected from its terminal organ, the whole neuron begins to atrophy (Weiss, Edds and Cavanaugh, °45; Sanders and Young, °46; Aitken, Sharman and Young, 747). If we disregard certain acute traumatic changes (“ascending degeneration,” ‘axon reaction” ) referable to the injury as such, the main long-range effect of the loss consists of a progressive reduction of the dimensions of the neuron, beginning with the nucleolus and spreading to the nucleus, the cell body, and finally the diameter of the axon (Cavanaugh, 51). Conversely, upon reconnection with a peripheral organ, the dimensions enlarge again. Moreover, when a neuron is “over- loaded,” that is, made to innervate a larger volume of peripheral tissue than originally (e.g., by collateral branching into a dener- vated field), nucleus and cell body enlarge above their normal dimensions (Terni, ’20; Edds, ’49: Cavanaugh, ’51). The production center of neuronal synthesis thus adapts itself sensitively to the demands of the peripheral innervation volume. Variations of functional activity have simi- lar effects. Not only do neurons tend to atrophy, when they are chronically de- prived of excitation (e.g., Edds, ’51), but they hypertrophy, again starting frorn the nucleus, in response to intensified physio- 365 logical demands (Hamberger and Hydeén, 49). Enlargement of the nuclei and nucleoli of the cells in certain brain centers following induced hyperfunction of these centers (e.g., antidiuretic center in the hypothalamus) has also been recorded (Ortmann, 51). The interpretation of all these facts requires cau- tion, because nuclear enlargerment may sig- nify either true protoplasmic growth gener- ally associated with increase in desoxyribo- nucleic acids (Hydén, ’50), or merely water uptake, accumulation of functional products and other transitory changes subserving functional activity rather than true growth (see Leuchtenberger and Schrader, ’52). At any rate, the realization that the size of a neuron is subject to continual upward or downward regulations in accordance with extraneous influences received from both its afferent and (ascendingly) efferent ends, certainly controverts the classic view which has endowed the central nervous system, at least in morphological regards, with a nim- bus of static and rigid fixity. The significance of this demonstrated plas- ticity for our concepts of central activity is evident. The backward projection of periph- eral conditions into the centers, which we encounter here for the first time in our discussion and which will be amplified be- low, is particularly noteworthy. This makes it all the more important to stress the fact that the reported influences do not ‘‘deter- mine” activity, growth, or size, of a neuron in any absolute sense, but simply enhance or depress its inherent activities, and within relatively narrow limits at that (for in- stance, even the chronically disconnected ganglion cell still retains about 60 per cent of its normal mass). Biochemistry of Neuron Growth. The morphological evidence for the localized growth of the neuron from the nuclear ter- ritory (Weiss and Hiscoe, 48) is supported by the cytochemical demonstration of abun- dance of desoxyribonucleic acids and a high rate of protein synthesis in the nuclear territory (Caspersson, 50; Hydén, ’50), with fluctuations that closely parallel the growth activity of the neuron. Thus, the intensity of these syntheses increases during the re- generative phase following nerve section (Bodian, *47), during the growth reaction resultine from hyneractivity (Hambereer and Hydén, ’45) and, embryonically, during the phase of the functional alerting of brain cells (Flexner, 750). In agreement with this monopoly of the nuclear territory as the growth center of 366 the neuron, the well-known impairment of nerve growth in thiamine deficiency (beri- beri) is observed only if the cell bodies are bathed by the deficient medium (in vitro), even though the rest of the nerve, including the growing tips, lies in normal medium (Burt, ’43a). The correlation of cytological, biochemi- cal and physiological data for normal and deficiency states of the nervous system is making encouraging progress (see p. 376). One of the major tasks in these studies which remains is to distinguish clearly between the relative contributions of the metabolic machinery of the neuron to its growth (i.e., protoplasmic reproduction) on the one hand, and to its functional activity on the other; a task rendered more difficult by the fact just outlined, that having reached a sta- tionary size, a neuron keeps “growing” with- out change of total mass. DEVELOPMENT OF GROUP RELATIONS Fasciculation. Outside the gray matter and the terminal nerve fibers arborizations, SPECIAL VERTEBRATE ORGANOGENESIS of pioneering fibers, the routing is achieved not by “attraction” but by a form of con- tact guidance. When a nerve source and a peripheral organ, e.g., a limb, are implanted at some distance from each other in a loose connective tissue bed (e'g., the dorsal fin of an amphibian larva), a strong nerve cable soon develops between them (Fig. 136) in the following manner (Weiss, *50a). Pio- neering fibers from the nerve center invade the surroundings. Those that happen upon the limb and succeed in connecting with its tissues thereby become somehow adhesive for other nerve fibers growing out subse- quently; older fibers thus become guides to the limb for younger ones, endowing them, in turn, with adhesiveness, and so forth, in a sort of chain reaction. It is not known what eventually terminates the agglomeration; perhaps fasciculation ceases automatically as soon as peripheral saturation (see p. 368) has been reached, that is, when further newcomer fibers no longer find functional attachments, hence, do not become adhesive. The lessons of nerve anatomy which show that nerve fibers of identical functional Fig. 136. Fasciculation. Nerve cable that has formed between a fragment of spinal cord (left) and a limb (right), both deplanted to the dorsal fin of a urodele larva (frontal section, showing the loose connective tissue of the fin between the two borders of epidermis). (From Weiss, ’50a.) rarely appear solitary. They commonly as- sociate with other nerve fibers to form bun- dles, as well as with the Schwann cells and connective tissue cells. The grouping of fibers into bundles seems to come about in two ways: (a) by a primary tendency of younger fibers to follow the course of older fibers in close application of one to the other (“fasciculation”); and (b) by the secondary gathering of small groups into larger assemblies through the formation of common connective tissue sheaths. Primary fasciculation is not a chance event but a systematic device to route a sufficient number of nerve connections toward a desti- nation in need of them. As in the outgrowth properties (e.g., fibers subserving pain or special sensations) often course together, both in peripheral nerve trunks and central funiculi, indicate an element of selectivity in the process of fasciculation. Young fibers of a given kind would apply themselves preferentially to older fibers belonging to the same, rather than to another, category. Some experimental proof for “selective fasci- culation” by contact affinities may be seen in the observation that supernumerary Mauth- ner’s fibers of the amphibian brain run along- side the normal Mauthner’s fibers of that side much more frequently than could be expected on mere chance; and in the further observa- tion that transected longitudinal fiber tracts NERVOUS SYSTEM of the spinal cord, when regenerating through a connective tissue gap, even over great dis- tances, tend to retain their fascicular iden- tity (Hooker, ’30; Holtzer, ’52b). Evidently, such selectivity of association imples two things: first, that nerve fibers of different functional designations are constitutionally (i.e., substantially) different, and second, that they can recognize and distinguish one another according to kind. Fasciculation occurs by non-detachment among fibers that have made contact, rather than by active association. It therefore is promoted by conditions furthering the chances of contact. These chances are low in compact tissues containing innumerable separate fibrous guide lines, but high in tissues with large liquid spaces, where the nerve pathways are crowded into the rela- tively few land bridges. Judging from tis- sue cultures, established nerve fibers may themselves bring about the latter condition by liquefying surrounding colloids (Weiss, 34a) and thus facilitate bundling. Moreover, liquefaction around attached nerve fibers will cause their floating stems to cling to- gether, according to the same principle that makes wet threads stick together in air, thus assembling them secondarily into trunks. These considerations furnish a ready expla- nation of the fact that fibers tend to remain separate in the gray matter, in the periph- eral tissues and in nerve scars, whereas they are aggregated into bundles in the more liquid-filled interstices, particularly along the blood vessels. However, much remains still to be found out about the mechanisms of fasciculation; particularly the factors ef- fecting the sorting and collecting of fascicles into still larger assemblies by wrappings of connective tissue are still wholly unexplored. Association with Sheath Cells; Myeliniza- tion. The specific affinity between Schwann cells and axons appears to be mutual. Sheath cells attach themselves to nerve sprouts (Speidel ’32). The tips of nerve fibers, con- versely, show a definite predilection for strands of Schwann cells (Nageotte, ’22; Dustin, *17), and it is doubtful whether naked sprouts, not enveloped by sheath cells, can persist over appreciable distances. The two cell types, when in contact, thus seem to form firm unions, which actively resist separation (Abercrombie, Johnson and Thomas, *49). Sheath cells have been seen to shuttle freely between nerve fibers of different kinds (Speidel, ’33), with a certain preference for transfer from unmyelinated to myelinated ones (Speidel, 50). Sheath cells, 367 however, do not share the specificity of the nerve fibers which they coat; this can be inferred from the fact that regenerating sensory or motor nerve sprouts penetrate Schwann cords of either nerve type with equal ease (see p. 362). Myelin is presumably formed in the sur- face of the axon, with the sheath cells (or, in the central nervous system, the glia cells) furnishing some essential stimulus or com- plement (Speidel, ’33, °35), but the details of the process are not known. Since the same sheath cell has been observed either to induce, or fail to induce, myelin formation, depend- ing on whether it joined the branch of a potentially myelinated or a potentially un- medullated fiber (Speidel, ’33), the differ- ential faculty for myelinization must be a property of the nerve fiber itself. The lami- nated structure of the myelin sheath (see p. 347) and the proportionality between its thickness and the caliber of the axon suggest that successive layers are shed by the sur- face as the axon grows in width (see Geren and Raskind, ’53). As Schwann cells line up in tandem along the axons, myelin formation progresses in a general proximodistal sequence, starting in each cell from the region of the nucleus. The length of the cell defines a myelin segment, and the junction of two cells, the node of Ranvier. The standard length of individual segments (internodes) amounts to about 300 micra, which agrees with the average length of an extended Schwann cell in tissue culture. The same average length is found in regenerated nerve fibers regard- less of diameter (Hiscoe, 47; Vizoso and Young, °48). The greater internodal length in primary (non-regenerated) fibers, which varies directly with fiber diameter attain- ing up to about 1500 micra, is presumably to be ascribed to passive elongation by the stretching of the nerves in tow of growing organs (Hiscoe, 47; Young, ’50). The longest fibers having undergone the greatest ex- tension would also end up with the longest segments, and since the longest fibers (within the same class) have the largest caliber (Kolliker, 1896), a general propor- tionality between fiber diameter and inter- nodal length would result (Thomas and Young, ’49), which fact has recently assumed added significance in connection with the saltatory theory of nerve conduction. Saturation Factors. Volume and density of peripheral innervation vary relevantly from organ to organ and from region to region within the same systems (skin, intestine, 368 connective tissue). This characteristic dis- tribution is grossly anticipated in the rela- tive allocations of neurons to different pe- ripheral sectors (see below, p. 374), in the frequency of preterminal branching and in the peripheral control of fasciculation. Since this rough preallocation would still leave a wide margin of variability in the number of fibers actually arriving in a tissue in a given individual, there are ad- ditional screening factors in operation in the terminal tissues themselves which adjust the final quota of terminals to a stable norm. Each tissue would thus maintain a char- acteristic ‘saturation’ density of innerva- tion. While not much is known about the means by which this control is exerted, it is already becoming obvious that they are different for different tissues, and that the mechanisms for upward regulation from a deficient source are of a different kind than those involved in the downward regulation from an excessive source. They will there- fore be considered separately. When supernumerary (Detwiler, ’36b) or excessively large (Harrison, 35a) limbs are transplanted to the limb region of urodele embryos, the enlarged periphery, while causing no adaptive increase in the central nerve source (see below, p. 382), yet de- rives from that undersized source the full contingent of nerve fiber branches appro- priate to its larger mass. Similarly, limbs transplanted in later larval stages, and provided with only a small branch for re- generative innervation, contain eventually the full contingent of fibers normal for a limb (Weiss, ’37a). In nerve regeneration in the adult, likewise, the number of re- generated fibers in the distal stumps approxi- mates the normal quota, even if the prox- imal source of fiber stems has been dras- tically reduced (Dogliotti, °35; Litwiller, 38a; Weiss and Campbell, °44; Billig, van Harreveld and Wiersma, ’46). This compensatory increase of the volume of innervation is due to more extensive peripheral branching of the individual neu- rons. Being the rule in the regeneration of transected nerves (see above, p. 350), such branching is satisfactorily accounted for in the cases just mentioned in which nerves have been severed. However, compensatory branches may also arise, as if in response to peripheral needs, from nerves that have not been deliberately traumatized. When musculature is partially denervated by the experimental elimination of part of its tribu- tary ventral roots, the residual healthy intra- SPECIAL VERTEBRATE ORGANOGENESIS muscular fibers develop “spontaneously” col- lateral branches that take over the innerva- tion of the neighboring denervated muscle elements (see Edds, ’53). The fact that such collateral branching can be artificially in- duced by intramuscular injection of extracts from degenerating muscle (Hoffman, °52) suggests an active participation of the de- nervated muscle fibers in tapping the locally available nerve sources for extra branches. As indicated in our discussion of branching mechanisms (p. 361), this need imply noth- ing more than the weakening of the axonal surface, particularly at the nodes, to permit ever-present abortive axonal leaks to yield durable offshoots. Perhaps substances re- ported to enhance nerve regeneration (von Muralt, 46) should be viewed in the same light, as facilitating the protrusion of branches or diminishing the resistance of tissues to penetration by them. Compensatory collateral innervation, com- parable to that observed in muscle, has also been described to occur in denervated sec- tors of skin, whose infiltration by side branches from nerves of the surrounding intact area has been either observed di- rectly in experimental cases (Weddell, Gutt- mann and Gutmann, *41) or deduced from clinical results (Livingston, °43). Again, the explanation may lie either in a true activation of branching by emanations from the denervated tissue or, conversely, in the removal of an active suppressing principle that could be assumed to emanate from nerve-saturated tissue as a bar to its invasion by nerve branches continually forming as a result of intra-axonal growth pressure and surface instability. The latter alternative seems plausible since the assumed abortive branching has actually been observed in living preparations (Speidel, 42). Negative microscopic evidence, on the other hand, would be meaningless in view of the elec- tronmicroscopic demonstration of fiber col- laterals far below the range of microscopic visibility (Fernandez-Moran, 52). In conclusion, if overproduction of branches is a regular occurrence, it would account for a reservoir of fibers sufficiently large to supply the needs of even a consid- erably enlarged periphery; on the other hand, it may be necessary in addition to assume peripheral factors that facilitate, if not the production, at least the further out- growth and consolidation, of branches in accordance with the size of the periphery to be innervated. Either mechanism insures to the peripheral tissue a full quota of NERVOUS SYSTEM innervation even from an undersized nerve fiber source. Equally important, however, is the upper limitation set to nerve density, referred to above as “saturation.” Even in the face of a superabundant supply of nerve fibers, the tissue restricts the admission of terminals to its normal quota. The best-studied ex- amples are muscle, peripheral nerve, and regenerating limbs. Individual muscle fibers rarely contain more than one ending of nerve fibers of the same kind. That this em- bodies an active self-protection of the muscle fiber against multiple innervation (‘‘hyper- 369 Of different causation, but of comparable effect, is the peripheral control of nerve fiber numbers in nerve regeneration. Irre- spective of the size of the nerve fiber source and even in the presence of an excessive amount of branches produced, the number contained in the regenerated distal stumps approximates the normal number closely enough to intimate active regulation (e.g., Davenport, Chor and Dolkart, ’37; Weiss, 37a; Weiss and Campbell, °44; Litwiller, 38a; Weiss and Cummings, ’43). Since new fibers course both inside and between old nerve tubes (Holmes and Young, °42), the NERVE SOURGE NORMAL REDUCED NORMAL NORMAL PIERIRIRE RY. NORMAL Fig. 137. Diagram summarizing the regulation of density of peripheral innervation in instances of abnormal ratios of nerve source to peripheral field. neurotization”), comparable to the self-pro- tection of fertilized eggs against multiple insemination (Harrison, *10), is evidenced by the fact that it is impossible to force appreciable surplus innervation upon a mus- cle even by inserting an excessive supply of nerve fibers right into it (Elsberg, 17; Fort, *40; Weiss and Hoag, ’46). Only unin- nervated muscle fibers seem to be ready to accept nerve endings, but once a connection has been effected, the muscle fiber shields itself somehow from further impregnation. Since this physiological insulation requires some time to develop, there exists, of course, an open interim during which additional endings can take. This explains why there is always a certain percentage of muscle fibers found with multiple endings, and why the incidence of “hyperneurotization” is higher when superabundant nerve masses are allowed to pervade the muscle simul- taneously (Hoffmann, °51). The nature of the protective reaction is unknown, but may be looked for in a change of surface prop- erties. capacity of the latter cannot be the limiting factor. Indeed, the numerical restriction is observed even if the peripheral nerve stumps have been evulsed (Litwiller, ’38a). Even more instructive is the similarly restrictive influence exerted by regenerating limbs on the quota of nerve branches they admit to their territory. When a urodele limb is am- putated, the cut nerves promptly sprout branches in numbers superabundant for the supply of a full-sized limb. Nevertheless, only a small fraction enters the young re- generation blastema, and this fraction in- creases only gradually, in direct proportion to the growth of the blastema (Weiss and Walker, ’34; Litwiller, 38b), with a satura- tion constant of ca. 40 nerve terminals per cubic millimeter of tissue. It is difficult to escape the conclusion that each innervated tissue fragment establishes an inhibitory field around it which prevents the penetra- tion of competing fiber branches (analogous to territorial dominance observed in other tissue complexes; see Wigglesworth, 48; Willier, 52; Weiss, 53). Similar factors may 370 be responsible for the great local differences in the density of the central neuropil (see Herrick, °48, pp. 29-39). It is interesting to note, in this connection, that topical application of carcinogens en- tails a marked rise in the density of the local cutaneous nerve net (Julius, ’30), and that the presence of mouse sarcomata in chick embryos similarly opens certain vis- ceral organs, e.g., the mesonephros, which normally would have remained uninner- vated, to profuse invasion by sympathetic fibers (Levi-Montalcini and Hamburger, 53). These observations seem to indicate that the mechanisms controlling saturation density can be suspended by certain agents, among which tumor agents seem to be prominent. The peripheral density control (Fig. 137), whatever its nature may be, serves to in- sure, in combination with the other numeri- cal controls outlined earlier, the attainment and maintenance of an adequate functional state despite wide fluctuations of the indi- vidual developmental histories. It should have become clear from our discussion that such purportedly goal-directed performances, when properly analyzed, are resolvable into chains of causal mechanisms. DEVELOPMENT OF THE CENTRAL NERVOUS SYSTEM (CNS) DETERMINATION OF THE CNS The foregoing chapters, devoted to nerve fibers, have taken the existence of nerve cells for granted. To consider the origin of the nerve cells themselves, we must now turn back to an earlier phase of the em- bryonic history. As for the manner in which neural differentiation is initiated and the primordia of the neural system are mapped out in the early embryo, we may simply refer to the article in this book by Holt- freter and Hamburger. We shall take up the story from the time when the antero- dorsal sector of the ectoderm has become irrevocably earmarked for the formation of the neural organs and endowed with the capacity to produce parts with the morpho- logical, histological and chemical charac- teristics of central nervous system even when isolated from the rest of the germ. At that stage, by virtue of anteceding inter- actions of induction and segregation, the neural plate has become constituted as a system of fields, the median ones directing the transformation to central nervous struc- tures—anteriorly, brain and its derivatives; SPECIAL VERTEBRATE ORGANOGENESIS posteriorly, spinal cord—while the marginal ones give rise to neural crest. That this area already has biochemical properties that are distinctly neural is indicated by its selective immune response to antibodies prepared against adult neural antigens (Ebert, °50). At least in anteroposterior direction, it con- stitutes a definite mosaic of fields (Dalcq, ’47; Nieuwkoop, °52), the individual re- gions of which already contain differential conditions guiding the subsequent steps of morphogenesis toward the formation of spe- cific localized parts of the CNS. It is these steps, involving mainly folding, cell move- ments, proliferation, cell growth, cytodif- ferentiation, secretion, and cell degeneration, that we shall now consider in greater detail. EARLY MORPHOGENESIS Neurulation. Transformation of the neural plate into the neural tube occurs by means of forces residing within the plate itself, for as was shown by Roux (1895), the fold- ing takes place even in excised and isolated plates. The expansive pressure of the sur- rounding epidermis in the germ plays merely an adjuvant role (Giersberg, ’24). The dynamics of the folding process, as those of similar invaginations (e.g., gastru- lation; see article by Costello in this book) are still not fully understood. They are based on the development of differences in surface expanse between the outer and inner sides of the plate. An earlier suggestion, attributing this difference to differential water uptake (Glaser, 714), has now been discounted (Glaser, °16; Brown, Hamburger and Schmitt, ’41). Differential cell growth or cell multiplication has likewise been ruled out (Burt, °43b; Gillette, °44; Hutchinson, °44). The most likely assumption is that of an active contraction of the outer surface of the plate, the contractile elements, pre- sumably fiber proteins, being either in the cells (Lewis, ’47) or in their outer coating (Holtfreter, °43), or perhaps in the inter- cellular fiber cement around the outer cell poles seen under the microscope as “terminal web” (Sauer, °35). Active elongation of the medullary cells may normally assist this process (Brown, Hamburger and Schmitt, °41; Holtfreter, °46); ultraviolet irradiation of the germ with wave lengths near the absorption maximum of sterols inhibits it (Davis, *44). When the raised folds meet from the two sides, epidermis fuses with epidermis, and neural layer with neural layer. Since this NERVOUS SYSTEM occurs even in asymmetrically mutilated embryos under mechanically wholly aber- rant conditions (Holtzer, °51), it cannot be simply a mechanical accident, but must be viewed as a case of selective fusion of tis- sues according to their respective affinities (see Holtfreter, °39; Chiakulas, ’52). Simi- larly, the extrusion at this stage of the neural crest cells from the confines of the neural plate might be an expression of a a7 1 cylindrical lumen. This actually occurs in isolated pieces of plate in homogeneous sur- roundings (Holtfreter, ’34). The slit-shape of the normal tube has been shown to de- pend on the presence of notochord (Leh- mann, 735) (Fig. 138C,E£). The effect may be credited to a vertical system of fibers, spanning the thickness of the plate along a median strip coextensive with the noto- chord and apparently attached to it, which Fig. 138. Shape of neural tube under different conditions (from Holtfreter, ’34). A, Solid neural mass developed in explantation: nuclei of gray matter crowded near the surface; white matter in the interior. B, Neural tube surrounded by mesenchyme: shape cylindrical with central lumen; nuclei massed at the inner (free) surface. C, Asyntaxia dorsalis (failure of the tube to close); thinning of the floor of the tube in contact with the notochord; gray matter along the free surface. D, Neural tube underlain by muscula- ture; lumen eccentric at far side; white matter at near side. E, Neural tube underlain by notochord; normal appearance; slit-shaped lumen, oriented towards notochord. /, Lumen; mm, mesenchyme; c, noto- chord; s, segmented musculature. transient incompatibility (or disaffinity) be- tween the two groups. Thus, visible move- ments and shaping processes appear as the observable results of more intimate physico- chemical distinctions within the prospective neural system. Early Morphogenesis of Brain and Cord. The gross shapes of the early brain and cord, respectively, are anticipated in the proportions of the neural plate, whose wide anterior part, upon folding upward, forms the large vault of a brain ventricle, while the narrower posterior part encloses the narrow lumen of the central canal of the spinal cord. The shape of the canal varies with the details of the folding process (Fig. 138). Uni- form curling of the plate would leave a holds the midline firmly anchored as a hinge about which the flanks of the plate fold up (Weiss, ’50c). A similar fibrous plane seems to define the border between the alar and basal plate cell masses; as the latter grow and bulge, it gives rise to the lateral sulcus. Because of their importance for the later regular distribution and group- ing of cell columns, such tangible traces of early subdivisions would merit more in- tensive study; at present, we have no more than vague hints as to their presumable role. After the closure of the groove, the turgor of the fluid in the lumen assumes the mor- phogenetic role of firm support for the limp walls, which otherwise would collapse. The source of this turgor has been found in 372 the secretion of fluid from the cells of the inner lining of the early ventricles (Weiss, °34b; Holtzer, 51). Furthermore, the ciliary beat of the lining may propel the fluid anteriorly, which in the normal embryo would help to maintain the distention of the brain cavity. In cyclostomes and teleosts, in which the CNS is laid down initially as a solid cell cord, this same secretion process seems to be the method by which the central lumen is secondarily established in the in- terior. The shrinkage of the central canal by partial fusion of its walls (see Ham- burger, °48), paralleled by the decline of mitotic activity, may reflect a reduction of turgor in the spinal portion. With hydrostatic pressure on the inside and the confining skull capsule on the out- side, continued enlargement of the brain wall by growth, cell migrations and the de- position of white matter must be expected to lead to deformations, which, depending on the local conditions, manifest themselves as Cave-ins, outpocketings, fissures or folds. Practically nothing is known about the me- chanics of these elementary shaping proc- esses, although there are at least some indi- cations that the fissures between major di- visions of the cortex actually arise as cave-ins along lines of least resistance in the wall which tends to expand in confined space (Clark, ’45; Kallén, 51). It must be empha- sized, however, that the systematic pattern, according to which such mechanical events take place, is intrinsically prepared by the inequalities established previously by the locally differing processes of proliferation, migration, aggregation and differentiation (see Bergquist and Kallén, ’53a); the gross mechanical factors do not create these dif- ferentials, but merely translate them into more conspicuous spatial configurations. Accordingly, the attainment of normal brain configuration depends not only on the typical development of the brain wall, but also on the proper harmony between the latter and the growth of the skull capsule (or in the case of the cord, the spine) on the outside, and the turgor of the cerebro- spinal liquor on the inside. If this harmony is disturbed, either by a genetically deter- mined imbalance between the component tissues or by later trauma or nutritional deficiencies, serious aberrations of the CNS will ensue. Genetically conditioned hyper- secretion of central fluid, for instance, leads to hydrocephalus and brain herniation (Lit- tle and Bagg, ’24; Bonnevie, ’34; see Section SPECIAL VERTEBRATE ORGANOGENESIS XIV, Teratogenesis, by Zwilling); delayed closure of the folds past the onset of secre- tion, to various grades of spina bifida with draining fistulae (for an example of me- chanical production of spina bifida, see Fow- ler, 53); and retardation of skull growth in vitamin A deficiency, with unimpeded growth of the CNS, to brain compression and herniation (Wolbach and Bessey, ’42). The early cartilaginous capsule, at least in the spinal region, can accommodate its size to the actual dimensions of the en- closed CNS (Holtzer, 52a), but this adapt- ability is certainly greatly reduced in later stages. Morphogenesis of the Neural Crest. The neural crest, which contains precursor cells for spinal ganglia, sympathetic ganglia, Schwann cells, pigment cells, and, in lower vertebrates, also branchial skeleton and ordinary mesenchyme, is regionally spe- cialized even before its cells start on their migrations away from the dorsal midline (Horstadius, *50; Niu, °47). The bilateral cell masses first move down in rather co- herent sheets (Detwiler, ’37b); the parts of prime interest here, those giving rise to the ganglia, then settle down in two major columns, one between spinal cord and myo- tomes, the other alongside the aorta, with further outposts moving into the viscera. The localization of the ganglionic columns is hardly a simple matter of filling open grooves between the tissues, but rather an expression of specific contact affinities be- tween the crest cells and the surrounding cell systems. We find a model of this process in the formation of pigment bands by neural crest-derived melanophores (Twitty, °49), where the migrating cells likewise aggregate along certain tracts pre- formed in the surrounding tissues. Later the continuous columns break up into segmental clusters. The tendency to separate into smaller groups seems to be intrinsic to the cells, but the segmental localization of these groups is determined by the segmental arrangement of the myo- tomes, for the experimental removal (or disarrangement) of the latter abolishes (or correspondingly disarranges) the segmental array of the ganglia (Lehmann, ’27; Det- wiler, ’34, ’35). The segmental arrangement of the nerve roots and neural arches is likewise dependent on the presence of axial mesoderm (Detwiler, ’37b), but just how these processes are causally interrelated is not yet clear. NERVOUS SYSTEM GROWTH PATTERNS Proliferative Sources. In the ganglia, cell division is rather ubiquitous. In the CNS, on the other hand, mitoses are, in post-neurula- tion stages, confined to the inner surface, lining central canal and brain ventricles (Fig. 139). More peripheral layers are es- sentially devoid of mitotic figures; the cells there continue to grow in size, but without ensuing divisions. Whether the inner “ger- minal” layer is subject to some active mito- genic stimulation by its exposure to the lumen, or rather division in the outer layers is actively inhibited by local conditions in the mantle, is uncertain. However, upon injury to the early CNS (Hooker, ’25), as well as after unilateral ablation (cord: Detwiler, °44; Holtzer, ’51; not observed in midbrain: Detwiler, ’46b), mitotic cells may appear throughout the mantle, which seems to disprove an early loss of divisory faculty. There is a remote possibility that the con- finement of mitotic figures to the inner surface might not truly express the position of the germinal cells in the resting stage, but that the latter might merely rise to the surface during mitosis (Sauer, °35). The observation that x-irradiation of embryos destroys a cell layer somewhat deeper than the inner lining (Hicks, 52) could be in- terpreted in two ways: either these deeper cells are the true germinal ones, but are im- paired in their normal premitotic centripetal movement, or else they are postmitotic cells defying the general supposition that germinal cells are the most sensitive to radiation (Hicks, *53). At any rate, once neuronal differentiation has become marked, none of the resulting nerve cells would ever divide again under ordinary circumstances. We therefore can confine our consideration of proliferative patterns to the inner, “ger- minal,” layer, whatever its precise delinea- tion may be. In this layer, mitotic density and rate vary characteristically both along the longitudinal and in the dorsoventral di- rection, as will be described more fully below. Growth and Shape. Daughter cells of a germinal mitosis may either remain in the germinal layer or move off into the deeper layers. In the former event, they can continue to proliferate, while in the latter event, they merely add to the bulk of the mantle with- out further reproduction. Since the relative frequency of the two events will depend on the spatial arrangement, the configuration of 373 early brain and cord assumes significance for their over-all growth rate. According to Hertwig’s rule, the plane of division of a radially elongate medullary cell should lie parallel to the surface; hence, of the two daughter cells only one would remain a germinal cell. We know neither what re- arrangements take place after division, nor what actuates and guides the movement of the proliferated cells. However, it seems that, at least in the earlier stages, they glide along other medullary cells stretching across the Fig. 139. Cross section of spinal cord (15th seg- ment) of 614-day chick embryo showing mitotic activity of germinal layer, predominantly in dorsal half (from Hamburger, ’48). neural tube and brain wall, or along radial fibers deposited by the medullary cells. If all new cells kept on being thus displaced in the radial direction, the neural tube would gain only in width, but not in length. The factor counteracting this trend seems to be the longitudinal stretch to which the tube is passively subjected by the lengthwise extension of the surrounding tissues, espe- cially the notochord (Horstadius, ’44; Kitchin, 49). This would divert a certain fraction of the new cells in the longitudinal, rather than radial, direction, and by extending the ger- minal surface area, also lead to a progressive expansion of the proliferative source. While still hypothetical, this view is supported by 374 the observation that isolated sections of the spinal cord, transplanted to the flank, where they are deprived of axial stretch, do become much thicker and shorter than if they had developed in continuity with the rest of the cord (Zacharias, ’38). The ease with which lateral halves of the cord are regenerated with the participation of transverse cell shifts from the intact half (Detwiler, ’47b; Harrison, ’47; Holtzer, °51), contrasted with the failure to repair major gaps in the longi- tudinal direction, likewise indicates the SPECIAL VERTEBRATE ORGANOGENESIS tivity declines at some levels and flares up at others. Moreover, the centers of proliferation often form a quiltlike pattern with foci at the intersections of four longitudinal columns with transverse bands (neuromeres) (Berg- quist and Kallén, 54). Some of the segmental peaks coincide with the appearance of promi- nent peripheral organs in the corresponding sector, for instance, in the limb segments. As will be shown later (p. 381), part of this correspondence is due to an active control ex- erted upon the centers by their respective Fig. 140. Mitotic pattern of the brain of Amblystoma (after Burr, ’32). A, Earlier stage (31, after Harri- son). B, Later stage. (35, after Harrison). To facilitate comparison, the two stages are represented side by side in symmetrical arrangement. The outline of the brain is indicated. Horizontal lines express the number of mitotic figures contained in serial brain slices of 30u each. Abscissa: Number of mitoses. Full black: Mitoses in forebrain and hypothalamus. Individual bars: Mitoses in the rest of the brain. Brackets indicate position and extent of sensory placodes. Note the change in the distribution of peaks from A to B. need of extraneous stretch for elongation. In the brain, the hydrostatic turgor deserves similar attention as a potential regulator of the expanse of the proliferative surface, hence of the over-all growth rate. These con- siderations would not apply, of course, to those parts of the CNS which arise, not as transformations of the original neural plate, but by secondary budding processes (e.g., the posterior parts of midbrain and spinal cord in amphibians). Regional Patterns of Proliferation. Regional variations of mitotic activity within the germinal layer have been mapped out for several species and stages (urodele spinal cord: Coghill, ’24, ’36; urodele brain: Burr, *32; chick cord: Hamburger, ’48; comparative studies: Bergquist and Kallén, 53b). Each stage has its own characteristic pattern (Fig. 140); that is, as development proceeds, ac- peripheries. But there are also some basic intrinsic growth differentials among differ- ent regions of the CNS (Coghill, 36; Det- wiler, 24a, °36b; Hamburger, 48), the origin of which remains in need of explanation. It has been proposed that local prolifera- tion is regulated by the number of intra- central tract fibers ending in a given locality (Detwiler, *36b). This has been inferred from experiments in which posterior cord pieces of potentially smaller end size (of urodele tail-bud stages) and anterior ones of potentially larger size had been mutually exchanged and found to develop in accord- ance with their new sites, the former growing beyond, the latter remaining below, the sizes they would have attained in their original positions (Detwiler, ’23, ’24a). Since more descending brain fibers terminate at more anterior than at more posterior levels, these NERvous SYSTEM fibers were thought to have determined the site-specific growth rates. Confirmatory evi- dence was seen in the fact that when the anterior cord segments were replaced by a supernumerary medulla as extra fiber source, the spinal host segments lying immediately posterior to the graft became abnormally large (Detwiler, ’25c). In the light of later work, however, this view has become untenable. For instance, when the normal medulla as the suppler of supposedly stimulating fiber tracts was re- placed by a spinal fragment of much lower fiber productivity, the host spinal cord be- hind developed quite normally, without the expected diminution (Detwiler, ’37c). Simi- larly, suppression of forebrain development entails no deficienies in the medulla (Det- wiler, 45). In the chick, cord segments which have been transected and prevented from re- ceiving down-growing fiber tracts still gain normal dimensions (Levi-Montalcini, °45; Hamburger, 746). Also, the assumption that descending fibers would have an intrinsic preassigned length at which they would stop is gratuitous; free nerve fibers do not stop spontaneously, but are stopped by their surroundings, usually a recipient cell. Hence, fiber tracts cannot be the major determinants of axial growth patterns, although they can exert some modifying effects (see p. 383). All evidence thus leads to the conclusion that the closed neural tube represents a longi- tudinal mosaic of specifically different local fields, each guiding the further fate of the area under its control; differential prolifera- tion is but one expression of these different fates, with which we shall deal more fully below in the proper context (p. 376). Besides the longitudinal pattern, there is also a notable dorsoventral differential of mitotic activity, both in regard to intensity and time course. Proliferation is more abun- dant and lasts longer in the dorsal than in the ventral half of the cord, with a rather sharp demarcation between the two zones (see Fig. 139). Generally noted in amphib- ians (Detwiler, ’25a; Maclean, ’32; Coghill, 33), this fact has been most conclusively demonstrated in the chick (Hamburger, ’48). It may be related to the general precocity of the ventral, as compared to the dorsal, portions, including the precession of motor over sensory function (Coghill, ’29). Dorsal and ventral halves also differ in other re- spects, the ventral one being far richer in alkaline phosphatase (Moog, ’43), and at the same time being the first to receive vascularization (Feeney and Watterson, ’46). 34/5 Growth Rate. Conventionally, the term “growth rate” refers to average increment of an organic object per unit of time. It is determined by measuring the object at the beginning and end of given intervals and dividing by the times elapsed. The resulting values are useful for rough orientation, but are often meaningless for analytical and comparative purposes, except for homogene- ous systems. If the dimensions of one part of the CNS increase faster than those of another part, this need not mean that the intrinsic growth activity of the former has been greater or that its mitotic rate has been higher. If we consider, for example, a given subdivision of the CNS bounded by certain landmarks by which it can be identified at successive stages, its volume increase is the resultant of the following tributary proc- esses: cell growth, accompanied by division in the germinal layer; cell growth without division in the mantle; immigration of nerve cells and other cell types (glia) from neigh- boring regions; emigration of cells; out- growth of axons and deposition of myelin (with that portion of the axons which mean- while has moved beyond our landmarks being unaccountably lost); passage of axons from other areas; invasion of blood vessels: accumulation of interstitial fluid; and cell disintegration. Large-scale destruction and resorption of cells is a common, if neglected, feature of embryonic development (Gliicks- mann, *51), and as we shall see presently, is quite prominent in the embryonic nervous system. From this listing, it should be plain that comparisons of the “growth” of different parts of the CNS on the basis of mere measurements of volume or mass cannot be very revealing and must be interpreted with due caution. There is an even wider margin for error when, instead of volume deter- minations, only one, supposedly representa- tive, dimension is sampled; e.g., cord diam- eter. The case is illustrated by the fact that the increase in area observed in cross sec- tions of isolated cord segments (Severinghaus, °30) often was interpreted as hyperplasia un- til more complete determinations disclosed that there had been a corresponding reduc- tion in length (Zacharias, ’38) (see p. 374). This comment should not detract from the value of the summary quantitative treatment usually accorded to CNS growth, which has been a true advance over earlier, purely verbal, descriptions; rather is it to stress the need for going even further and identifying precisely and quantitatively the various 376 component processes of unequal kind and weight, whose disparate contributions are indistinguishably lumped in any bulk de- termination. REGIONAL AND CYTOLOGICAL DIFFERENTIATION Histochemistry of Differentiation. Superim- posed upon the growth processes just dis- cussed appear the cytological specializations usually designated as “‘differentiations.” It is an axiom of development, however, that the morphological (i.e., microscopically discern- ible) “differentiation” of a cell (other than trivial geometrical changes) is but the visible expression of intimate changes in the compo- sition and distribution of at least part of the molecular population which constitutes that cell (cf. Weiss, ’49, 53). Any funda- mental distinction between physicochemical (sometimes called ‘“physiological’”) and morphological differentiation is thus purely artificial, for it refers not to any dichotomy in the properties of a cell but merely to two different techniques of observation, both with definite limitations, hence, supplementing each other. Accordingly, marked differences in morphology point to antecedent physico- chemical changes, even though the latter may not yet be detectable by our relatively crude analytical tools, and conversely, de- monstrable differences of physicochemical constitution (including differential staining in histological preparations) signify basic dif- ferences of protoplasmic properties (cytodif- ferentiation) even if these have failed to ex- press themselves in corresponding differences of structural detail. It is with these qualifica- tions in mind that the results of correlated studies between morphological and chemical ontogeny of the nervous system should be viewed (see also Section III, Chapter 1, by Schmitt). A mental separation must be made between those chemical systems that are common to all cells (e.g., the ones engaged in respira- tion, energy transfer, protein reproduction, etc.) and those that are peculiar to the ner- vous system. In practice, this is not feasible because of the fragmentary state of our pres- ent knowledge and because there are un- doubtedly quantitative variations of the former class which are as distinctive of nerve cells as are qualitatively specific compounds. Cytochemical studies of neuron growth have already been referred to above (p. 365). Even less is known about the specific biochemistrv of neural differentiation. SPECIAL VERTEBRATE ORGANOGENESIS Among the most profitable contributions to this field have been studies on the develop- ment of the cerebral cortex in which cyto- logical, chemical, metabolic and functional observations were correlated for a series of sample stages. They have led to the recogni- tion of certain critical phases during which differentiation advances in spurts. Within the same brief interval, the nucleus reaches mature dimensions, Nissl bodies appear in quantity, dendrites become more numerous, neurofibrils more prominent, the activities of cytochrome c, adenylpyrophosphatase, and succinic dehydrogenase rise sharply, and electric brain potentials are recordable for the first time (Flexner, ’50). In amphibians, motility develops in close parallelism with the production of acetylcholinesterase (Youngstrom, °38; Sawyer, °43; Boell and Shen, °50), which is instrumental in nerve conduction, and experimental modifications of the size of the brain (see p. 383) are reflected in corresponding changes in _ its content of these products (Boell and Shen, 631 These examples may suffice to discourage the practice of divorcing morphological from underlying physicochemical considerations. At the same time, much of this work must still be counted in the descriptive class, fur- nishing important data of information but not yet much causal understanding. Appearance of Regional Differences. As mentioned above, different levels along the longitudinal axis of the early CNS enter different developmental courses, which subse- quently express themselves in the overt mosaic of morphological, histological, and eventually, functional specializations. Al- though some gross mosaic features may be conceded to the neural plate from its very first appearance (e.g., Nieuwkoop, °52; see above, p. 370), its finer parcellation is a continuing process the progress of which can be tested by appropriate experiments (see Section VI, Chapter 1, by Holtfreter and Hamburger). The standard test consists of displacing the part in question (by explanta- tion, inversion, or heterotopic transplanta- tion) in a graded series of stages and es- tablishing precisely from what stage it can carry on a course of development typical of its original site even under aberrant environ- mental conditions. This test presupposes that features attained under the original and under the aberrant conditions are sufficiently distinct to be used as criteria. Individual blood cells, muscle cells and pigment cells, for instance, can be NERvouS SYSTEM easily distinguished by their inclusions, dif- ferent gland cells by their secretions; but in- dividual nerve cells, judged by their shapes, whose normally great variability is further exaggerated under experimental conditions, S888 02.0528, rf o 8 2. ° ° 30 “oS. ° © 2°0000 'o ou SGo%00 06, 470%2° 0\ 6 80%9 9°00 200 fe) to oo’ Yoo rand Ce ee i vy, 200000 88 elo ento So Poe Set eO Ge 0o? th \ oe 20 922 o%g00e _& OG © 906250 059 0 0040 O29 889 C29 0%GO%, ey tacos BS gees ptt S717 within each class, which must be postulated on functional evidence (see below, p. 384), for instance, between motor cells innervating different muscles or between neurons sub- serving different sensory modalities. The Loo Po 5 0°70: oS ode SDRPE So a Sa} 0 cies y) €or SOS" SS ee See G3 “~9c0a,.0° 20 PC) ° 0900 0990 ° Fig. 141. Diagrammatic frontal sections of the spinal cord of chick embryos of 4 days (A), 5 days (B) and 8 days (C), showing the regionally differing formation of the motor columns from a morphologically rather uniform condition. In B, degeneration of cells in the cervical region (c, black circles), and centripetal emi- gration of cells in the thoracic and sacral regions (th) have set in, resulting in the distribution shown in C. (From Hamburger and Levi-Montalcini, ’50.) cannot always be so readily identified as to the precise type to which they belong. Large motor cells, Rohon-Beard cells, spinal gan- glion cells, commissural cells, and a very few others are sufficiently distinctive to be used as cytological indicators for the respective neuron types. But there are no corresponding microscopic signs for the finer distinctions only individual cell that can be strictly identified as such is Mauthner’s neuron, of which there is in many species only a single pair, situated at the otic level of the hind- brain of lower vertebrates, and which can be recognized by its giant size. For the rest, one must rely on general morphological group criteria, such as cross-sectional con- 378 figuration, characteristic location and distri- bution of cell columns within the unit, and the lke. Applying these tests, it has been found that the consolidation of regional properties fol- lows a definite time sequence: in general, more anterior parts have a head start over more posterior ones, and fixation along the longitudinal axis seems to precede fixation in the mediolateral direction. Presumptive medulla oblongata of a urodele tail-bud em- bryo transplanted in the place of the first cord segments of another embryo develops into a supernumerary medulla oblongata (Detwiler, ’25c). Even an isolated lateral half of presumptive medulla grafted into a lateral gap of spinal cord develops according to its origin, quite incongruously for the location, into half a medulla (Detwiler, ’43a). Yet, while anterior spinal levels at the same stage show similar capacity for self-differen- tiation (Maclean, ’32), more posterior seg- ments are still able in some measure to con- form to site-specific determinative influences (a 387): All existing evidence leads to the conclu- sion that, in the later neurula stage, alloca- tion of differential developmental properties within the CNS has already made notable progress. The cord seems to lag slightly be- hind the brain, but by the time different cord segments have first become externally distinguishable (in birds), each one after transfer to another region develops essen- tially as if it were still in the original place, mosaic-fashion (B. Wenger, ’51). The region- ally specific “developmental properties” contain the whole ground plan for all sub- sequent transformations, that is, they deter- mine pattern and basic rate of proliferation, specific differentiation of cell types, migra- tion, establishment of specific mutual group relations, patterns of cell destruction, and the physiological mechanisms underlying functional activity and coordination. Thus segmental organization in the chick embryo, for instance, which begins to emerge on the fifth day of incubation, shows up (Fig. 141) in abundant cell degeneration in the cervi- cal segments, segregation of a large lateral motor column in the brachial segments, ab- sence of a lateral column and presence of a characteristic column of preganglionic autonomic cells (nucleus of Terni) at the thoracic and sacral levels (Levi-Montalcini, 50; Hamburger and Levi-Montalcini, ’50). In the amphibians, where functional tests have been carried further, it includes such properties as the ability of the limb seg- SPECIAL VERTEBRATE ORGANOGENESIS ments to engender coordinated limb move- ments, in contrast to the lack of this faculty in the trunk and head regions (see below, p. 387). Similar early differences of properties at different levels of the main axis have been demonstrated for the spinal ganglia in re- gard to cell proliferation and regulatory ability (amphibians: Lehman and Youngs, 62) and incidence of cell degeneration (chick: Hamburger and Levi-Montalcini, ’49). It must be stressed that even as the mosaic of diverse parts of the CNS emerges, each area at first seems to operate “field’’-fashion, that is, while its general fate is fixed, the detailed course of its component elements would still be variable. This is indicated by such observations as the following. At a time when the development of the medulla oblongata has already become irrevocably identified with the part of the tube lying at the presumptive hindbrain level (see above), one can still turn that piece around, reversing its anteroposterior axis, and yet see it yield a medulla of normal (unre- versed) configuration (Detwiler, ’43b). Cord regions of the same stage after anteropos- terior reversal regulate similarly (Detwiler, 24a). However, the interpretation of these regu- latory effects calls for some caution. While shape and proliferation rates, which served as criteria, may indeed be as adaptable as indicated, this need not apply equally to the segregation of the various qualitatively specialized cell strains, which might follow a totally different time pattern and whose determination might be either precocious or delayed relative to those other features. Experiments with Mauthner’s cell (M- cell) seem to be pertinent to this prob- lem. Unilateral extirpation in a post-tail-bud stage of the hindbrain at the prospective M-cell level leads to permanent absence of an M-cell on the operated side (Detwiler, °33). Since heterotopic grafts of the same area from a neurula yield supernumerary M-cells (Piatt, 44; Stefanelli, ’50)—and as- suming that the history of the M-cell is representative of that of other, less con- spicuous, cell types—we must conclude that even at this early stage a rather detailed fixation of type characteristics has occurred in the various precursor cells of the later cerebral neuron strains. Repetition of the experiments in earlier stages using the M-cell as marker has narrowed the critical period of determination to the late gastrula NERVOUS SYSTEM (in frogs). At that stage, while an explant or transplant from the presumptive M-cell region can already give rise to an M-cell independently, another M-cell may develop at the normal site from amidst the residual neural cells that have closed the gap from adjacent levels (Stefanelli, °50). This proves that the factors localizing M-cell develop- ment in the hindbrain had already been in operation but were still enough active to turn out a second set; furthermore, that there were still cells in the surrounding brain regions sufficiently labile in char- acter to respond to redirection. These results suggest that the character of cell strains is fixed much earlier than are the numbers, distribution and arrangement of their de- scendent cells. Determination in the transverse direction, that is, along the mediolateral axis of the neural plate, is likewise a gradual process. This can be concluded from the fact that at stages at which transverse strips of neural plate manifest definite “self-differentiation” of regional character after displacement, they still are capable of extensive regulation in the lateral direction (Roach, *45). How- ever, at the stage of the closure of the neural folds, some further mosaic subdivision has also developed in the mesiolateral direction (which after folding appears as ventro- dorsal direction); for in birds, localized defects placed in the neural epithelium at that stage result in cords in which the whole radial sector that would normally arise from and cover the destroyed patch is completely missing (E. Wenger, 750). It is to be noted that these results prove only the loss of capacity for regulative re- distribution of tasks within the residual fragment of nervous system itself; no re- generation of the missing sectors had been initiated. If regeneration, i.e., mobilization of new cell material to replace missing parts, can be activated, as has been possible in amphibians, the regulative faculty ex- tends into considerably later stages; again, it is far wider in the lateral than in the longitudinal direction. After the excision of a lateral half of presumptive midbrain, medulla or spinal cord segments from the neural tube, increased proliferation and mi- gration of cell masses across the midline from the residual intact half restore the missing portion with remarkable morpho- logical and, in earlier stages, also histologi- cal, perfection (Detwiler, *44, °46b, ’47b; Harrison, ’47; Holtzer, ’51). A comparative view of all these results 379 leads us to distinguish rather sharply be- tween determination of individual cell fate, strictly cell-wise, on the one hand, and “determination” of a cell complex, on the other, the latter connoting the imparting of some frame of conditions to the cell group which only in further consequence would gradually fix the characters of the individual component cells. In the case of the CNS, this presents us with the alternative that either (a) the diverse cell types of the later mantle are already preformed as such in a corresponding variety of precursor cells in the neural epithelium, or (b) the cells of the neural epithelium are still equipotential and acquire their differential type character- istics only through local influences of the different mantle portions in which they come to lie. There is not enough evidence on hand to decide this alternative crucially one way or the other. Except for some indirect mor- phological and _ pathological indications (Globus and Kuhlenbeck, 44), it has not even been definitely settled whether the dichotomy between glia (spongioblasts) and nerve cells (neuroblasts) is already effected in the germinal epithelium or whether both are derived from common stem cells. In the case of Mauthner’s cell described above one could assume, according to (a), that during neurulation a particular cell of the plate is endowed with the ability either to turn into an M-cell itself or to undergo an orderly sequence of unequal divisions eventuating in the segregation of one of the descendents as an M-cell. A model for such a process is known, for instance, in the pro- duction of the mother cells of scales in certain insects (Henke, *53). The occasional occurrence of twin M-cells in haploid em- bryos (Fankhauser, 52) could be taken as a sign of the disturbance of the regular cell lineage because of the undersized mass of the haploid cells. According to (b), one would assume that a particular cell block in the presumptive hindbrain would be endowed with “inductive” activities that would reach a rather sharp peak at a given focus, and that the cell that happened to be thus pin-pointed would thereby be singled out to grow up into an M-cell. There is some suggestive, but meager, evidence pointing to early cell type diver- gence according to (a). In the experiments on the repair of excised halves of spinal cord from cell sources of the opposite half (see above, this page), it was noted that when the operation was performed in_ successively 380 older embryos, definite cell types would be- gin to be missing in otherwise morphologi- cally well restituted halves, with the large motor cells dropping out first, commissural neurons next and general internuncials last (Holtzer, 51). This indicates that the re- generated cell types stem each from the homologous cell type of the intact half; that the various types lose their mobility or be- come otherwise unavailable as sources of replacement, one by one, in the observed time sequence; and that the descendent cells of different strains can no longer sub- stitute for one another. If this be the case, the term “indifferent,” commonly given to cells of no particular morphological distinctiveness, is misleading. Actually, these cells would constitute a heterogeneous population, each with defi- nite differential type characteristics, which may or may never come to the fully mature expression amenable to morphological classi- fication; they would not be a common pool of truly equivalent elements, which could still be switched into the various types of specific neurons by determinative local influ- ences. It is by no means unlikely that eventu- ally both assumptions (a) and (6) will turn out to be partly correct in the sense that some distinctive type specificity is already inherent in the cells leaving the germinal layer for the mantle, but that additional di- versity is imposed upon them by conditions along their path and at their final locations. The appearance of qualitative diversity among sensory neurons has also been dem- onstrated for the spinal ganglia. Aside from indirect deductions from the fact of selec- tive fasciculation (see above, p. 366), tangi- ble microscopic, topographical and_ be- havioral differences between cell groups subserving different functions have been revealed under the microscope (Levi-Mon- talcini and Levi, 43); as will be shown below, they likewise represent qualitatively different segments of a heterogeneous neuron population. In conclusion, there is ample evidence for the early emergence in the CNS of quali- tatively diverse cell strains, the number of recognized varieties being severely limited by the inadequacy of our means of dis- crimination; there is some evidence that the diversity of strains can at least partly be projected right back to a corresponding di- versity within their production source, the neural epithelium; and that this mode of development of qualitative regional diversity leaves an adequate margin for quantitative SPECIAL VERTEBRATE ORGANOGENESIS adjustments of numbers within each type— the adjustments which will form the sub- ject of the following pages. PERIPHERAL EFFECTS ON CENTRAL DEVELOPMENT Historical Remarks. A quantitative corre- spondence between nerve centers and their peripheral area of innervation has long been inferred from comparative and pathological studies. Congenital absence of an extremity, for instance, was found to be reflected in uni- lateral underdevelopment of the correspond- ing spinal segments (e.g., Edinger, °21). However, whether the missing parts had failed to develop from the start or had been formed but secondarily degenerated from lack of peripheral outlets could not be de- cided by such static observations. The first attempt to reproduce the results experi- mentally by removing limb buds in chick embryos (Shorey, ’09) led essentially to a confirmation of the fact that centers faced with a reduced periphery became (or re- mained) undersized; they also proved that the relation was a causal one, without, how- ever, elucidating its nature. In further cor- roboration, removal of an eye in early am- phibian larvae was found to entail reduced size of the optic centers in the corresponding (i.e., contralateral) midbrain hemisphere (Steinitz, °06; Diirken, °13). Yet, not until these defect experiments were supplemented by overloading experiments could the ac- tively stimulating nature of the peripheral influence be regarded as firmly established. The first well attested case was the ex- cessive development of spinal ganglia in trunk segments whose peripheral mass had been increased by the addition of a grafted limb (Detwiler, ’20). Continued experimen- tation in amphibians (chiefly by Detwiler and his school), and later even more pene- tratingly in birds (by Hamburger and co- workers), has reiterated an old lesson of biological research: a relation that on first acquaintance appears simple and transpar- ent, when subject to more minute analysis more often than not turns out to be much more complex, if not more obscure, than originally suspected. Significant differences were discovered in the response of different species, of spinal ganglia vs. spinal cord, between brain parts, and among different regions of the cord. Meanwhile, other types of peripheral influences upon central de- velopment have been discovered, such as the control of neuronal size reported above NERVOUS SYSTEM (p. 365) and the functional specialization of centers by their end-organs discussed be- low (p. 384), all of which add to our con- viction that the potent role of the periphery is exerted not by a single unitary mecha- nism, but by a multiplicity of interlocking ones. In the earlier studies of peripheral re- bound on central development, attention was focussed almost solely on the final size (num- ber of cells and total mass) of the affected central part. Depending on whether it was above or below the expected normal size (mostly calculated from the asymmetry be- tween the experimental and opposite control halves of normally symmetrical systems), the difference was described as “hyperplasia” or “hypoplasia,” signifying an overproduc- tion or underproduction of cells. However, since final size is determined not only by proliferation rate but also by cell migration, cell growth, cell destruction, etc (see above, p. 375), these terms are apt to be misleading and will not be used in the following dis- cussion (see also Hamburger and _ Levi- Montalcini, 49). When speaking of “periph- ery,” we shall mean tissues to be innervated, not just any extra-neural environment. Peripheral Rebound on Primary Neurons. Ganglia. No conspicuous excess of brachial over trunk ganglia can be detected after the respective peripheries have been roughly equalized by the suppression of the develop- ment of the limb (Detwiler, ’24c; Hamburger and Levi-Montalcini, °49). The intrinsic development in the absence of a limb may therefore be taken as the baseline over which effects of peripheral increase build up. Sup- pression of trunk muscles by myotomectomy results in still smaller ganglia (Detwiler, 27), but in view of the relative uniformity of myotomes in the normal animal, this fact can have no influence on the shape of our “baseline.” When a single limb develops (1.e., in the normal case) there results then a rise of this base value to the (normal) magnitude typical of intact limb levels. Adding a limb to a trunk segment raises the base value of the latter so that the spinal ganglia turn out larger than those of normal trunk segments (Detwiler, ’20; Hamburger, ’39b). In amphibians, the effect can be obtained throughout the larval stages until after metamorphosis (Carpenter, ’32, 33). Cranial ganglia likewise enlarge when they are made to innervate a supernumerary organ grafted to the head (Detwiler, ’30b). These increases are only partly due to in- creased proliferation; for the most part they 381 result from the fact that more of the “in- different” neuroblasts (see above, p. 380) in the ganglion are caused to mature into large typical dorsal root neurons (Hambur- ger and Levi-Montalcini, ’49), which thus add not only to the tally of identifiable sensory cells, but being larger, also to the mass of the ganglion. Besides, there occurs normally in various ganglia a_ certain amount of cell degeneration, which in the presence of a larger periphery, e.g., a limb, is held in check (Hamburger and Levi- Montalcini, 49). Thus, the final count of cells is regulated through at least three de- vices: proliferation, maturation and elimina- tion. Whether the effect is of a generalized kind, involving all neurons of the overloaded region, or a selective response of specifically matching types is still an open question. That there is some selectivity is definitely indicated by the ganglionic response to tu- mor transplantation. Ganglia at the level of a transplanted mouse sarcoma (in chick embryos) show the typical increase com- monly observed under conditions of periph- eral overload (Bueker, ’48). The effect is selective in that it exempts the motoneurons of the cord, remains confined to the medio- dorsal neurons in the spinal ganglia, and reaches its greatest intensity in the purely sympathetic para- and prevertebral ganglia (Levi-Montalcini and Hamburger, *51, ’53). This whole problem has recently been complicated by the discovery that the sym- pathetic ganglia of the chick embryo develop excessively even when the inducing sarcoma graft has been placed on the allantoic mem- brane, beyond the reach of actual fiber con- nections, evidently exerting its effect by some humoral agent (Levi-Montalcini, *52; Levi-Montalcini and Hamburger, *53). One could tentatively assume that the primary effect of the tumor agent consists of a gen- eral unstabilization of cell surfaces. The nerve cell bodies would thereby be enabled to issue more sprouts, and the nerve fibers more branches (see above, p. 361); this effect has actually been observed in tissue culture (Levi-Montalcini et al., 54). The visceral organs, on the other hand, would lose their surface protection against fiber invasion and could absorb the outgrowing branches (see p. 370). But the relation, if any, between these events and the ganglionic hyperplasia is by no means clear, and the factual analysis will have to be driven much further before any definitive explanation can be adopted. Spinal Cord. As in the spinal ganglia, the 382 peripheral rebound on the development of the spinal cord is superimposed upon an intrinsic pattern of regional differences (see p. 374), which may again serve as “base- line.” Even without limbs, this baseline is higher in limb segments than in cervical and thoracic segments (Bueker, 43; Ham- burger, 46). As for its potential altera- tion by changes in peripheral area, the results seemed at first to vary according to species. In urodeles, the presence of a limb caused no increase in the number of cells in the “motor” half of the respective cord region (Detwiler, ’24c), although in accord- ance with the conclusions reported under Factors Controlling Neuron Growth, p. 365, the individual neurons, hence the cross sec- tions of the motor roots, were larger (Det- wiler and Lewis, ’25). In anurans (May, 33) and birds (Hamburger, °34), on the other hand, the presence of a limb entailed a considerable enlargement of the “motor” cell columns of the cord. This apparent discrepancy turned out to be one of ter- minology. Whereas in the latter group the large cell bodies of motoneurons are com- pactly assembled in separate cell columns (“motor horns”) and therefore can be tal- lied separately, in urodeles they lie inter- mingled with other cell types so that for practical reasons the total number of cells in the ventral half of the cord was counted as “motor.” Subsequent cell counts in the chick (Hamburger and Keefe, 44) showed that in this form likewise the total cell num- bers in the ventral cord were not appreciably different whether or not a limb was there, but that in the presence of one, the ratio of “indifferent” to fully matured motor horn cells was shifted in favor of the latter, and only the latter had been counted in the earlier studies. It is clear from these facts that one of the effects of the actual presence of a limb in the limb segments is a recruitment process by which neuroblasts that would otherwise have remained less distinctive are induced to mature into large motoneurons. However, this is only part of the story. First, there seems to be also a certain, though minor, stimulation of (mitotic) cell proliferation (Hamburger and Keefe, ’44); why this does not lead to an increased cell number in cases with overloaded peripheries has not been resolved. Second, the ways in which the “baseline” values are attained in dif- ferent regions vary markedly and the ways of the peripheral influences vary accordingly. As outlined before (p. 378), in the chick the SPECIAL VERTEBRATE ORGANOGENESIS various levels of the spinal cord, initially of comparable cell content, gradually assume unequal sizes owing to increased cell degen- eration in the cervical region, increased maturation of ventral horn cells in the brachial region, and characteristic migra- tions of the cell bodies (or just the nuclei?) of preganglionic sympathetic and parasym- pathetic neurons in the thoracic and sacral regions, respectively (Levi-Montalcini, ’50) (see Fig. 141). A limb added to any of these regions will then cause segmental enlarge- ment by either reducing degeneration, or promoting maturation, or checking emigra- tion, respectively. This diversity of ways in which the final cell tally can be altered complicates the search for the underlying mechanisms and raises doubts in the assumption of a single common mechanism. As in the case of the spinal ganglia, one could maintain that it takes a primary connection between a center and its peripheral district by some pioneer- ing fibers in order to furnish that center with an estimate of its peripheral domain. But how does the pioneering neuron con- vey its information to others still in im- mature state? We have already commented on those changes in metabolic activity and other properties of a successfully connected neuron that express themselves in its size (p. 365) and fasciculation (p. 366). One need only assume that certain effects of these changes spread to neighboring cells (Ham- burger, ’39b; Hamburger and Keefe, 44). Observations in sheep embryos, showing co- incidence between the arrival of motoneu- rons at the periphery and the development of their dendritic fields, have led to the contention that spreading dendrites might be the transmitters of the inducing stimulus (Barron, *43, 46). But this explains little. For one thing, it could not apply to spinal ganglion cells, which lack dendritic inter- connections, and moreover, the unknown influence becomes no better known by being transferred from the perikaryon to the den- drites. The numerical increase of peripherally overloaded centers is limited by the output capacity of the respective centers. Thus, whereas the addition of a limb produces a marked increase over the limbless state, the further addition of one, two or even three extra limbs to the plexus fails to produce an appreciable further augmentation (Verzar and Weiss, 30; Weiss, ’37a; Bueker, °45). It is perhaps for this reason that large limbs transplanted orthotopically to a small body INERVOUS SYSTEM fail to evoke a corresponding cellular in- crease in the spinal limb centers (Schwind, 31; Harrison, ’35a). On the other hand, genetically hyperdactylous mice have been reported to have larger limb cord segments (Tsang, *39; Baumann and Landauer, 743). Whether this increase of central cell number as a result of the presence of extra toes is as limited as that observed in urodeles or whether the limb centers of rodents have perhaps retained a higher output capacity from their phylogenetic past, when they pos- sessed more toes, is an open question. Rebound on Secondary Units. Modifications of the size of spinal ganglia due to altera- tions of the periphery are also reflected in corresponding variations of the sensory col- umns of the spinal cord with which the former connect (Hamburger, ’34; Barron, ’45). This demonstrates the existence of transneuronal effects “in series” similar to the transneuronal effects “in parallel” just discussed. Instead of a non-neural periphery affecting its correlated neurons, one neu- ronal group now influences the quantitative development of another to which it bears an effector relation. No numerical increase of the internuncial cells discharging into the motor columns has as yet been seen to follow an induced increase of the latter, although the reverse, secondary degenera- tion of more proximal neurons in conse- quence of destruction of central fiber tracts, has been observed (see Bodian, ’42). Results comparable to those in the spinal segments have also been obtained with cranial nerves. Elimination of the labyrinth including the acoustic ganglion entails underdevelopment of some associated cell groups of the medulla (Levi-Montalcini, ’49). After the early removal of one eye in larval amphibians, the midbrain roof of the opposite side, end station of the crossed optic nerve fibers, develops defec- tively. Its size remains subnormal (Steinitz, 06; Diirken, ’13; Larsell, ’31), mitotic ac- tivity being reduced and the segregation of typical cell strata being impaired (Koll- ros, 53). As in the spinal centers, the effect is complex in nature, involving proliferation, migration and cell enlargement, although in the present case it is transmitted not directly but through intermediary neurons. It is not surprising to find that the mor- phological underdevelopment of an eyeless midbrain hemisphere is reflected in a deficit of its chemical products. Thus, the activity of acetylcholinesterase, an obligatory con- stituent of neural tissue, is reduced in pro- 383 portion to the reduced number of nerve cells (Boell and Shen, ’51). Midbrain centers connected with eyes of subnormal size (transplantation from small to large animals) are intermediate between those of normal and anophthalmic specimens (Twitty, °32). Genetically determined re- duction or suppression of eye development (microphthalmia, anophthalmia) has the same effect on the size of the optic brain centers as has the corresponding experi- mental interference (Chase, ’45). It is note- worthy that in insects, too, eye reduction, either experimentally produced (Kopec, ’22) or genetically caused (Power, °43), is cor- related with reduction of optic ganglia. Cephalopods react similarly (Ranzi, ’28). An artificial increase in the volume of optic nerve fibers reaching the brain yields the expected central enlargement. This has been obtained in the midbrain after replace- ment of the normal eye by one of excessive size (Harrison, 29; Twitty, °32), or after adding a supernumerary eye (Pasquini, ’27); and in the medulla, in response to the entry of an aberrant optic nerve from an eye grafted in the place of an ear (May and Detwiler, ’25). In conclusion, there is widespread evi- dence of quantitative regulation of the maturation of nerve centers from both the effector and receptor ends. Although the modes of action may differ, the principle is the same whether the “receptors” and “ef- fectors” concerned are sensory organs and muscles or other neuron groups. Presumably some of the intracentral regulations out- lined below (p. 386) are manifestations of this same principle. Its operation provides the nerve centers with the necessary latitude of adjustment to insure adequate central control despite the wide individual variabil- ity and unpredictability of the detailed pat- terns of innervation illustrated throughout this article. It is important, however, to remember that the degree of adaptive lati- tude is limited by intrinsic properties of the responding centers which date back to their earlier prefunctional and even preneural stages. Specific Modulation and Resonance. The pe- ripheral encroachment upon central develop- ment reaches its climax of refinement in the process of qualitative adaptation (‘‘modula- tion”) of neurons in conformance with the type of effector or receptor organ with which they connect. Let us explain this phenome- non in the case of muscle innervation where it was first discovered (Weiss, ’24). 384 In discussing nerve outgrowth, we have referred to several provisions for the over-all euidance of masses of motoneurons to mus- cle masses as their appropriate destinations (selective contact guidance, selective fascicu- lation, etc.), as well as for the preclusion of peripheral connections with inappropriate kinds of tissues. It must be remembered, however, that orderly motor function pre- supposes that the whole muscle mass of a region be not thrown into contraction indis- criminately or all at once, but that at any one moment, only definite selections of indi- vidual muscles be made to contract in com- binations yielding a “coordinated” move- ment. Physiological concepts of coordination are still rather controversial but they all agree on one point, namely, that the co- ordination of muscle contraction is the re- sult of the selective excitation of the motor ganglion cells connected with the muscle fibers to be actuated. Thus, in order to effect an orderly movement, the motor centers must “know” precisely which ganglion cells are hitched to what muscles. Theoretically, the way in which this knowledge is acquired could be conceived of about as follows: Either (a) there is a predestined motor cell group for each indi- vidual muscle and some detailed mechanism exists by which the axons of that cell group are routed precisely to the matching muscle to the exclusion of all other muscles, so that the pattern of connections would be stereo- typed for all individuals; or (b) lacking such stereotvped connections, the centers would “learn” about their specific relations to the periphery by “trial-and-error,” actuating muscles at first in random combinations and then fixing somehow those central linkages that had incidentally yielded useful move- ments. Most past and current thinking about coordination implies either one or the other of these assumptions. Yet, both are contra- dicted by the facts. As for (a), predestina- tion of motor fibers for particular muscles (rather than just muscles in general) is ruled out not only by the normal variability in plexus formation. but above all by the experimental proof that any motor fiber will innervate equally readily any individual muscle (see above, p. 363) and that, as will be described presently, coordinated function can be obtained even after deliberate ran- domization of the peripheral pathways. And as regards (b), not only has the “trial-and- error” period supposed to produce basic co- ordination by gradual learning never been SPECIAL VERTEBRATE ORGANOGENESIS observed (see later, p. 391), but, as we shall show below, the very crux of this thesis, that the degree of utility to the individual of an achieved movement is at all critical for the development of coordination, has been experimentally invalidated. Neither alterna- tive being tenable, the solution came from a third and unexpected direction. In condensed version, it is as follows: (1) Each individual muscle has some _ consti- tutional specificity by which it is distin- guished (presumably in its finer protoplas- mic chemistry) from all other muscles (ex- cept homologous ones); (2) it imparts its specificity to the motor nerve fiber to which it has become attached, and through the axon, to the ganglion cell; this progressive specification of motoneurons by, and in con- formance with, their individual terminations has been termed ‘“‘nerve modulation.” It is through this direct epigenetic backward pro- jection of the mosaic of muscular specifici- ties upon the population of motor ganglion cells that the centers are informed as to just where their communication lines ter- minate. For fuller information, the reader may be referred to earlier reviews (Weiss, °36, °50b, °52a). Only the basic experiment will be briefly summarized here. When a limb muscle of a larval am- phibian is transplanted near a normal limb and provided with innervation from a nerve branch diverted from the normal limb plexus—any limb nerve branch—the trans- planted muscle is always found to contract simultaneously with the muscle of the same name in the normal limb (principle of “myo- typic” function). With several supernumer- ary muscles, the rule applies to each of them separately, so that if a whole limb with a full extra set of muscles is added, their total activity duplicates the overt ac- tions of the normal limb; or if transplant and normal limb are of reverse asymmetry, their movements mirror each other (Fig. 142). Thus, ganglion cells of the limb level of the cord that happen to innervate muscles of the same name (i.e., of identical consti- tutional specificity) become functionally linked, even though the functional effects of supernumerary muscles or limbs are use- less, or outright detrimental, to the animals. Since in these experiments the choice of ganglion cells for the test muscle was en- tirely a matter of chance or assignment by the experimenter, it is evident that the muscle must have conveved its name to the central cell, and since this happens even in NERVOUS SYSTEM the complete absence of sensory innervation (Weiss, ’41a), it must have occurred by way of the motor axon itself. While the earlier experiments were car- ried out in functional larval stages, requiring the remodulation of neurons that had already been modulated once before, the results are the same after embryonic transplantations (Detwiler, ’25b, ’42). On the other hand, modulation has thus far been proven experi- mentally only in larval amphibians. After 385 the locality of origin of the transplanted patch, rather than bearing the “local sign” (Miner, *51). Similar results have been de- scribed for vestibular neurons (Sperry, ’45). The retina likewise consists of a mosaic of sectors of different constitutional specifici- ties which are projected into the optic nerve fibers, thereby enabling the latter to establish selective discharge relations with a corre- sponding central mosaic of specific receptor units in the midbrain roof (Sperry, °43, Fig. 142. Myotypic function of supernumerary muscles. A transplanted limb with reversed symmetry (right limb) near a normal left limb mirrors the movements of the latter. (From motion picture, Weiss, 52a.) metamorphosis, neurons lose their plasticity and retain the specificity acquired previ- ously. Rat nerves transposed to other muscles postnatally likewise failed to undergo re- modulation (Sperry, 41). At the same time, there is strong evidence that modulation takes place in prenatal stages in all mammals, including man (Weiss, ’35). Sensory neurons are subject to the same qualitative modulation by their respective end-organs as are motoneurons. Propriocep- tive fibers connected with any kind of mus- cle signal to the cord the correct name of the particular muscle (Verzdr and Weiss, ’30). Trigeminal neurons of the skin newly made to innervate transplanted cornea thereby ac- quire corneal character and corresponding reflex relations (Weiss, ’42; Kollros, ’43b). A transplanted larval skin patch from a for- eign sector imparts its foreign specificity to the local cutaneous fibers that innervate it; thereafter when these fibers are stimu- lated they evoke reflexes characteristic of ’44). In urodele embryos this qualitative mo- saic condition is attained about the neurula stage (Stone, *44). While the rather general validity of the principle of specific neuron modulation by effector and receptor organs seems thus well established, the nature of the processes in- volved is still undefined. To judge from its slow rate, as well as its qualitative diversity, modulation belongs to quite a different class of processes than nerve conduction. Its speci- ficity sets it apart from more general “trophic” effects. Perhaps it resembles most closely phenomena of induction, infection and immunological sensitization. Modulation, it must be stressed, clarifies only one aspect of what really is a two- sided phenomenon. By labelling, as it were, the ganglion cells according to their ter- minations, it produces a qualitative point- for-point replica in the centers of the periph- eral receptor and effector units. Possibly the “labelled” neurons can, in turn, transfer 386 their specificities to other, more proximal neurons (Weiss, "41a; Sperry, 51). None of this, however, touches on the problem of coordination and its origin as such; that is, on the mechanisms by which the individual units are actuated in such orderly groupings and sequences that their total effect yields an integrated movement, such as walking, swimming, feeding, etc. As the physiological study of the neural elements has far out- distanced the understanding of their group behavior, we are still without a concept of coordination that could claim cogency or general applicability. This matter would be of no particular concern to us in the present context but for two reasons: first, no con- sideration of neurogenesis would be complete unless it included some account of the de- velopment of the “integrated activity” of behavior; with this we shall deal briefly later. And second, much light can be shed on the nature of coordination from a study of its ontogeny; for this, the experiments just reported offer a relevant example. The experiments have shown that not only does an extra set of muscles operate in the correct combinations and time sequences re- quired for normal coordination (as expressed by the near-by normal limb), but a single original set of muscles, when scrambled or otherwise abnormally arranged and inner- vated, also operates in the same stereotyped order; that is, each muscle as an individual contracts at such time and in such strength as would be called for in the particular movement of a normal limb (Weiss, *41a), regardless of the fact that, owing to the anatomical disarrangement, this blind execu- tion of sequences designed for normal ar- rangement results in a wholly senseless performance. Thus, an amphibian provided with limbs of reverse symmetry (by ex- changing right and left limbs) executes all movements in reverse, e.g., walks backwards whenever it is due to advance, and vice versa, all its life (Weiss, ’37b). Since this occurs likewise in animals in which the limbs had been reversed as buds and which therefore had never experienced the use of normal limbs, it is plain that the basic coordinating mechanisms, which call the differerit muscles for a given movement into operation in the proper selection and se- quence, are intrinsic and stereotyped prod- ucts of CNS development and operate blindly regardless of the effectiveness or inefficacy of the resulting movements. These central mechanisms in which the various coordina- tion patterns are preformed might be called, SPECIAL VERTEBRAE ORGANOGENESIS with a non-committal term, “central action systems.” They deal not with the muscles as such, but with the ganglion cells modu- lated by the latter. Modulation merely sets the muscles into the proper response rela- tions with the central action systems, but it does not govern their construction. For- mally, the relation between central action systems and the modulated receptor and effector neurons resembles communication by “resonance.” In conclusion, the response relations within the CNS are now recognized to be ruled by qualitative specificities of great subtlety, far beyond morphological detec- tion, and not simply by geometrical relations of otherwise equivalent units. DEVELOPMENT OF CENTRAL ACTION SYSTEMS The realization that in order to call forth, for instance, a coordinated elbow movement, the CNS must have developed the specific means to excite ganglion cells modulated by elbow muscles, presents us with a prac- tical test of the presence or absence of spe- cific action systems in a given central sector: a muscle group transplanted to the sector will not operate unless the center contains the matching set of specific activators. Ap- plying this test, it was found that coordi- nated limb activities are engendered only within the normal limb segments of the cord (Detwiler and Carpenter, ’29). Even a completely isolated brachial cord section can yield coordinated activities in a limb innervated by it, whereas an isolated piece of trunk cord in otherwise identical circum- stances cannot (Rogers, 734). The function of limbs innervated by trunk nerves alone remains abortive. Similarly, limbs trans- planted to the head and innervated by cranial nerves, while twitching in associa- tion with head muscles, never exhibit or- derly independent movements (Detwiler, ’°30b); moreover, such movements as are ob- served are attributable chiefly to local eye, gill or gular muscles that have attached themselves to the skeleton of the grafted limb, with the limb muscles themselves be- ing essentially uninvolved (Weiss, ’36; Piatt, ’41). We learn from these results that the regional differences within the cord, some of whose quantitative expressions we have previously encountered (under Regional and Cytological Differentiation, p. 376), are really much more profound, pertaining not only to numbers and configurations of cells NERVOUS SYSTEM but to all those other properties on which coordinated function depends. Since the mere attachment of a limb to a foreign spinal or cerebral region does not induce there the differentiation of effective limb control, the experiments also confirm our conclusion that modulation plays no constructive part in the design of central coordination patterns. The prevalent tendency to base orderly central function on precise patterns of neu- ronal connections gains little support from the experimental work in embryonic and larval stages. Any sufficiently large fraction of the limb level of the cord contains the full coordinative machinery for a limb; in am- phibians one-third of the normal segments is sufficient (Detwiler and Carpenter, ’29; Detwiler and McKennon, °30; Weiss, 736). Not only reduction but considerable morpho- logical disarrangement may be _ inflicted upon the spinal limb centers without abol- ishing the essentially coordinated develop- ment of their typical action systems. For in- stance, lateral halves of the limb cord resti- tuted after ablation by regeneration from the opposite half (Detwiler, ’°47b; Holtzer, 51), as well as limb segments grafted in dorso-ventral inversion (Holtzer, 50), still yield the typical limb coordination patterns. Antero-posterior reversal of the early tail- bud medulla (Detwiler, 51) or midbrain (Detwiler, 48) likewise fail to impair func- tional activity appreciably (Detwiler, °52). The existence of serial functional locali- zation in the spinal cord raises the question of the time sequence and manner of its origin. Is it already inherent in the early neural plate organization, or is it acquired only in the course of subsequent morpho- genesis as a result of intracentral segrega- tions and inductive interactions? The ap- propriate test is the standard one of hetero- topic transplantation or isolation. Overt morphological criteria, such as architecture, growth rates, cell numbers, and the like, indicate that some of the regional differences are rather firmly laid down as early as the neural fold stage (see above, p. 376). How- ever, we do not know to what extent these morphological features signify distinctive functional properties. Only a combination of morphological and functional tests can tell. Of the few thus far made, the follow- ing are pertinent. When the prospective limb segments of the cord of a urodele embryo in the neurula stage are replaced by a corresponding length of trunk cord, which would normally have remained smaller and incapable of control- 387 ling a limb, the grafted piece acquires the approximate size, as well as the functional qualifications, of true limb segments (Det- wiler, °23). Conversely, prospective limb segments shifted to the trunk region remain undersized and fail to develop the action systems for limb control (Moyer, ’43). Since this positional adaptation appears prior to the development of the limb, it must be ascribed to intracentral regulatory inter- actions, rather than to peripheral influences. If trunk segments are grafted to the brachial region in the later tail-bud stage, however, the adjustment is incomplete and limb func- tion remains defective (Table 6 in Detwiler, *36b). Evidently, the spinal action systems become fixed during that period. The spinal region anterior to the limb cen- ters, when tested at the comparable stage as before, shows less plasticity and tends to “‘self-differentiate” in disharmony with its new site (Detwiler, ’25a). In its turn, how- ever, 1t Causes certain conforming modifica- tions in the adjacent posterior host segments (judged by their enlargement, as functional tests have not been carried out). This ten- dency toward self-differentiation followed by some “inductive” influence spreading caudad is even more marked when brain parts are transplanted to more posterior sites, e.g., a supernumerary medulla oblongata in the place of anterior spinal cord (Detwiler, 125e)) It thus appears that functional localiza- tion along the spinal axis continues into post- neurulation stages, proceeding in antero- posterior sequence. Whether the determina- tion of regional cell number, the “baseline” of our earlier discussion, is effected by a separate agency from that determining func- tional pattern is uncertain. If it were, some sort of transneuronal stimulation such as that noted above (p. 383) at the junction of optic fibers and midbrain centers might be invoked (Detwiler, ’36b; see, however, the objections raised on p. 375 regarding this hypothesis). There are so few facts to go by that speculation has free rein. What is needed is more critical analytical research. Parenthetically, we may point to the po- tential usefulness of heteroplastic transplan- tation in the further exploration of this field. Since exchange of parts between dif- ferent species is feasible during the younger stages of lower vertebrates, unique test com- binations could readily be produced. Ex- change of parts of essentially similar func- tion would, of course, be less instructive than exchange of those with regard to which 388 the two species differ crucially. So far, heteroplastic work has dealt largely with the former. The limbs of a given species of salamander are readily controlled by hmb spinal cord of another species, no matter whether the strange combination is ettected by transplanting the limb to the foreign host (Detwiler, 30a; Twitty and Schwind, ‘31; Harrison, 95a) oc by substituting foreign limb cord segments for those of the host (Wieman, ’26; Detwiler, 31); even newt cord can coordinate salamander leg function (Hertwig, °26). But then, activities do not differ markedly between these species. The exchange of limbs between anurans and urodeles should be more rewarding, since the combination is feasible (Guyénot, ’27), and one could test whether or not a urodele center can make a frog leg jump. Recent studies on the behavior of animals with hybrid vestibular organs show clearly the great potentialities of such research (An- dres, *45). Taking all these sketchy fragments of in- formation together, the conclusion emerges that spinal cord and brain develop early a gross topographic mosaic of functionally specialized areas, each with specific physico- chemical and structural peculiarities not fully shared by the others, whereas each sector of this mosaic within itself manifests wide powers of regulation and substitution; so much so that any theory of coordination that would rely on a rigidly predetermined order of microconnections among neurons (rather than merely statistical regularities) seems clearly controverted by the facts of experimental embryology. It is highly in- structive that these embryological data paral- lel closely the results of work on functional localization in the cortex, which have lke- wise revealed macrolocalization of func- tional districts without microlocalization on the cellular level (Lashley, *42). It is in problems of this kind that embryology, physiology and psychology become confluent so that a conjoint approach promises to lead to much deeper insight than we now pos- sess.* The techniques of experimental mor- phology, able to manufacture crucial test situations in young animals never attainable in the adult, have hardly yet been called upon to contribute their share to this team- work. Their exploitation seems to hold rich prizes, but space restrictions do not permit us to elaborate the matter in this place. *See the Survey of Neurobiology; Publication No. 237 by the National Academy of Sciences—Na- tional Research Council, Washington, D. C., 1952. SPECIAL VERTEBRATE ORGANOGENESIS THE HUMORAL MILIEU IN NEURAL DEVELOPMENT It is beyond the scope of this article to review the specific nutrient requirements of the various components of neurogenesis. As in other organ development, the internal milieu must provide not only all factors requisite for general cell life, growth and ditferentiation, but must in addition satisfy the more specialized needs of the countless steps of which “the development of the ner- vous system” is composed. Even in its early formative stage, the CNS is already distin- guished from other organ rudiments by its different metabolic requirements (Spratt, 62). Neuropathology, on the other hand, has been able to trace many neural defects of the adult to deficiencies in the availabil- ity or utilization of nutrients, thus identify- ing the role of the latter in neurogenesis. Nerve degeneration in thiamine deficiency (beriberi) is a familiar example. Yet, the still obscure ‘“‘demyelinization” diseases prove that we do not even yet know the requirements of such a common process as myelin formation. Mental derangements have been related to the lack of certain metabolic enzymes (Hoagland, °47), but the effects of metabolic disturbances on the earlier phases of neural development have not yet been adequately explored. There is need for much more systematic investiga- tion. The condensed outline of neurogenesis presented in this article ought to have given some idea of the almost endless array of peculiar conditions of chemical and physical nature called for at every turn of this com- plicated course. To satisfy these ever-chang- ing needs, the milieu presumably under- goes phase-specific changes of composition. In this, non-neural parts or even some other parts of the nervous system may act as pro- viders of the required supplements—in a sense, as sources of specific nutriment. Yet this “symbiotic” interdependence among the tissues is practically unexplored. It can readily be seen that the role of hormones, the circularized products of en- docrine glands, is but a special case in this general category of relations. Just a few cursory remarks can be made here on this topic. For convenience, one may distinguish functional from developmental hormone ef- fects, the former modifying the performance of nervous systems that already possess their full complement of neurons, the latter in- fluencing the growth and differentiation of NERVOUS SYSTEM still incomplete nervous systems; although in view of our concept of the continuous growth of neurons (see p. 364) and the in- stability of their connections (see p. 363), particularly in the submicroscopic realm, the distinction between the two groups is apt to fade. The former group is best illus- trated by the hormonal effects on sexual behavior (see Beach, ’47), the latter by the hormone dependence of the transformation of neural structures and functions from the larval to the adult state in metamorphosing species. 50 40 WwW oO FREQUENCY 389 thyroxine carrier) near the abducens nu- cleus, which effects the central linkage (Kollros, °43a). Similarly, the morphologi- cal signs of brain metamorphosis, such as increased proliferation, cell increase and histogenetic changes, can be locally evoked (Kollros et al., 50; Weiss and Rossetti, ’51). Particularly instructive in the latter sense is the response of Mauthner’s neuron, which normally regresses during metamorphosis; if metamorphosis of the hindbrain is en- forced by the local application of a_ thy- roxine source, the nerve cell bodies enlarge NON-M-GELLS ——©—THYROIP sO THY ROXINE ——-®--CONTROLS M-CELLS 6 7 8 9 10 DIAMETER CLASSES Fig. 143. Metamorphic size changes of opposite sign in Mauthner’s cells (M-cells) and their surrounding neuron population (non-M-cells) in urodele larvae with thyroid or thyroxine-agar implants near brain. Histograms of nuclear diameters in 1000 non-M-cells each of control, thyroid and thyroxine cases, as well as mean diameters of M-cells for the corresponding groups of animals. (From Weiss and Rossetti, *51.) As outlined in Section XII, amphibian metamorphosis is under the direct control of the thyroid hormone; insect metamorphosis similarly depends upon the hormones of cer- tain head glands. The directness of the con- trol is proven by the fact that upon localized topical application of hormone, only the sur- rounding local region of the tissue undergoes transformation. The same has now been dem- onstrated for the “adaptive” changes in the nervous control mechanism which must necessarily accompany the other bodily transformations if the metamorphosed ani- mal is to function properly. That these neural changes are likewise under the pri- mary control of the hormone is revealed by the following experiment. The “wink” re- flex of the frog, which is normally not ex- ecuted by the larva although sensory and motor elements are present and individually capable of functioning (Kollros, *42), can be made to mature precociously by the im- plantation of thyroid gland (or an artificial markedly (Fig. 143) save for Mauthner’s cell, which shrinks (Weiss and Rossetti, *51). This result proves that the various neuron types respond to the hormone each in its own distinctive fashion, determined by preformed metabolic patterns, which further substanti- ates the qualitative diversity within the neu- ronal population emphasized throughout our discussion. Again these experiments are only modest openings into a rich field of future investi- gation, for in no other event of neurogenesis do we find such a favorable constellation of circumstances as in metamorphosis, where dramatic changes in neural composition, architecture and function, crowded into a relatively brief span of time, supervene in a nervous system that had already attained an advanced state of functional perfection, whose parts are distinct, relatively large and easily accessible to manipulation, and whose transformations can be set off by a _ con- trollable agent. 390 DEVELOPMENT OF BEHAVIOR No account of neurogenesis can be com- plete without relating itself to the prob- lems of behavior. After all, behavior is not only another overt sign of the molecular, cytological, morphological and functional or- ganization of the nervous system, but is its dominant manifestation to the accomplish- ment of which all other features have been made subservient by phylogenetic and onto- genetic adaptation. Unfortunately, space re- strictions do not permit me to give the topic its due treatment. However, in order to indi- cate at least some major aspects, I repeat here, with slight modifications, a summary pub- lished on an earlier occasion (Weiss, ’50c). It is a healthy sign that the sharp separa- tion once advocated between a purely phe- nomenological study of behavior, on the one hand, and the physiological study of its possible neurological foundations, on the other, has not been generally adopted. The pursuit of any one scientific field under an injunction against trespassing into another is neither rational nor productive, especially if both have common objects. It simply is not true that nothing can be learned about the “organism as a whole” by studying its constituent parts and their interrelations. On the other hand, it would, of course, be equally erroneous to assume that mere pre- occupation with the elements will tell the full story of their collective behavior. In the light of developments, it would seem unwarranted to subscribe to either a purely holistic or a purely elementarian theory of neural functions and behavior to the exclu- sion of the other, or to pursue studies on behavior alone or on its neurological foun- dations alone without the benefits that each field can derive from the advances of the other. Regardless of the pertinence of his detailed propositions, it certainly has been the historical merit of Coghill to have built a strong case for the conjoint attack on the problems of behavior and against the sepa- ratist trends of technical disciplines. The realization that much can be learned about behavior by the study of its develop- ment is of relatively recent date. But, as frequently happens in the history of science, the formation of theory outraced the acqui- sition of factual knowledge, and soon stu- dents of the development of behavior were found to be rallied around two opposite doctrines, one stressing the primacy of the holistic, the other the elementarian, view- point. Each centered its arguments on cer- SPECIAL VERTEBRATE ORGANOGENESIS tain objects, observations, and techniques different from those of the other, and evi- dently each party felt justified in consid- ering its particular niche as a fair sample of the behavioral universe. Thus what in sober evaluation would have become a fruitful stimulus to further clarification of the issues assumed the dogmatic aspect of an irreconcilable antithesis. Again, as often happens in the course of scientific history, the conflict is turning out to be a matter of one-sided viewpoints and undue generali- zations rather than of facts. Contrasting views on whether neural functions emerge as mass actions (Coghill, ’29; Hooker, °52) or in localized fragments (Windle, ’50) can be reconciled if the diversity of sample species and techniques is duly considered and if one refrains from raising observations gathered from a limited field to the dignity of doctrines of universal and unqualified validity. PHENOMENOLOGY OF THE DEVELOP- MENT OF BEHAVIOR The phenomenology of behavioral devel- opment is actually an old discipline. It started with the recognition of the fact that behavior does have a stepwise ontogenetic history, and it went on to describe the steps involved. Only in the second instance did it proceed to test the significance of the steps as instruments or causal links in the develop- ment of the whole sequence. However, ever since the demonstration that embryos raised in narcosis would develop behavioral pat- terns of normal organization (Carmichael, 26; Matthews and Detwiler, ’26), even though the overt expression of the whole series of precursor steps had been suppressed, it has been clear that the behavioral steps are merely external manifestations of under- lying intrinsic developments rather than practice steps. The complex performances of later stages cannot possibly be founded upon the tested success of their simpler precursors, since they seem none the worse for the omission of the intermediate func- tional tests due to narcosis. Again, undue generalizations must be avoided, and what is said here for the early and fundamental steps of behavioral development does not apply equally to the terminal phases, in which the inherent developmental patterns are polished and perfected by actual prac- tice and adjustment. The phenomenological study of the de- velopment of behavior has revealed that, NERVOUS SYSTEM like all development, it follows a trend from the general to the specific and from more widespread involvement of elements to more restricted and differential activation. Cog- hill’s principle of “individuation” from a background of mass reaction is based on this realization. It still holds true as designating a trend of events, even if the initial per- formance under consideration has never been a total activity of the whole body. In some cases and for some functions the pri- mordial activity undoubtedly involves all the neural apparatuses capable of function- ing at the same time (Hooker, ’52; Barron, 60), while in other cases and for some other functions, it seems equally clear that activity is territorially localized from the beginning (Windle, ’50). To give a clear-cut example of the latter type, we need only refer to the appearance of the lid-closure reflex in amphibians (Koll- ros, 42, °43b). This reflex appears only at metamorphosis after having been completely absent in the otherwise fully functional larva (p. 389). But from its very onset it constitutes a strictly localized and circum- scribed act, which has never been part of a “total pattern” of central functions. Evi- dently, individuation from mass action does not apply to this type of response, but this, in turn, does not invalidate the principle for other performances. The sobering lesson from all this work has been that we cannot find a key formula for the development of behavior which will save us the trouble of investigating each component of behavioral development in its own right. THE NEURAL BASIS OF INDIVIDUATION The development of behavior shows clearly two phases—an early expansive and a later restrictive one (Carmichael, ’33). During the expansive phase, wider and wider areas of the body come under neural control. This sequence has been clearly correlated with the gradual expansion of intracentral nerve connections and pathways (Coghill, ’29). In a sense this correlation is obvious, as it merely expresses the fact that where there is no neural pathway, there can be no neural function. Reactions during this early phase are remarkably stereotyped, indicating ab- sence of discriminative response mechanisms. However, the more the nervous system ap- proaches structural completion, the more prominent becomes its ability to activate restricted portions and patterns of the exist- 391 ing network independently in selected and coordinated combinations. It is this restric- tive “individuation” for which the proper neural correlate is still to be revealed. It had been thought that the development of limb innervation in amphibians could serve as an exquisite model of ‘“‘individua- tion” in strictly anatomical terms (Cog- hill, ’29; Youngstrom, ’40). The limbs move at first only in association with trunk move- ments, which was explained by the fact that their early innervation consists of col- laterals from the motor neurons of trunk muscles. Later “dissociation” from the trunk appeared linked to the development of a secondary separate fiber system from the limb segments of the cord (Youngstrom, ’40). Other studies (Taylor, 43), however, suggest that the so-called “primary” associated limb movements are, in reality, passive movements effected through the trunk muscles of the shoulder, while the intrinsic true limb move- ments do not appear until after the limb muscles have received the independent set of segmental “secondary” neurons. Hence the intrinsic limb function arises as a sepa- rate and individualized activity from the very first rather than as an “individuated” offshoot of an earlier mass response. Al- though some aspects of the situation are still in doubt (Herrick, °48, p. 128), it is quite evident that this singular instance cannot possibly serve as the key model for “individu- ation” in general, as originally proposed. The neural correlate of individuation thus remains as obscure as ever, nor is it very likely to reveal itself in gross microscopic features. Moreover, the progressive refine- ment and localization of coordinated function are only in part attributable to improvements in the central action systems, as the progres- sive muscle-specific modulation of effector neurons (see p. 384), establishing more dis- criminatory relations with the central action systems, undoubtedly has a share in the process (Weiss, ’36; Barron, *50). CENTRAL ORIGIN OF COORDINATION Many modern concepts of neurophysiology attempt to derive the properties of the out- put of the nervous system directly from the pattern of the sensory input. Such a concept is clearly contradicted by the studies on development. The fact that the appearance of motor performance antedates sensory con- trol has often been stressed (see Coghill, ’29; Herrick, ’48). Even more striking is the evi- dence of animals in which the development 392 of the sensory nervous system had been experimentally suppressed. The basic pat- terns of motor coordination in such anesthetic areas develop without major impairment, and limbs lacking sensory innervation from the beginning function coordinately without sensory control having ever had a chance to play a constructive part in the develop- ment of the motor patterns (Weiss, “41a; Yntema, °43; Detwiler, ’47a). Since neither learning nor patterns of sensory stimuli have any basic part in the development of orderly central functions, we must look to the auton- omous processes of central development it- self, outlined earlier in this chapter (pp. 376— 388), as the main source of coordination pat- terns. The one fact that has been conclusively established by experimental results is that the central nervous system develops a finite repertory of behavioral performances which are pre-functional in origin and ready t« be exhibited as soon as a proper effector ap- paratus becomes available (Weiss, 41a). A clear distinction must be made between the mere generation of a central discharge and the pattern of its distribution (coordina- tion). Contrary to a widespread belief, a central discharge does not depend for its generation upon afferent influx but can originate within the centers. Many instances of rhythmic automatisms of nerve centers have been reported (see Bremer, 53) and referred to underlying fluctuations in the GENIC ENDOWMENT i CELL STRAIN SPECIATION Ny eee eae Ms SUBSPECIATION DEGENERATION ROE EATON MYOTOMES ml Gore ON Nee (G I SS es ray Y See <=. oh oF ae DIFFERENTIATION py SELeCuNInN aEDeA ee =——_= es REST ee > Sit ecroocam ) ~—EAR ECTOOERM STAGE 13 DORSAL VIEW STAGE IS ANTERICA VIEW STAGEIS DORSAL viEW Fig. 145. Relative positions of ectodermal rudiments of the head in neurular stages of Amblystoma punctatum (Carpenter, ’37; Harrison, ’45). % ¢ Mets & ROMIAL py MILLI m S ||| @' @?))1| fll oS ane ™@— SOMITE © —ENTODERM S.R-SINUS ‘ RHOMBOIDALIS @-HEART % —NEURAL CREST @—CHORDA L.P-LATERAL PLATE M —HEAD MESODERM © —NEPHROS e®-ERYTHROCYTES Fig. 146. Map of prospective areas in the primitive streak blastoderm of the chick (Rudnick, °44). a mosaic its differentiation may be adversely affected by relationship with parts of the central nervous system other than the region with which it is normally related (Detwiler and Van Dyke, ’50; Detwiler, *51). It has been concluded that the medulla exerts a late morphogenetic influence on the otic vesicle. However, it can be shown that the medulla is not necessary for late differentia- tion. If the hindbrain is removed at an ear plate stage a normal labyrinth forms (Yntema, unpublished). Mechanical factors are of importance, especially the formation of a capsule (Kaan, ’38) and a functional endolymphatic sac which may regulate hy- drostatic pressure of the endolymph (Harri- son, *36a). However, presence of capsule and of endolymphatic sac do not in them- selves assure normal development (Detwiler and Van Dyke, ’50). The size of the ear in the salamander influences the number of cells in the acoustic area of the medulla (Richardson, *32), and absence of the laby- rinth in the chick results in less marked dif- ferentiation of some associated medullary nuclei (Levi-Montalcini, °49). Neurons of the auditory ganglion are derivatives of the wall of the auditory vesicle (Yntema, ’37); hence they serve as an index of differentia- EAR AND NOSE tion rather than an inductor of the ear along with the facial ganglion as held by Szepsen- wol (733). Some of the results indicated above in- volve comparisons between a system at one stage with the same system at another stage. As will be evident in later discussion, the states of one variable in such a continuum are not obvious unless other variables are 417 the four cardinal positions so that in respect to the two main axes of the ear, both, either one, or none are harmonic with the host axes (Fig. 147). The resulting vesicles may be classed as belonging to four types: (1) harmonic, (2) disharmonic, (3) redupli- cated, (4) irregular or vesicular. In trans- plants with the AP axis disharmonic, the asymmetry is reversed and a_ harmonic D D D D D V D V PIPR AIA Pia R PA PIAL Pla PIP L AIA V D Vv D Vv V V Vv Fig. 147. Scheme of operations to test the effect of changed orientation. The profile of an Amblystoma embryo (stage 21) is shown above. The curved broken line represents the embryonic axis; the stippled area, the rudiment of the ear placode; the square area, the graft. The four squares below show the positions of the axes in four orientations for the right side of the embryo; R, a graft from the right side; L, a graft from the left side. The letters inside the squares give the orientation of the graft; those outside, the direction of the cardinal points of the embryo (Harrison, ’45). held constant. The results do show a _ pro- gressive loss of ear-forming ability on the part of the system as a whole but do not demonstrate whether such loss is due to changes in one or more parts of the system. POLARIZATION OF THE EAR ECTODERM Before and during development of a mo- saic pattern the ear rudiment of Amblystoma becomes irreversibly polarized along its main axes, anteroposterior (AP) and dorsoventral (DV), as the experiments of Harrison (’24, 36a, ’36b, °45) have shown. The ear rudi- ment from a donor is grafted into the ear region of a host at the same stage of de- velopment as the donor (homostadic trans- plantation). The graft is placed in one of ear develops in a majority of experiments done when the neural folds are approxi- mated but not fused (Fig. 148A, stage 19). In later stages the disharmonic AP polarity is usually not reversed. The DV axis be- comes fixed during tail-bud stages (stages 25-27); later rotations of the DV axis tend to produce disharmonic labyrinths. Simi- larly, the AP axis is reversible in Bufo when the neural folds appear but is fixed when the neural folds are fused (Choi, ’31). If the AP axis is reversed during critical stages at the end of neurulation in Am- blystoma the polarization of its ectoderm is frequently adjusted only partially to its new environment and an enantiomorphic twin ear results (Fig. 149). “These are al- Ways mirrored across the transverse plane and may consist of either two anterior or Donor and Host Stages 15 (SEIS ZO Rela ee2s. A. Ear Ectoderm APDD to Ear Region. Harrison ‘36a Donor Stages 13 14 16 18 20 22 24 B. Ear Ectoderm APDD to Region Anterior to Ear. Host Stage 22. Yntema ‘39 Host Stages 16 18 C. Ear Ectoderm APDD to Ear Region. Donor Stage 20. Yntema, new data Donor Stages oe Ue ee [te D. Ear Ectoderm APDD to Ear Region. Host Stages 21-23. Hall ‘39 Host Stages [4 elion 6 KEY AP AXIS ae tate cabaret to Ear Region. Donor Stages 21-22. I EAR AADD Hall ‘41 sti © ©REDUPLICATION ORIGINAL ASYMMETRY EAR APDD Fig. 148. Differentiation of ear ectoderm placed APDD in various environments. A, Host and donor are at the same stage of development at the time of operation. Harmonic labyrinths usually develop from grafts in which the AP axis has been rotated before fusion of the neural folds (stage 20); in operations on older stages, the labyrinths are more frequently disharmonic. B, The AP axis is disharmonic in the heterotopic labyrinths which have recognizable axes; at stage 14 the ear ectoderm may show indication of AP polarization. The site and stage used for receiving the graft show no indication of reversing polarity to produce harmonic ears. C, The influence of the stage of the host on expression of the graft in homeotopic position is illustrated. Stage 20 is more likely to produce harmonic ears from disharmonic stage 20 ectoderm than earlier or later stages. The high incidence of reduplication when stage 14 is host indicates that this stage influences the expression of polarity more than slightly younger or older embryos. In comparison with A it is seen that reduplications occurred in the stage 20 to 20 combination more frequently than in the results of Harrison. D, A polarity of the ear ectoderm is indicated at the beginning of neurulation which may be expressed in hosts at head process stages at the time of operation. E, A reversal of polarity is accomplished in some cases by grafting older ear ectoderm to neurulae; as shown in A, this ectoderm would more frequently develop disharmonically if grafted to hosts at the same stage of development as the donor. EAR AND NOSE two posterior halves. Partial twinning, in- volving either the semicircular canals, which develop from the ventral half of the ear plate, or the saccule, which develops from the dorsal half, also occurs” (Harrison, °45, pp. 295-296). On the basis of his experiments, Harrison has concluded that the ectoderm of the ear Cir ll TWO ANTERIOR HALVES 419 of harmonic equipotentiality after fixation of the AP axis and before that of the DV axis. These observations led to further in- vestigations on the polarity of the ear ecto- derm. In the experiments indicated above, both host and donor are at the same stage of development for any one experiment. As has TWO POSTERIOR HALVES Fig. 149. Normal right labyrinth and two types of enantiomorphic twins, medial surface. The large arrow points to the posterior end of the embryo; the small arrows point posteriorly with respect to the labyrinth or half labyrinth to which they refer. c.a, Anterior semicircular canal; c.p, posterior canal; cr.a and cr.p, the corresponding sensory cristae; cr./, lateral crista, projected from the ventrolateral wall but not visible in this view; d.e, endolymphatic duct; m.1, macula of lagena; m.s, macula of sacculus; m.u, macula of utriculus; p.a, papilla amphibiorum; s.e, endolymphatic sac (Harrison, ’45). region and the surrounding area is at first isotropic about the axis perpendicular to its surface. Just as the neural folds are closing a change supervenes. The ear plate becomes polarized with respect to its AP axis. A little later conditions again change, indicating a transition from an indifferent DV axis to one which is definitely polarized. A correla- tion between the polarization of the ear ectoderm and a regularity in the molecular arrangement of the constituent cells has been postulated and an attempt to demon- strate this by x-ray diffraction, though un- successful, indicates a possible means of extending fundamental knowledge of em- bryology (Harrison, Astbury, and Rudall, ’40). In Amblystoma Hall (37) studied laby- rinths whose ectoderm had been rotated 90 degrees. In these experiments, the ear rudi- ment may still respond to some of the tests been indicated, conditions in the hosts as well as the grafts are variables. When the stage of either host or donor is constant in a series, a further analysis of the problem is possible; this has been done in hetero- stadic transplants to homeotopic and hetero- topic positions in Amblystoma (Hall, ’39, "41; Yntema, °39). The results of these ex- periments call for modifications in the con- clusions drawn from homostadic operations (Fig. 148B-E). Expression of polarity de- pends in part on the strength of polarization in the rudiment, in part on the intensity of the polarizing factors in the environment. A labile AP polarity is found in the ear rudi- ment of the early neural plate stage; this anisotropism is masked in Harrison’s experi- ments by the ability of the hosts used to reverse it. A similar inclination along the DV axis was not demonstrated. Host factors concerned with the arrangement of struc- 420 tures along the two main axes of the laby- rinth may act both before and after the critical stages for the axes indicated by Harrison. The AP polarity of the ear ecto- derm appears to differentiate gradually along with other characteristics of the ear ectoderm. Some studies on the ciliary beat of the ectoderm within the ear vesicle bear on this subject (Woerdeman, ’41). The direction of the beat is already fixed in the presumptive ectoderm of the Triton gastrula, although the stages for polarization of the main axes of the labyrinth are comparable with those of Amblystoma. Hence a polarity is demon- strable in the ectodermal region of the gas- trula. The polarization of the ear rudiment may be a local expression of a general po- larity: this is indicated by some experiments on Amblystoma (Yntema, 48) in which prospective gill ectoderm was placed apdd (anteroposterior axis inverted, dorsoventral axis not inverted) into the ear region, using donors and hosts shortly before and after the end of neurulation. As a result the for- eign ectoderm may form disharmonic or reduplicated labyrinths. Under these condi- tions polarization and induction are sep- arable processes. Fixation of polarity of the ciliary beat precedes that of the ear vesicle, as Woerdeman and others have shown. From this Woerdeman concluded that the proc- esses of polarization were not the same for the two systems. It is also possible that polarities of various organs, fixed at different stages, are differential derivatives of a basic but labile pattern in the cells involved. In such a system one derivative of polarity might be fixed, the basic polarity then re- versed, and a second derivative of the same cells would then be disharmonic with the first. The amphibian ear vesicle has been ro- tated experimentally after it has become a mosaic and its axes fixed (Spemann, °10; Streeter, 07, °14; Ogawa, °21, ’26; Tokura, "24, °25; Hall, ’37). Under some conditions the vesicle develops in its new position; under others it tends to turn back to its normal position, either partially or com- pletely. The factors which produce such a turning remain undetermined; Streeter (714, 21) has suggested some possibilities. INDUCTION OF THE AUDITORY VESICLE As Harrison (735, ’38) has pointed out, the formation of a normal ear results from SPECIAL VERTEBRATE ORGANOGENESIS the interplay of several factors: the position of the rudiment, as well as ectoderm, meso- derm, and myelencephalon, is involved. Two relations of tissues are to be noted especially: (1) that chordamesoderm comes to lie under the prospective ear ectoderm during gastrulation, and (2) that the raising of the neural folds brings the hindbrain rudiment in relation to the ear ectoderm (Yntema, 46; Ginsberg, *46). A mesodermal inductor, implied by some earlier results (e.g., Dalcq, °33; Holtfreter, 33), was demonstrated in Amblystoma by Harrison (735, °38, ’45). Belly ectoderm is grafted in place of the neural plate and folds of the hindbrain; subsequently two vesicles form on each side, a lateral one from the presumptive ear ectoderm, a medial one from the graft overlying the chorda- mesoderm which has served as the inductor. Induction by mesoderm of Rana has been reported by other workers (Albaum and Nestler, ’37; Zwilling, 41), and the ob- servations of Kogan (44) on Triton neu- rulae indicate that chorda of the hindbrain level is an active inductor. However, when chorda from the Triton neurula is trans- planted to the gastrula, posterior chorda may induce ear formation in the trunk of the host (Borghese, °42). Ability to induce ear is spread throughout the roof of the archen- teron but it is greatest at the anterior level of the notochord; heat destroys this capacity. These same cells while still part of the dorsal lip of the blastopore can induce ear. Even the uninvaginated mesoderm on the posterior neural plate of the early neurula retains this ability in Triturus (Kawakami, °43, °49). Development of vesicle in absence of adjacent brain in the chick (Szepsenwol, ’33), in Discoglossus (Pasteels, ’39), or after its early removal in the chick (Levi-Montal- cini, 46), and studies on Rana hybrids by Moore (’46) indicate that paired mesodermal inductor regions are localized in the pri- mary organizer of the early gastrula. The induction of the ear begins as a part of the primary organization occurring during gastrulation and continues through later stages. The beginning of the second period of induction is characterized by the approxima- tion of the presumptive ear ectoderm to the adjacent neural fold. The inductive capacity of the hindbrain rudiment has been demon- strated by the formation of the ear vesicle from ectoderm next to a heterotopic hind- brain in Amblystoma (Stone, ’31), and in Triton, Rana and Bombinator (Gorbunova, EAR AND NOSE °39). A developmental dependence of the ear vesicle upon the hindbrain is shown by ex- plantation experiments in Rana and Bufo (Guareschi, ’°35), and in Triton (Mangold, 37); it is also implied when auditory vesi- cles occur along with induced neural tis- 421 neural tube is associated with induction of the ear. The normal inductors of the ear are not species specific in Amphibia (Holtfreter, °35; Albaum and Nestler, ’37; Schmidt, ’38; Kogan, °39). Ears induced xenoplastically Fig. 150. Fig. 151. Fig. 150. Three-dimensional graph indicating the normality of ears induced from gill ectoderm grafted into the ear region. The two horizontal axes are determined by the duration of stages at a given tempera- ture; the vertical axis represents the normality of labyrinths induced from various combinations of donor and host stages. The view indicates primarily the results obtained when older donors are used with hosts ranging from yolk-plug to motile stages (stages 12 to 35). The peak of the graph represents the normal response of the stage 12 to 12 combination; no other combination of stages consistently produced normal labyrinths. The two ridges labelled A and B represent periods of maximal activation at stages 14 and 20, respectively. The normality of vesicles falls off as hosts and donors become more developed. Fig. 151. Another aspect of the graph, showing the more complicated results which are produced when younger donor stages are used. The face of the wall labelled C indicates the acquisition of competence to respond to neural activation by stage 18 gill ectoderm. The wall to the right of the label D indicates a marked rise by stage 13 ectoderm in competence to respond to mesodermal activation. The ridges to the left of these walls separated by a shallow recession represent the stages of maximal competence; D indicates the period of maximal competence to respond to mesodermal activation—stage 13; C indicates the same for neural activation—stage 18. The region E represents negative results obtained by placing ectoderm lacking neural competence on older hosts. The region F represents combinations in which tissue of the central nervous system is induced with or without associated ear vesicles. Since ear vesicles alone did not form, the ability to produce ears was not measured. sue, e.g., Triton (Holtfreter, °35). The re- sults of Trampusch (’41) with Amblystoma, and Barbasetti (48) with Rana indicate that migrant neural crest cells may induce ear formation. If these cells are a part of the normal induction system they may aug- ment activation by the rudiment of the hind- brain itself. Experiments on Acipenser (Ginsberg and Dettlaff, ’44) and the chick (Waddington, °37) also indicate that the may become functional (Andres, *45) and their rate of development is characteristic of the donor species (Andres, 49). The results of Woerdeman (’38) show that urodele optic vesicle or fibers arising from its retinal layer can serve as an abnormal inductor for ear formation. Other specific abnormal in- ductors of the ear may well be found, as is implied by the work of Chuang (’39) and Toivonen (740). 422 Further characteristics of induction of the ear have been obtained by replacing ear ectoderm by foreign ectoderm; grafts of prospective gill ectoderm from various stages are placed in the ear region of various stages of Amblystoma (Yntema, *50). The relative responses are represented in a three- dimensional graph (Figs. 150 and 151). Mesodermal activation is found to be great- est in the early neural plate stage and to diminish rapidly during neurulation. Neural activation appears during late neurulation and is greatest at the time of closure of the neural folds. It persists longer than the mesodermal and indications of it are present in motile stages. Periods of maximal re- sponse occur in the foreign ectoderm; these precede by short intervals the respective periods of maximal activation. In Am- blystoma, foreign ectoderm of the early neu- rula, competent to respond to mesodermal activation, is not as yet competent to respond to neural activation. This second competence is acquired during neurulation. Both com- petences, once acquired, are retained for a considerable period in decreasing intensity. The observations that the first competence is qualitatively different from the second and that ectoderm with only the first com- petence does not respond to the second or neural activation show that the two acti- vations, mesodermal and neural, are also qualitatively different. In anurans, the com- petence to respond to neural activation is present in prospective ectoderm of gastrular stages (Ponomarewa, ’38; Schmidt, ’38). The relative importance of the two types of inductions for normal development varies among different groups of amphibians (Gins- berg, ’46, 50). The normality of ears result- ing from mesodermal induction only and the increments of response brought about by neural activation can be demonstrated by removing neural plate and fold related to the ear ectoderm before and during neural induction in Amblystoma (Yntema, unpub- lished). Mesodermal induction alone results in imperfect small ears. By the time the ear plate has formed the rudiment is able to develop into a normal labyrinth without further activation by neural tube or crest. FORMATION OF CAPSULE OF INNER EAR Development of a cartilaginous capsule for the membranous labyrinth of the inner ear depends upon the presence of the latter (Lewis, ’07), and the normality of the mem- SPECIAL VERTEBRATE ORGANOGENESIS branous labyrinth depends upon presence of the capsule according to Kaan (’38). This latter relationship has been questioned by Detwiler and Van Dyke (750). In absence of the auditory vesicle, capsular cartilage does not differentiate in amphibia (Lewis, ’°07) and in the chick (Reagan, °17; Yntema, 44). In a fish, Acipenser, a solid cartilag- inous body forms in the ear region after extirpation of the vesicle (Filatow, °30). In the former cases the vesicle induces the mesenchymal cells to form cartilage. In the latter case the differentiation of cartilage is doubly assured since it occurs in absence of the vesicle and since the vesicle can in- duce a capsule heterotopically. The dependence of capsule formation has been further analyzed by heterotopic and heteroplastic transplantations of the vesicles. Following heteroplastic exchange of audi- tory vesicles, the amount of capsular car- tilage formed is influenced by the nature of both the graft and the host, e.g., Amblystoma punctatum and A. tigrinum (Richardson, 32). In all Amphibia tested a heterotopic vesicle near the ear region will induce a more or less complete capsule (Luther, ’24) and the induction is interspecific, since an anuran vesicle will induce capsule from the tissue of a urodele host (Lewis, ’07). Re- ports as to the ability of the ear vesicle to induce capsule formation in the trunk are contradictory. According to most stud- ies on urodeles, no capsular cartilage de- velops except in some instances in which the vesicle lies next to a developing vertebra and apparently induces hyperplasia of the vertebra (Balinsky, ’25; Yntema, °33). In anurans a capsule frequently appears but from no such apparent source (Filatow, ’27). In xenoplastic combinations the cartilage about anuran labyrinths in urodele flanks is anuran without contribution from the host (Balinsky, ’27; Kaan, ’30) and so has arisen from the graft, though care was taken by Kaan to exclude mesectoderm of the neural crest and mesentoderm. Heterotopic vesicles on the head of the urodele have capsules which are derived in part from the host and in part from the grafted ear ectoderm (Yntema, 39). These results indicate the conclusion that the ear rudiment or the ectoderm surrounding it may be a source of mesectoderm for the ear capsule, at least in the heterotopic position. This is contrary to the view generally held that the capsule arises from mesentoderm alone both in the normal position (e.g., Stone, °26) and heterotopically (e.g., Kaan, ’30). The mes- EAR AND NOSE enchyme which constitutes much if not all of the orthotopic capsule arises from mesen- toderm of the ear region and the anterior somites (Stone, ’26; Kucherova, ’35; Ichi- kawa, °36; Kaan, ’38). The ear vesicle has the ability to draw mesenchymal cells toward it (Filatow, ’27). In the flank these may be from sclerotomes and form capsule, or they may be from other sources and organized into a more or less com- plete limb. In the orthotopic position cells so attracted form the normal capsule (Kaan, °38). If other structures such as nasal placode, optic cup, or lens placode are substituted for the otic vesicle, no cartilaginous capsule forms (Lewis, ’07; Ichikawa, ’36; Kaan, ’38). Orientation of mesenchymal cells about a celloidin block in the ear region has been reported (Filatow, ’27). FORMATION OF MIDDLE EAR Effects of absence of the inner ear upon the development of the middle ear in Rana have been observed by Luther (’24) and Violette (28, ’30). The tympanic ring and membrane as well as the middle and external parts of the columella develop independ- ently of the inner ear. The inner end of the columella and the operculum are dependent on the presence of a normal labyrinth for their full development but only the oper- culum is absent when this factor is removed. In the chick, formation of the stapedial plate depends upon presence of the inner ear: the remainder of the columella forms after extirpation of the otic vesicle (Reagan, °17). The metamorphic transformation of the frog ear has been described by Witschi (749). Formation of the tympanic membrane of the frog depends upon the presence of the annular tympanic cartilage (Helff, 40). The quadrate, another visceral cartilage, can also induce formation of the membrane if in contact with the skin. A cartilage of the appendicular skeleton is less effective. The induction is probably brought about by some chemical product of the cartilage, since the tympanic cartilage still retains inductive influences after being killed by various methods. THE NOSE DIFFERENTIATION OF THE NASAL PLACODE Studies on the development of the nasal sac show that it undergoes the typical embryonic process of a gradually increasing differentia- 423 tion along with a progressive loss by the embryo of the ability to constitute the organ from adjacent tissue. Early differentiation of the nasal rudiment has occurred by the early neurula stage in Amblystoma (Carpenter, °37) and in Rana (Zwilling, ’40); this has been demonstrated by heterotopic transplants. Nasal sacs inde- pendent of brain tissue may form from pro- spective ectoderm of the middle gastrula when transplanted to a blastema (Emerson, 45). This result may indicate a determina- tion on the part of the ectodermal rudiment or may be due to an inductor in the ab- normal environment. In any case differentia- tion of the nasal rudiment does not wait upon formation of the neural plate and folds and its induction may start during gastrula- tion, as does that of the ear. During neurulation and later stages the rudiment acquires greater power of self- differentiation (Kawakami, °36; Zwilling, 40). Following extirpation of the nasal rudiment during neurular and head process stages, a new sac develops from the replac- ing ectoderm (Bell, ’06; Luna, *15). This ability is subsequently lost (Burr, ’16). Belly ectoderm of the early neurula has the ability to form a normal olfactory sac (Kucherova, °45). The formation of the nasolacrimal duct depends upon the presence of the nasal pit (Ogawa, ’29). INDUCTION OF THE NASAL PLACODE The problems concerned with the induc- tion of the nasal sac are peculiarly like those of the ear vesicle; they involve the questions of a mesodermal and subsequent neural in- duction. Considerable evidence points to the pres- ence of an early mesodermal induction of the nasal sac: (1) the more normal differen- tiation in explants (Emerson, ’45) and heter- otopically (Luna, °15; Cooper, ’43) in pres- ence of underlying mesoderm; (2) induction of sac by mesoderm with no induced brain (Zwilling, ’40; Moore, 46; Kawakami, ’43), and induction of sacs which have no connec- tion with brain parts, as in Bufo by Triton (Holtfreter, ’36); (3) differentiation in situ of nasal organs after removal of anterior neural plate (Spemann, ’12; Raunich, ’50); (4) formation of e nasal organ following re- moval of anterior neural fold and olfactory rudiment, and its absence if underlying mesoderm is included in the _ extirpate (Siggia, ’36); (5) the demonstration of dif- ferentiation of the rudiment prior to forma- 424 tion of the neural folds and any intimate re- lation between nasal and neural rudiments (Carpenter, °37; Zwilling, °40; Schmal- hausen, ’50). A second induction by the neural crest and fold adjacent to the nasal rudiment has been indicated by several types of observations: (1) the typical association of the nasal placode with the forebrain in cyclopia (e.g., Adelmann, °37) and in experiments pri- marily concerned with neural inductions (Raven, ’33; Mangold, ’33b; Holtfreter, ’36) ; (2) dependence for differentiation upon pres- ence of the underlying neural fold or fore- brain rudiment in heterotopic positions (Luna, 715; Siggia, °36; Kawakami, °41) and in explants (Mangold, ’33a); (3) induction from prospective gastrular ectoderm by neu- ral tissue (Woerdeman, 38); (4) induction by heterotopic forebrain rudiment from flank ectoderm (Zwilling, ’34; Kucherova, °45; Kawakami, °36). Zwilling (40), however, has carefully repeated this last experiment xenoplastically with negative results and ascribes the positive reports to inclusion of the nasal rudiment with the transplant. It may be noted that regeneration of the nasal organ occurs along with reconstitution of the forebrain in the chick (Waddington and Cohen, ’36), and that the regeneration was found by Luna (715) to be dependent upon presence of adjacent neural rudiments in the frog. The results of Siggia (’36) indicated pre- viously contradict this observation of Luna. The evidence indicates a series of two inductions of the nasal rudiment, but clari- fication of the problem is obviously needed. The normal inductors are not species spe- cific in Amphibia (Holtfreter, ’35; Zwilling, 40). As in the case of the ear, abnormal in- ductors may bring about differentiation of a nasal rudiment without accompanying brain tissue. These include the eye rudiment with or without adjacent neural ectoderm (Ikeda, *37; Woerdeman, ’38), heated mouse liver (Holtfreter, °34) and alcohol-treated guinea pig thymus (Toivonen, ’40). FORMATION OF SENSORY EPITHELIUM AND NASAL PASSAGEWAYS The olfactory ectoderm and its derivative, the olfactory nerve, can differentiate with no central nervous connection in Amphibia (Bell, ’06; Siggia, ’°38; Cooper, 43) and in the chick (Street, ’37). Fibers of the olfactory nerve have a proliferative effect on the region of the brain which they may enter in both normal and heterotopic locations (Burr, ’24a, SPECIAL VERTEBRATE ORGANOGENESIS 24b; May, ’27). Shaping of the ectodermal nasal passageways in heterotopic transplants depends upon presence of the mesectoderm surrounding the nasal pit (Cooper, °43). Moreover, most of the ectodermal nasal structures can form without nearby ento- derm. In the chick epithelia and cartilage can differentiate independently but their parts are more normal if they interact (Street, ’37). FORMATION OF NASAL CARTILAGES The dependence of the nasal cartilages upon the nasal sac in urodeles has been described in contradictory fashion by Burr (16) and Schmalhausen (739). According to the earlier work, the cartilages differenti- ate independently of the sac. In absence of the sac they collapse and join the rudiment of the trabecula, thereby enlarging it. In the latter finding the olfactory cartilages proper, that is, the capsule exclusive of its medio- basal part, form only in presence of the nasal sac and arise in part from mesectoderm de- rived from the sac itself. The size of the trabecula is not increased following extirpa- tion of the nasal rudiment but varies with the size of the adjacent brain. These findings are contradictory and call for further investi- gation. The report of Burr recalls the in- dependent formation of the ear cartilage in Acipenser; the report of Schmalhausen in- dicates an inductive action of the nasal sac on mesenchyme such as the amphibian auditory vesicle exerts. PROBLEMS Many of the general problems of embryol- ogy can be applied to the ear and nose. The nature of organization can be studied in a system apart from neural induction. Work on the specificity of inductors and the inductive effects of chemical compounds has only been initiated. The factors which produce struc- tural specialization remain unknown; in re- gard to the ear, the question exists as to why a crista forms in one part, a macula in another. Polarization of ectoderm in relation to intracellular organization has been dis- cussed but a means of analysis is as yet un- known. An outstanding lack in studies on the ear and nose is work on the structural basis of the origin of function. Some correlation would be desirable between the structure and composition of the sense organs at vari- ous stages and the functional derivatives of EAR AND NOSE those states. Some problems on the physiol- ogy of the organs have been studied in animals prepared by techniques of experi- mental embryology. The amphibians have served for the greater part of the experiments on the sense organs. Further determination of conditions in other groups would broaden understanding and afford a safer basis for generalizations. Con- tradictions and incomplete observations have been indicated in the discussion. These prob- lems offer good prospect for immediate solu- tions, and the findings are needed to secure understanding of the conditions involved. REFERENCES Adelmann, H. B. 1937 Experimental studies on the development of the eye. IV. The effect of the partial and complete excision of the prechordal substrate on the development of the eyes of Am- blystoma punctatum. J. Exp. Zool., 75:199-237. Albaum, H. G., and Nestler, H. A. 1937 Xeno- plastic ear induction between Rana pipiens and Amblystoma punctatum. J. Exp. Zool., 75:1-9. Andres, G. 1945 Ueber die Entwicklung des Anurenlabyrinths in Urodelen (Xenoplastischer Austausch zwischen Bombinator und Triton al- pestris). Rev. suisse Zool., 52:400—-406. 1949 Untersuchungen an Chimaren von Triton und Bombinator. Teil I. Entwicklung xenoplastischer Labyrinthe und Kopfganglien. Genetica, 24:387-534. Balinsky, B. I. 1925 Transplantation des Ohr- blaschens bei Triton. Roux’ Arch. Entw.-mech., 105:718-731. 1927 Xenoplastische Ohrblaschentrans- plantation zur Frage der Induktion einer Ex- tremitatenanlage. Roux’ Arch. Entw.-mech., 170: 63-70. Barbasetti, M. A. 1948 Sulle relazioni causali di sviluppo ‘“‘otocisti-cresta neurale” in _trapianti embrionali negli Anfibi. Rend. Acad. naz. Lincei, Ser. 8, 4:489-493. Bell, E. T. 1906 Experimental studies on the de- velopment of the eye and the nasal cavities in frog embryos. Anat. Anz., 29:185-194. Borghese, E. 1942 Transplantation der Chorda von Neurulen unter die prasumptive Rumpfepi- dermis mittlerer und spater Gastrulen in ver- schiedener Orientierung bei Triton. Roux’ Arch. Entw.-mech., 742:53-82. Burr, H. S. 1916 The effects of the removal of the nasal pits in Amblystoma embryos. J. Exp. Zool., 20:27-57. 1924a Hyperplasia in the brain of Am- blystoma. Proc. Soc. Exp. Biol., N. Y., 21:473- 474, 1924b Some experiments of the trans- plantation of the olfactory placode in Amblys- toma. 1. An experimentally produced aberrant cranial nerve. J. Comp. Neurol., 37:455-497. Carpenter, E. 1937 The head pattern in Amblys- toma studied by vital staming and transplanta- tion methods. J. Exp. Zool., 75:103-129. 425 Choi, M. H. 1931 Determination of the ear and side-specificity of the ear region ectoderm in am- phibian embryos. Fol. anat. Japon., 9:315-332. Chuang, H. H. 1939 Induktionsleistungen von frischen und gekochten Organteilen (Niere, Leber) nach ihrer Verpflanzung in Explantate und verschiedene Wirtsregionen von Triton-Kei- men. Roux’ Arch. Entw.-mech., 739:556-638. Cooper, R.S. 1943 An experimental study of the development of the larval olfactory organ of Rana pipiens Schreber. J. Exp. Zool., 93:415—-451. Dalcq, A. 1933 La détermination de la vésicule auditive chez le discoglosse. Arch. Anat. micro., 29:389-420. Detwiler, S. R. 1951 Recent experiments on the differentiation of the labyrinth in Amblystoma. J. Exp. Zool., 118:389-406. , and Van Dyke, R. H. 1950 The role of the medulla in the differentiation of the otic ves- icle. J. Exp. Zool., 713:179-199. Domacavalli, A. 1937 Le influenze regionali nello sviluppo dell’otocisti negli anfibi anuri. Riv. Biol., 22:245-248. Diirken, A. 1951 Die Wirkung von Ultraviolett- bestrahlungen auf die Ohranlage von Triton al- pestris. Roux’ Arch. Entw.-mech., 144:521-554. Emerson, H. S. 1945 The development of late gastrula explants of Rana pipiens in salt solution. J. Exp. Zool., 100:497-521. Evans, H. J. 1943 The independent differentia- tion of the sensory areas of the avian inner ear. Biol. Bull., 84:252-262. Filatow, D. 1927 Aktivierung des Mesenchyms durch eine Ohrblase und einen Fremdké6rper bei Amphibien. Roux’ Arch. Entw.-mech., 770:1-32. 1930 Entwicklungsmechanische Unter- suchungen an Embryonen von Acipenser giild- enstddii und Acipenser stellatus. Roux’ Arch. Entw.-mech., 722:546-583. Ginsberg, A. S. 1939 Some data on the deter- mination of the ear in Triton taeniatus. Compt. Rend. Acad. Sci. U.R.S.S., 22:370-373. 1946 Specific differences in the deter- mination of the internal ear and other ectodermal organs in certain Urodela. Compt. Rend. Acad. Sci. U.R.S.S., 54:557-560. 1950 Arteigentiimlichkeit der Anfangs- stadien der Entwicklung des Labyrinths bei Am- phibien. Compt. Rend. Acad. Sci. U.R.S.S., 73: 9299-332. , and Dettlaff, T. 1944 Experiments on transplantation and removal of organ rudiments in embryos of Acipenser stellatus in early devel- opmental stages. Compt. Rend. Acad. Sci. U.B.S.S., 44:209-212. Gorbunova, G. P. 1939 Concerning inductive capacity of medulla oblongata in embryos of amphibians. Compt. Rend. Acad. Sci. U.RB.S.S., 23:298-301. Guareschi, C. 1930 Studi sullo sviluppo dell’- otocisti degli anfibi anuri. Roux’ Arch. Entw.- mech., 722:179-203. 1935 Studi sulla determinazione dell’ orecchio interno degli anfibi anuri. Arch. ital. Anat. Embriol., 35:97-129. Hall, E. K. 1937 The determination of the axes 426 ' of the embryonic ear: an experimental study by the method of 90° rotations. J. Exp. Zool., 75:11— 39. Hall, E.K. 1939 On the duration of the polariza- tion process in the ear primordium of embryos of Amblystoma punctatum (Linn.). J. Exp. Zool., 82:173-192. 1941 Reversal of polarization in the ear primordium of Amblystoma punctatum. J. Exp. Zool., 86:141-151. Harrison, R. G. 1924 Experiments on the devel- opment of the internal ear. Science, 59:448. 1935 Factors concerned in the develop- ment of the ear in Amblystoma punctatum. Anat. Rec., 64 (suppl. 1) :38-39. 1936a Relations of symmetry in the de- veloping ear of Amblystoma punctatum. Proc. Nat. Acad., Wash., 22:238-247. 1936b Relations of symmetry in the de- veloping embryo. Coll. Net, 77:217-226. 1938 Further investigation of the factors concerned in the development of the ear. Anat. Rec. 70 (suppl. 3):35. 1945 Relations of symmetry in the de- veloping embryo. Trans. Conn. Acad. Arts & Sci., 36:277-330. , Astbury, W. T., and Rudall, K. M. 1940 An attempt at an x-ray analysis of embryonic processes. J. Exp. Zool., 85:339-363. Helff, O. M. 1940 Studies on amphibian meta- morphosis. XVII. Influence of non-living annular tympanic cartilage on tympanic membrane for- mation. J. Exp. Biol., 17:45-60. Holtfreter, J. 1933 Der Einfluss von Wirtsalter und verschiedenen Organbezirken auf die Differ- enzierung von angelagertem Gastrulaektoderm. Roux’ Arch. Entw.-mech., 127:619-775. 1934 Ueber die Verbreitung induzieren- der Substanzen und ihre Leistungen im Triton- keim. Roux’ Arch. Entw.-mech., 732:307-383. 1935 Ueber das Verhalten von Anuren- ektoderm in Urodelenkeim. Roux’ Arch. Entw.- mech., 133:427-494. 1936 Regionale Induktionen in xeno- plastisch zusammengesetzten Explantaten. Roux’ Arch. Entw.-mech., 134:466-550. Ichikawa, M. 1936 Experimental studies on the formation of the auditory capsule of amphibians. Bot. Zool., 4:1211-1223. Ikeda, Y. 1937 Beitrage zur entwicklungsmech- anischen Stiitze der Kupfferschen Theorie der Sinnesplakoden. Roux’ Arch. Entw.-mech., 136: 672-675. Kaan, H. W. 1926 Experiments on the develop- ment of the ear of Amblystoma punctatum. J. Exp. Zool., 46:13-61. 1930 ‘The relation of the developing audi- tory vesicle to the formation of the cartilage cap- sule in Amblystoma punctatum. J. Exp. Zool., 55:263-291. 1938 Further studies on the auditory vesicle and cartilaginous capsule of Amblystoma punctatum. J. Exp. Zool., 78:159-183. Kawakami, I. 1936 Self-differentiation of nose and induction of the same organ by the fore-brain SPECIAL VERTEBRATE ORGANOGENESIS in Triturus pyrrhogaster (Boie). Bot. Zool., 6: 1841-184:7, 2006-2012. 1941 Nose-inducing capacity of fore- brain of the sense-organs in Triturus pyrrhogas- ter. Zool. Mag., 53:147-157. 1943 Inductions of the cerebral sensory organs. II. Inductive effect of the archenteron roof. Kagaku, 77:399-402. III. Inductive effects of the uninvaginated portion of the dorsal blastopore lip. Bot. Zool., 77:859-862. (Japanese). 1949 Inductive effects of the heated archenteron roof and uninvaginated portion of the blastoporic lip. Siebutu, 4:41-45. Kogan, R. 1939 Inductive action of medulla ob- longata on the body epithelium of amphibia. Compt. Rend. Acad. Sci. U.RB.S.S., 23:307-310. 1944. The chordamesoderm as an inductor of the ear vesicle. Compt. Rend. Acad. Sci. U.R.S.S., 45:39-41. Kucherova, F. N. 1935 Eksperimentalnoe opre- delenie istochnikov mezenkhimy, idushchei na obrazovanie slukhovoi kapsuly. Rusk. Arkh. Anat., 14:361-370. 1945 Inductive influence of fore-brain upon body epithelium. Compt. Rend. Acad. Sci. U.RB.S.S., 47:307-309. Levi-Montalcini, R. 1946 Ricerche sperimentali sulla determinazione del placode otico nell’em- brione di pollo. Rend. Acad. naz. Lincei, Ser. 8, 1:443-448. 1949 The development of the acoustico- vestibular centers in the chick embryo in the ab- sence of the afferent root fibers and of descend- ing fiber tracts. J. Comp. Neur., 97:209-242. Lewis, W.H. 1907 On the origin and differentia- tion of the otic vesicle in amphibian embryos. Anat. Rec., 7:141-145. Luna, E. 1915 Ricerche sperimentali sulla mor- fologia dell’organo dell’olfatto negli anfibi. Arch. ital Anat. Embriol., 74:609-628. Luther, A. 1924 Entwicklungsmechanische Un- tersuchungen am Labyrinth einiger Anuren. Comment. biol., Helsingf., 2:1-48. Mangold, O. 1933a Isolationsversuche zur An- alyse der Entwicklung bestimmter Kopforgane. Naturw., 21:394-397. 1933b Ueber die Induktionsfahigkeit der verschiedenen Bezirke der Neurula von Urodelen. Naturw., 27:761-766. 1937 Isolationsversuche zur Analyse der Entwicklung der Gehér-, Kiemen- und Extremi- tatenregion bei Urodelen. Acta Soc. Fauna Flora fenn., 60:3-44. May, R. M. 1927 Modifications des centres nerv- eux dues a la transplantation de l’oeil et de l’or- gane olfactif chez les embryons d’Anoures. Arch. Biol., Paris, 37:335-396. Moore, J. A. 1946 Studies in the development of frog hybrids. 1. Embryonic development in the cross Rana pipiens 9 X Rana sylvatica ¢ . J. Exp. Zool., 101:173-219. Norris, H. W. 1892 Studies on the development of the ear of Amblystoma. I. Development of the auditory vesicle. J. Morph., 7:23-34. Ogawa, C. 1921 Experiments on the orientation Ear AND NOSE of the ear vesicle in amphibian larvae. J. Exp. Zool., 34:17-43. Ogawa, C. 1926 Einige Experimente zur Ent- wicklungsmechanik der Amphibienhérblaschen. Fol. anat. japon., 4:413—431. 1929 Eliminationsversuch der Nase bei den Amphibienlarven. Fol. anat. japon., 6:703- 710. Pasteels, J. 1939 Les effets de la centrifugation axiale de loeuf fécondé et insegmenté chez les amphibiens anoures. Bull. Acad. Belg. Cl. Sci., 25:334-345. Ponomarewa, W.N. 1938 Untersuchungen iiber die Dauer des induktiven Einflusses in der Bildung der Horblaschen. Russk. Arkh. Anat., 78:345- 352, 478. Raunich, L. 1950 Ricerche sperimentali sopra Vinduzione dell’organo olfattorio negli Anfibi Urodeli. Arch. Sci. Biol., 34:309-314. Raven, C. P. 1933 Zur Entwicklung der Gangli- enleiste. III. Die Induktionsfahigkeit des Kopf- ganglienleistenmaterials von Rana fusca. Roux’ Arch. Entw.-mech., 130:517-561. Reagan, F. P. 1917 The role of the auditory sensory epithelium in the formation of the stapedial plate. J. Exp. Zool., 23:85-108. Richardson, D. 1932 Some effects of heteroplastic transplantation of the ear vesicle in Amblystoma. J. Exp. Zool., 63:413-445. Réhlich, K. 1929 Experimentelle Untersuchun- gen iiber den Zeitpunkt der Determination der Gehorblase bei Amblystoma-Embryonen. Roux’ Arch. Entw.-mech., 778:164-199. Rudnick, D. 1944 Early history and mechanics of the chick blastoderm. Quart. Rev. Biol., 79:187- SY. Schmalhausen, O. I. 1939 Role of the olfactory sac in the development of the cartilage of the ol- factory organ in Urodela. Compt. Rend. Acad. Sci. U.RB.S.S., 23:395-398. 1940 Development of ear vesicles in the absence of medulla oblongata in amphibians. Compt. Rend. Acad. Sci. U.R.S.S., 28:277-280. 1950 Lokalisation und Entwicklung der Nasenanlage der Wirbeltiere im Zusammenhang mit der Frage ihrer Entstehung. Compt. Rend. Acad. Sci. U.R.S.S., 74:1045—-1048. Schmidt, G. A. 1938 Die morphogenetische Be- deutung des Nervensystems. I. Die Korrelation in der Entwicklung des Gehororgans. Russk. Arkh. Anat., 18:298-344, 475-477. Sidorov, O. A. 1937 Transplantation in certain Anura of the auditory vesicle in diverse stages of its development in order to discover the moment of its determination and its influence on the mes- enchyme. Russk. Arkh. Anat., 16:25-71, 145-162. Siggia, S. 1936 Contributi allo studio della de- terminazione del placode olfattivo in Discoglossus pictus. Monit. zool. ital., 47 (suppl.):116-119. 1938 Ulteriori contributi allo studio della determinazione del placode olfattivo in Discoglos- sus pictus. Monit. zool. ital., 48 (suppl.) :161- 164. Spemann, H. 1910 Die Entwicklung des inver- tierten Hoérgriibchens zum Labyrinth. Ein kri- 427 tischer Beitrag zur Strukturlehre der Organan- lagen. Roux’ Arch. Entw.-mech., 30:437-458. 1912 Zur Entwicklung des Wirbeltier- auges. Zool. Jb., Abt. 3, 32:1-98. Stcherbatov, I. I. 1938 Transplantation of audi- tory vesicle of chick embryo into chorio-allantois. Bull. Biol. Méd. exp., 6:511-514. Stone, L. S. 1926 Further experiments on the extirpation and the transplantation of mesecto- derm in Amblystoma punctatum. J. Exp. Zool., 44:95-131. 1931 Induction of the ear by the medulla and its relation to experiments on the lateralis system in amphibia. Science, 74:577. Street, S. F. 1937 The differentiation of the nasal area of the chick embryo in grafts. J. Exp. Zool., 77:49-85. Streeter, G. L. 1907 Some factors in the develop- ment of the amphibian ear vesicle and further ex- periments on equilibration. J. Exp. Zool., 4:431- 445. 1914 Experimental evidence concernin the determination of posture of the membrangjlg labyrinth in amphibian embryos. J. Exp. Zool., 16:149-176. 1921 Migration of the ear vesicle in the tadpole during normal development. Anat. Rec., 21:115-126. Szepsenwol, J. 1933 Recherches sur les centres organisateurs des vésicules auditives chez des embryons de poulets omphalocéphales obtenus expérimentalement. Arch. Anat. micro., 29:5- 94. Toivonen, S. 1940 Ueber die Leistungsspezifitat der abnormen Induktoren im Implantatversuch bei Triton. Ann. Acad. Sci. fenn., 55:1-150. Tokura, R. 1924 Zur Frage der Horblaschenin- version. Fol. anat. japon, 2:97-106. 1925 Entwicklungsmechanische Unter- suchungen uber das Hoérblaschen und das akus- tische, sowie faciale Ganglion bei den Anuren. Fol. anat. japon., 3:173-208. Trampusch, H. 1941 On _ ear-induction. Acta neerl. Morph. 4, 195-213. Violette, H. N. 1928 An experimental study on formation of middle ear in Rana. Proc. Soc. Exp. Biol., N. Y., 25:684. 1930 Origin of columella auris of Anura. Anat. Rec., 45:280. Waddington, C. H. 1937 The determination of the auditory placode in the chick. J. Exp. Biol., 14:232-239, , and Cohen, A. 1936 Experiments on the development of the head of the chick embryo. J. Exp. Biol., 13:219-236. Waterman, A. J., and Evans, H. J. 1940 Mor- phogenesis of the avian ear rudiment in chorioal- lantoic grafts. J. Exp. Zool., 84:53-71. Witschi, E. 1949 The larval ear of the frog and its transformation during metamorphosis. Zeit. Naturf., 4b:230-242. Woerdeman, M. W. 1938 Inducing capacity of the embryonic eye. Proc. konink. nederl. Akad. Wetens., 41:336-343. 1941 On the development of polarity in 428 SPECIAL VERTEBRATE ORGANOGENESIS the ectoderm of amphibian embryos. Proc. ko- ear vesicle in the salamander embryo. Coll. Net, nink. nederl. Akad. Wetens., 44:262-267. 19:30-31. Yntema, C. L. 1933 Experiments on the deter- 1948 The symmetry of ears induced from mination of the ear ectoderm of Amblystoma disharmonic ectoderm. Anat. Rec., 100 (suppl. 1): punctatum. J. Exp. Zool., 65:317-357. 95. 1937 An experimental study of the origin 1950 An analysis of induction of the ear of the cells which constitute the VIIth and VIIIth from foreign ectoderm in the salamander embryo. ganglia and nerves in the embryo of Amblystoma J. Exp. Zool., 773:211-244. punctatum. J. Exp. Zool., 75:75-105. Zwilling, E. 1934 Induction of the olfactory 1939 Self-differentiation of heterotopic placode by the forebrain in Rana pipiens. Proc. ear ectoderm in the embryo of Amblystoma punc- Soc. Exp. Biol., N. Y., 37:933-935. tatum. J. Exp. Zool., 80:1-17. 1940 An experimental analysis of the 1944 Experiments on the origin of the development of the anuran olfactory organ. J. sensory ganglia of the facial nerve in the chick. Exp. Zool., 84:291-323. J. Comp. Neurol., 87:147-167. 1941 The determination of the otic ves- 1946 An analysis of the induction of the icle in Rana pipiens. J. Exp. Zool., 86:333-342. Section VII CHAPTER 4 Limb and Girdle J. S. NICHOLAS VERTEBRATE APPENDAGES VERTEBRATE appendages are outgrowths of materials located in the body wall. The localization and determination of the con- stituent elements occur early in the history of the embryo. Morphologically, there are commonly distinguished two borders, the anterior or preaxial and the posterior or postaxial, and two surfaces, dorsal and ven- tral, which will later become. respectively the extensor and flexor surfaces of the free appendage. In the embryo, an ectodermal fin fold clearly marks the pre- and postaxial borders. The pectoral and pelvic fins of fishes and the extremities of all tetrapod vertebrates are rotated away from the primitive embryonic position during deyelopment. Frequently, the pre- and postaxial borders are not readily distinguishable in the devel- oping tetrapod limb. This is particularly so in amphibians where there is no ecto- dermal fold, the limb developing as a cylin- drical outgrowth with a rounded tip. In the lizards, an ectodermal fold is present (Mol- lier, 1895; Peter, 03; Braus, ’04a). The fold when first observed is longitudi- nal, the preaxial portion being anterior and the postaxial portion being posterior. Dur- ing the course of development, there is a tor- sion in the forelimb which brings the preaxial or radial border ventrally while the postaxial or ulnar border becomes dorsal in position. This process is reversed in the hind limb. Since the ectodermal fold is lacking in Amblystoma the pre- and postaxial borders are first recognizable in the forelimb when they have undergone a partial rotation about halfway between the assumed original rota- tion and the final degree of limb torsion. The distal portion of the extremity later des- tined to form the hand serves as an index of the torsion. The plane of flattening of the hand is inclined 45 degrees to the horizontal 429 plane of the embryo. Because of the late appearance of the limb borders, it is im- possible to ascertain by observation which radius of the limb disc represents the future radial border. It is indicated, however, in Swett’s (23) work that the material which later is distributed along the proximal border is found in the anteroventral quadrant of the limb disc not far from the center and about a radius which lies from 30 to 45 de- grees anterior to the dorsoventral axis. The materials composing the limb bud of Am- blystoma are localized much nearer to their definitive location than they are in the lizard. This seems to be the case in even the earliest limb transplantations (Detwiler, ’29). This means that as the materials grow into the free extremity they develop without torsion into a limb which is apparently partially twisted. After the digits have formed, further torsion takes place, turning the radial border to a ventral position and the ulnar to a dorsal one. THE DEVELOPMENT OF THE FORELIMB IN AMBLYSTOMA*® In the beginning tail-bud stage (Fig. 152A, stage 25) the pronephric swelling is visible and the somites may be observed through the ectoderm. There is no distinct limb bud present, but the region centered under the fourth segment just ventral to the pronephros contains the material that will give rise to the limb. When the tail bud is more marked (Fig. 152B, stage 29) the somatopleure ven- tral to the pronephros is thickened. The material is, however, a region rather than a “limb bud” on the surface of the embryo, since the chief cause of the swelling in that region is the pronephric swelling, the somato- pleural thickening merely serving to round * Modified from Harrison’s (718) description of the course of normal development of the forelimb of the spotted salamander, A. punctatum. 430 it ventrally. After this period the prominence on the side of the embryo becomes distinct (Fig. 152C, stage 33), but it is not until several days later that the extremity itself appears on the surface as a sharper elevation in the region of the fourth somite. The limb bud at first is a nodule about one and a half somites in diameter, and is SPECIAL VERTEBRATE ORGANOGENESIS being nearest the body. A little later, the first trace of the digitations appears at the extreme tip of the limb, the depression rep- resenting the notch between the first two digits (Fig. 152G). The digits elongate rapidly, as does the whole limb, but the joints are at this time not distinct. The dorsal border of the limb becomes distinctly A B Cc Fig. 152. almost radially symmetrical. It soon ac- quires greater convexity on its dorsoposterior border (Fig. 152D), and may be said to point in that direction, though the surface is rounded. From this period growth is rapid. The tip of the bud frees itself from the body wall, the axis of the limb making an angle of 30 to 35 degrees with that of the body when viewed from above, and pointing dor- sally at about the same angle to the hori- zontal. The bud elongates into cylindrical form, being attached to the body wall obliquely at its base. During this process the axis of the limb is more nearly parallel to the median plane (Figs. 152E and F). The distal part of the limb becomes flattened in a plane about 45 degrees to vertical, the dorsal border convex, and at the same time the hand is so twisted as to lie in a vertical instead of an oblique plane. The latter change is in reality partial pronation. The more lateral, which is morphologically the preaxial (radial) border, becomes ventral, the pollex lying on this side. The hand broadens and the forearm becomes somewhat flattened also. The elbow joint is now slightly flexed to- wards the ventral side (Fig. 152H). The limb is not motile, the changes being due to growth and not to muscular action. The third and fourth digits appear successively on the ulnar (dorsal) border of the hand, first as nodules which slowly elongate, the fourth being considerably behind the third in its development (Fig. 1527). The arm LIMB AND GIRDLE from above is bowed toward the body. The form changes of the limb are concerned largely with the lengthening of the various segments, notably the digits, and the more distinct demarcation of the arm, forearm and manus. Rotation takes place at the shoulder, the arm is directed more laterally and ventrally, so that the tip of the first digit rests on the substrate. Further rotation at this joint, coupled with flexion at the elbow, brings the manus much further forward beneath the gills, and the animal now rests upon two digits of each limb. The balancers, which serve to support the larva, are lost at stage 46. The first muscular movements take place at the shoulder at stage 44, and later, move- ment begins at the elbow (stage 45) and wrist joints (stage 46); the limb is then used in crawling, the positions just described be- ing those at rest. These changes are com- pleted just about the time the yolk is entirely gone and the larva begins to feed (Harri- son, *18). BASIC EXPERIMENTAL WORK ON THE LIMB The limb bud of the anuran larva con- stitutes a self-differentiating system which develops into a normal limb when trans- planted to new and strange surroundings (Braus, ’03, ’04b, 09; Banchi, ’04, ’05; Har- rison, 07). The various problems which have evolved from this fact of amphibian development have become so numerous and so widespread that they form overlapping sequences in modern experimentation. The review of only part of the limb prob- lem can be attempted here. This centers about the problems of limb development, limb specificity, limb and girdle relation- ships and reduplication. These taken to- gether form a combination of studies which have to do with the axial relations, their development, determination and_ behavior. Chronologically, the first problems to be attacked were problems of the outgrowth of the nerve fibers. Their results yielded the facts which later formed the basis for the modern concepts of nerve outgrowth (Harri- son, 715, (24): We may now attempt to locate more precisely the borders and surfaces of the definitive limb in the disc of tissue out of which it develops. Assuming them to be in the same relative position at the stage at which the operations are done as later, when these features first become visible, an oblique 431 line crossing the disc through the postero- dorsal and the anteroventral quadrants would pass through the ulnar and radial borders and hence divide the flexor from the extensor surface. The only reliable information bearing on this question is that given in Swett’s (723) paper. The posterodorsal sector forms the whole flexor surface and the distal portion of the extensor; the anterodorsal sector forms the rest of the extensor surface and extends along the radial border almost to the tip. The anteroventral portion forms the shoulder, while the posterior sector does not partici- pate in the formation of the free limb at all but enters only into the shoulder and body wall. If instead of dividing the limb bud vertically and horizontally, as in Swett’s experiments, one were to split it obliquely at the proper inclination, the cut would no doubt pass through the growing point of the limb and separate the flexor from the extensor surfaces. Balfour’s (1878) discovery of the composite nature of the elasmobranch fin raised also the question of the possible relationship be- tween the mesodermal somites and the de- veloping limb. That this relationship is variable is shown by the work of Harrison (1895) on teleosts, where a complete series of fin structures either dependent or inde- pendent of the mesoderm can be obtained. The limb will develop normally after the damage or complete extirpation of the meso- dermal somites. There is no demonstrable contribution of the somites to the muscula- ture of the limb in amphibians (Byrnes, 1898; Lewis, "10; Detwiler, ’18, ’29). Byrnes’ (1898) analysis of the situation in the frog embryo was obtained by burning the somites in the hind limb region with a hot needle. The defects in the somites pro- duced no defect in the development or muscu- lature of the limb. This experiment was repeated upon the urodele embryo by Lewis (10), using a more refined technique. He extirpated the somites, cutting out the meso- derm and observing the subsequent defects in the larva. While the ventrolateral muscu- lature was deficient, the limb developed normally, showing no defects in skeleton or musculature. Detwiler (718, ’29), in his study of girdle formation and also in his spinal cord studies, has performed identical experiments with the same _ results—per- fectly independent limb development. The forelimb region of the urodele embryo is a self-differentiating system. The materials which later form the limb can be located 432 much earlier than they can in anuran forms (Harrison, *17). The area for forelimb formation has been located in the body wall mesoderm of Amblystoma punctatum at stage 25 (Har- rison, °15, *17, °18); stage 18, high neural fold (Detwiler, °18); slit blastopore, stage 13 (Detwiler, ’29); Vogt (29) located it in Triton at the beginning of gastrulation. The time at which a definite structure can be located in the embryo has an important bearing on its usefulness in transplantation. In anuran forms the larvae are quite well developed before the limb bud can be located and removed. In Amblystoma punctatum the definitive tissues which later go into the formation of the lmb are located before any visible surface indications of the bud can be observed. This location of tissue, consisting of a portion of the somatopleuric mesoderm covered by overlying ectoderm, was originally found in the tail bud stage (Harrison, °17) (stage 29). Since that time, Detwiler (18), by the use of Nile blue sulfate transplants, was able to locate the limb area in the stage of high neural fold (stage 18), and later traced the definitive material back to the slit blastopore stage (stage 13). The limb material can then be _ trans- planted before it develops into any sort of structure. Its effect upon the organism upon which it is growing can be tested, as well as its normal and abnormal relationships to it. It affords us, then, a perfect experimental tool for studying various effects and has been used as such to great advantage. The limb disc, at the usual operating stage (29), is localized as a thickening of the somatopleure centered ventral to the fourth myotome and extending ventral to the third and fifth myotomes. The area involved is usually described as a circular area three and one- half somites in diameter (Harrison, °17, 18). The constant localization of the limb ma- terial with reference to other structures which can plainly be observed in the em- bryo is a very important factor in the ease with which the limb materials can be located. Harrison’s experiments in determining the extent of the limb-forming materials furnish an interesting example of the capacity of the limb region to regenerate. The transplanta- tion experiments test its capacity to develop in new and strange surroundings. The experiments were performed by re- moving the material (both ectoderm and somatopleuric mesoderm) from a given area. The loose mesoderm left behind after the SPECIAL VERTEBRATE ORGANOGENESIS removal of the majority of the imb-forming tissue was either caretully removed or lett im place. When the mesoderm was carefully removed there was less likelihood of re- generation. When the operative area was large the number of regenerating limbs was reduced. Atter the removal of tissues from this re- gion, the tissues torming a ring around the outside of the wound tend to migrate toward the center and torm a new limb-forming region. [This is an area of mesodermal ma- terial which under normal conditions would never form a limb but which under the imposed operative conditions moves into the region tormerly occupied by limb mate- rial. ‘This region Harrison would term as normally possessing a weak potency for limp formation. Under normal conditions it is marginal tissue. It is only under unusual circumstances of limb removal that it pos- sesses the capacity for limb formation. It is interesting that when a transplant (5 somites in diameter) larger than 34% somites is removed from the limb region and trans- planted to the flank region, and then the central portion of the graft is removed, the behavior of the ring of material around the limb is the same as though it were in normal position. A considerable number of regen- erating limbs can be secured after this pro- cedure. When a wound of 342 somites in diameter is made in the limb region and the meso- derm carefully cleaned from the floor of the wound, regeneration of a limb seldom occurs. This Harrison (18) has taken as the region which contains the preponderance of normal limb-forming material. The limb is formed by the rapid multiplication of the cells con- tained within the limb disc and not by the migration of tissues which le outside of the disc. This fact again emphasizes the inde- pendence of the limb. The histology of the so-called limb bud shows numerous mitoses while the regions surrounding it show many less karyokinetic figures. This is a histologi- cal indication of the limb independence which is proved by transplantation and also by the lack of relationships to the meso- dermal somites. When the limb disc of 31% somites is re- moved and the wound is covered by indif- ferent ectoderm, no limb develops (Harrison, 18). The extirpation of the limb region with the cleaning of mesodermal cells is suf- ficient to reduce to 14 per cent the chances of regeneration of the limb. When the limb LIMB AND GIRDLE region is removed but the wound not cleaned of mesoderm the amount of regeneration is 82 per cent. When, however, the wound is cleaned and a piece of indifferent ectoderm from any part of the body is healed over the wound, there is no regeneration of the limb. The transplantation of indifferent tis- sue has blocked the inwandering of the sur- rounding tissues and prevents reconstitu- tion. If the ectoderm of the normal limb re- gion be thoroughly freed from the closely adherent mesoderm cells, it reacts as indif- ferent ectoderm taken from other regions of the body and prevents limb growth. The material which forms the limb con- stitutes an equipotential system; any part can form the whole; two superimposed limb discs can form a normal limb (Harrison, "18, °21; Schwind, ’31). Limb equipotenti- ality is based upon the results of experiments in which (1) after the extirpation of any half of the limb bud, the remaining half gives rise to a complete normal limb; (2) two superimposed buds form a limb which at first is large but which rapidly regulates to the normal size; (3) a normal limb may develop from two ventral or two dorsal halves if they are properly oriented; (4) after inversion of the limb disc, the radial portion of the limb gives rise to the ulnar portion and vice versa, changing the pro- spective significance of practically the entire cellular constituency of the bud; (5) the inoculation of mesoderm from the limb region, even though disorganized by the operation, can give rise to a normal limb; (6) a composite limb formed of a half-limb disc of an Amblystoma tigrinum embryo transplanted upon an A. punctatum embryo in place of a half of the A. punctatum limb region gives rise to a perfect composite limb with morphological characteristics of each species recognizable. The evidence seems overwhelming for an equipotential system in Driesch’s (’05) sense. The limb-forming materials are localized in the mesoderm. This is shown by experi- ments in which the limb ectoderm is trans- planted but does not develop a limb; the mesoderm of the limb disc may be removed and the normal limb ectoderm left in its normal location, but no limb develops; the transplantation of mesoderm alone to a strange environment produces limb develop- ment (Harrison, 718, 725). The experimental proof for these state- ments is so rigorous that one would hardly expect any question to be raised with re- gard to the truth of the location. However, 433 Filatow (28) reports that there is no develop- ment of a forelimb in the axolotl after the transplantation of mesoderm alone. “In 20ige Fallen ist das Mesenchymtransplantat nach einigen Tagen verschwunden und dement- sprechend hat sich auch die aussere Vor- wolbung des das Transplantat bedeckenden Epithels vollstandig geglattet.” In two other cases, the formation of cartilages wunre- lated to the extremity was found. In these cases, the transplanted material remained for a longer time. In all others there was no histological trace of the transplant or struc- tures influenced. From these facts, Filatow considers that the materials transplanted have not yet been determined as limb form- ing. Unfortunately, the exact stage of the ani- mal is not given, but the period of develop- ment utilized extends from the _ tail-bud stage through early motile stages, in which one would expect, on the basis of Harrison’s work, that determination would have taken place. Harrison’s work receives confirma- tion from Ruud’s (726) analysis of sym- metry relation, but unfortunately this work does not employ the mesoderm alone so that we have no definite experiments which cor- respond to those of Filatow. Balinsky (731), working upon Triturus taeniatus, finds that the mesenchyme plays an important part in limb differentiation. The results are con- flicting and the experiments are not strictly comparable, since the limb and its mesoderm develop much earlier in A. punctatum than in either the axolotl or Triturus. Steiner (21) has published a short series of experiments in which he has seared the epithelial covering of the hindlimb bud of Rana. His conclusion is that the formative influence of the ectoderm acts upon the mesoderm in the production of a limb and that in this the amphibians studied are simi- lar to the higher vertebrates. The mass of evidence points to the deter- mination of the limb mesoderm of urodeles as the positive factor of limb formation. The evidence so far amassed for the anuran group, which is scattered and scanty, points to an interaction between the ectoderm and mesoderm. Development of the limb is in all cases taken at later stages than the urodele observations and absolutely no criti- cal experiments have been carried through to show the sequence which is hypothesized as causative. Until clear-cut experiments can give more light upon anuran develop- ment, it would seem wiser to continue the concept of mesodermal determination as a working hypothesis; it is a simpler view, 434 it is adequate for describing development, and it keeps one from using the term “organ- izer” indiscriminately and not in Spemann’s sense. THE GIRDLE THE PECTORAL GIRDLE In the amphibian, Harrison (718) showed that if the limb disc (3% somites in diam- eter) were removed the central parts of the scapula would not develop. The peripheral HUM, Fig. 153. S. Sc., Suprascapula; Sc., scapula; P. Cor., procoracoid; Cor., coracoid; Hum., humerus (from Harrison, 18). parts, suprascapula, coracoid and _ procora- coid, however, were represented by small stubs of cartilage with a shape fairly char- acteristic of the part which should appear in the specific region occupied by the above structures. (See Fig. 153.) This finding corroborates Braus’ (’09) results and indicates an early localization of the parts of the girdle. If, Lowever, the animals were kept for fairly long periods (ca. 100 days) the rudiments of the outlying girdle parts coalesce into a single cartilag- inous mass. This process was studied in de- tail by Detwiler (718), who described the steps in restitution of the total structure and who further circumscribed the parts giving rise to the girdle. From his results, Swett’s (23) work and some unpublished experiments of my own, a diagram can be constructed which shows the localization of parts as we now under- stand it. These are projected upon the lateral flank of Amblystoma (stage 29) as shown in Figure 154. The suprascapula develops as a good sized rudiment in the absence of the limb. This SPECIAL VERTEBRATE ORGANOGENESIS is the most constant of the developing single parts, maintaining its morphology and posi- tion with a greater degree of constancy than the other two outlying parts, both of which tend to le closer than normal to the center of where the limb would have developed. The procoracoid particularly is affected by limb disc removal and is frequently a much smaller vestige, whereas the coracoid is less affected. The suprascapular rudiment develops from an area that impinges upon the borders of the mesodermal somites; the area indicated by hatching in Figure 154 possesses a greater degree of regional deter- mination than is found in the other two parts and is least affected. The chondrogenesis of the girdle seems to follow a similar pattern in the Amphibia. Wiedersheim (1889) found the same se- quence of formation in Triturus, Siredon and Salamandra that Detwiler finds in Am- blystoma. Braus (’09) records just about the same course of events in Bombinator. There are three centers, one for the scapula, one for the coracoid and one for the procoracoid. There is normally no separate center for the suprascapula, but in the absence of the limb this does chondrify separately from the other two outlying elements. In the normal animal the girdle and its parts are readily recognizable by dissection after a light staining in toto with Ehrlich’s haemotoxylin. Before the limb has elongated and while it is still a shelf-like projection from the body wall, the preparations will show an aggregation of heavy mesenchyme which differentiates into characteristic pro- chondrin before the stage of trifurcation, which indicates the appearance of the first digit. The prochondrin differentiates directly into chondrin, and cartilage is formed about the twentieth day of larval life. At this time the ulnar digit is just beginning to form. The scapula and the outlying parts which enter into the formation of the glenoid cavity will ossify after metamorphosis, the remaining parts retaining their cartilaginous character. During development the limb and the girdle present embryonic systems in com- bination in a most interesting way. Det- wiler’s (718) work shows the mosaic char- acter of the girdle; Harrison’s (718) results prove the equipotentiality of the limb. These two systems differentiating in normal con- tiguity produce a harmonious morphologi- cal system, the appendicular skeleton in which the girdle elements, anchored as they are to the body wall, act as the regulators LIMB AND GIRDLE of the posture assumed by the free extremity. In the diagram (Fig. 154) the nearly central region of the limb area (between 3 and 5) is comparatively free of girdle-form- ing cells. Gradually the limb posture be- comes dominated by the girdle structures at its proximal extremity. This is shown by experiments in which the developing tissues around the limb area are rotated, producing a corresponding rotation of the limb without reversing its laterality. In this process of postural control the outlying rudiments must presumptively be considered as the most effective factors. When the limb is trans- 435 tated the girdle may, instead of showing its normal morphological components, de- velop as a plate-like sheet of cartilage. Since the forelimb mesoderm will grow into a forelimb under various experimental conditions, including the inversion of the mediolateral axis, why is it that a limb disc does not always develop two limbs, as experimentally demonstrated by Nieuw- koop (’46) in embryos deprived of the yolk mass? It is, of course, easy to assume a simple mechanical inhibition caused by the presence of the heavy yolk. When, how- ever, a small bit of head ectoderm is placed Fig. 154. The various portions of the girdle-forming region have been tested by transplantation experi- ments for which the parts marked out by the vertical lines (a to e) and the horizontal lines (7 to 7) were employed singly or in combination (from Detwiler, ’18; Swett, ’23; and Nicholas, unpublished). planted heterotopically with but a small part of the scapula, the imparted rotation is not corrected in later stages. The relation between the girdle and the free extremity is important also in redupli- cation, reversal of symmetry and inversion. The work of Swett (732, ’45) throws some light on these problems but much remains to be explained. Usually the girdle is ab- normal when the limb is duplicated after operations involving rotations of various de- grees. Usually after orthotopic inverted limb disc transplantations the limb is reversed in its asymmetry and so is its girdle, ie. a right limb disc upside down on the right side of the embryo becomes a left limb with a left-sided girdle. This speaks strongly for a change in the mosaic structure of the gir- dle parts, for their localized developmental potencies cannot be irreversibly fixed or determined. When areas as large as 5 somites in diameter are used for transplan- tation there is seldom formed a reduplicat- ing limb, whether orthotopically or hetero- topically transplanted. If this graft is ro- over the limb mesoderm, replacing its nor- mal covering ectoderm, a similar result is obtained, and here the mechanical effect can- not be assumed to be perceptibly greater than in the normal ectoderm. Yet there is inhibition of differentiation. The presence of the developing girdle might also be thought of as a possible factor that causes the ex- tremity to develop outside the flank instead of inside the coelom, but it lacks the ability to make the mesoderm form a free extremity when covered with head ectoderm. There are still many problems of morphogenesis to which the limb-girdle combination may con- tribute solutions. THE PELVIC GIRDLE Stultz (36) has performed a series of ex- periments which yield information regarding the development of the hindlimb and girdle of Amblystoma. The same course of events is followed as in the forelimb but the mo- saic constitution of the girdle is not so striking as in the forelimb. In his analysis 436 Stultz is dealing with a structure which is developing relatively and actually later than the forelimb and therefore will show vari- ants from its pattern. It is remarkable that the course of events is so similar when the differentiation time is so widely separated. THE LIMB AND GIRDLE IN REPTILES The amphibian embryos have served as the tool for most of the experimental analy- IX Vill Vit VI Vi VIE VU IX SPECIAL VERTEBRATE ORGANOGENESIS ses. A critical morphological study of the development of the forelimbs and_ hind- limbs and their girdles in Lacerta has been completed by Romer (742, *44). The first stage which he treats shows the develop- ment of the limb-girdle mass in an embryo of 5 to 6 mm. crown-rump length. It is more advanced in development than is the case in Amblystoma, for the mesenchyme composing the limb bud is_ perceptibly thicker and consists of three layers ar- Fig. 155 (Abbreviations). Bi — biceps Bri — )brachialis infe- rior Cb — coraco-brachia- lis Gbb ~~ —‘coraco-brachia- lis brevis Cbl |= —coraco-brachia- lis longus Cl — clavicle Cor — coracoid Decl —_— deltoides clavic- ularis Dse —-— deltoides scapu- laris Ect — ectepicondyle Ent — entepicondyle Ext — extensors of forearm Fent — entepicondylar foramen — flexors of fore- arm — humerus — tissue probably representing humero-radialis muscle — humerus — latissimus dorsi — cutaneous nerve — median nerve — radial nerve — supracoracoide- us nerve — ulnar nerve — olecranon — pectoralis — radius — subcoraco-scap- ularis — subscapularis — scapula — scapulo-humer- alis anterior — supracoracoide- us — sternum — triceps ina VI-IX — spinal nerves (From Romer, *44) LIMB AND GIRDLE ranged dorsoventrally. Mollier (1895) inter- preted these layers as derivatives of the myotomes, but Romer finds that they are completely separated from the myotomes at this stage and he is certain that their sub- sequent history is independent of the myotomes, In his stage II (Figs. 155A4—-F) Romer shows the rapid differentiation which has occurred in the limb and girdle with a slight increase (1 mm.) in length of the STAGE 4 FE J cirove 22772 yez4 GLENOID REGION P2722) V4 UPPER ARM FOREARM = wrist AND HAND f=] wees STAGE 6 A APICAL ECTODERM 437 polarized at different times in development. In Lacerta with the very rapid differentia- tion one might expect that the axes are not separable by time. LIMB AND GIRDLE DEVELOPMENT IN BIRDS AND MAMMALS Normal development of the limbs and girdles is well described in detail in birds and mammals (Hamilton, 52; Patten, ’51; Whites SS = ill (ii, : Xi == STAGE 7 C DORSAL PORTION OF CORAGCOID R_ RADIAL REGION OF -FOREARM S SCAPULA : U ULNAR REGION OF FOREARM Fig. 156. Maps showing the approximate areas for tissues of the future wing parts in stages 4 through 7. The arrows indicate the direction of the future long axes of the wing parts designated. (From Saunders, ’48.) embryo. There is a regional blocking out of the heavy condensing mesenchyme. The halves of the girdle are remote from each other and no dermal elements are present. When the embryo has reached stage III (Figs. 155 G—J), the girdle and the limb parts are plainly in evidence. The rapidity of limb growth and of its differentiation are both much greater than one would find in a similar stage in Amblystoma. Lacerta raises interesting problems; first, because of the simultaneity of development of the girdle and the limb; second, be- cause of the three layered composition of the mesenchyme which gives rise to these structures; and third, because of the speed with which differentiation in the whole appendicular skeleton occurs. One would expect results quite different from those obtained in Amblystoma, particularly with reference to the polarization of the limb axes. In Amblystoma the three axes are Bardeen and Lewis, 01; Broman, ’11). There are, however, some experimental findings which demand attention. Hamburger’s (738) implants of the wing primordium into the coelom (cf. Rudnick, ’45) showed its capacity for complete self-differentiation. By marking with carbon particles Saunders (48) has de- termined the sequential order of the develop- ment of the tissues of the wing and their relation to its future parts. Almost all of the wing bud, at stage 4, consists of the materials which will form the proximal parts of the wing and its girdle. On the basis of his experiments in which the outgrowth has been carefully observed and described, Saunders concludes that de- velopment of the limb parts is in. proximo- distal order and that this is controlled by the ectodermal cap of the wing. primordium. The results of the carbon marking experi- ments are confirmed by the remoyal -of the ectodermal cap in successive developmental 438 stages, which results in suppressing the apical growth zone. The earlier the removal of the ectodermal cap, the greater the de- ficiency in wing parts. Saunders rightly interprets his findings in stating that the apical ectoderm is essen- tial. He draws, however, too great a con- trast with the results obtained in Am- blystoma. His statement that previous work- ers have assumed a completely passive role for the ectoderm is not justified, since in no case was the role of the ectoderm important to the problems under consideration. Harri- son (718) first showed that when the limb mesoderm was removed and the ectoderm which before formed the cover to the limb mesoderm was used as a cover for the wound, no limb development occurred, al- though if the wound was left uncovered limb development frequently took place. In the light of this experiment in which meso- derm which possessed the potentiality to form limb was covered by ectoderm which would have formed the covering of the limb, it hardly seems possible that the same conditions prevail in the essentiality of the ectoderm. This is likewise shown in trans- plantation of limb mesoderm to the flank with a covering of flank ectoderm, for the limb develops with its normal asymmetry if the mesoderm is dorsodorsal in orienta- tion and has reversed asymmetry if oriented dorsoventrally. The flank ectoderm has dorsodorsal orien- tation and, if directive, should give the re- sult which Saunders obtained in transplant- ing the wing primordium with inverted orientation to the hindlimb region. The ini- tial asymmetry was maintained. As Saunders himself states, more evidence is needed on this point, and he has outlined the experi- mentation which would give the added infor- mation. Just where the mammalian limb is going to fit is problematic but there are already some leads to future work. Rothfels, in a pre- liminary unpublished study, has removed the limb primordium of the rat embryo at stage 29 and has grown the limb rudiments in tissue culture. Her results in a small group of experiments showed the develop- ment of recognizable cartilages representa- tive of complete limbs with deficiencies, in the. main, confined to the digits. The defi- ciencies were about the same whether the ectoderm was present or had been removed from the primordium before culture. More recently, Moscona (’51) performed dissociation experiments using trypsin to SPECIAL VERTEBRATE ORGANOGENESIS separate the cells of chick limb buds after removal of the dorsal and ventral ectoderm. The dissociated cells were then cultured. Chondrogenic and myogenic cells tended to aggregate, forming an inner nodule of car- tilage around which the layer of myoblasts was formed. This is, of course, a far different experiment from that of Saunders, with the chick in place upon its substrate of the yolk mass, but the conditions of the ex- periment seem to throw some light on the capacity of the dissociated cells to attempt to complete their potential destiny without the ectoderm acting in an essential manner. The problems of the limb, regarded as solved by all too many, are still with us. The present paradoxical results will be reconciled by those who will complete the thoughtful analysis of future experimen- tation. REFERENCES Balfour, F. M. 1878 A Monograph on the Devel- opment of Elasmobranch Fishes. Macmillan and Co., London. Balinsky, B. I. 1931 Zur Dynamik der Extremi- tatenknospenbildung. Roux’ Arch. Entw.-mech., 123:565-632. Banchi, Arturo 1904 Sviluppo degli arti ad- dominali del Bufo vulgaris innestati in sede ano- mala. Monitore Zool. Ital., 15:396-399. 1905 Sviluppo degli arti pelvici del Bufo vulgaris innestati in sede anomala. Arch. Anat. Embriol., 4:671-693. Bardeen, C. R., and Lewis, W.H. 1901 Develop- ment of the limbs, body wall, and back in man. Am. J. Anat., 7:1-36. Braus, H. 1903 Versuch einer experimentellen Morphologie. Miinch. mediz. Wochenschrift, 47: 1-5. 1904a Die Entwickelung der Form der Extremitaten und des Extremitatenskeletts; in Hertwig’s Handbuch der Entwickelungslehre der Wirbelthiere, Bd. 3, T. 2, pp. 167-338. Gustav Fischer, Jena. 1904b Eimige Ergebnisse der ‘Trans- plantation von Organanlagen bei Bombinator- larven. Verhand. der Anat. Gesellschaft, Jena, pp. 53-65. 1909 Gliedmassenpfropfung und Grundfra- gen der Skelettbildung. I. Die Skelettanlage vor Auftreten des Vorknorpels und ihre Beziehung zu den spateren Differenzierungen. Morphol. Jahrb., 39:155-301; also in Exp. Beitr. Morphol., 1:284-430. Broman,I. 1911 Normale und abnorme Entwick- lung des Menschen. J. F. Bergmann, Wiesbaden. Byrnes, Esther F. 1898 Experimental studies on the development of limb-muscles in Amphibia. J. Morph., 14:105-140. Detwiler, S. R. 1918 Experiments on the devel- opment of the shoulder-girdle and the anterior LIMB AND GIRDLE limb of Amblystoma punctatum. J. Exp. Zool., 25:499-538. Detwiler, S. R. 1929 Transplantation of anterior- limb mesoderm from Amblystoma embryos in the slit-blastopore stage. J. Exp. Zool., 52:315- 324. Driesch, H. 1905 Die Entwickelungsphysiologie von 1902-1905. Ergeb. Anat. Entwick., 74:603- 807. Filatow, D. 1928 Ueber die Verpflanzung des Epithels und des Mesenchymes einer vorderen Extremitatenknospe bei Embryonen von Axolotl. Roux’ Arch. Entw.-mech., 773:240-244. Hamburger, V. 1938 Morphogenetic and axial self-differentiation of transplanted limb primor- dia of two-day chick embryos. J. Exp. Zool., 77: 379-400. Hamilton, H.L. 1952 Lillie’s Embryology of the Chick. Henry Holt and Co., Inc., New York. Harrison, R. G. 1895 Die Entwicklung der un- paaren und paarigen Flossen der Teleostier. Arch. mikr. Anat., 46:500-578. 1907 Experiments in transplanting limbs and their bearing upon the problems of the devel- opment of nerves. J. Exp. Zool., 4:239-282. 1915 Experiments on the development of the limbs in Amphibia. Proc. Nat. Acad. Sci., 7: 539-544, 1917 Transplantation of limbs. Proc. Nat. Acad. Sci., 3:245-250. 1918 Experiments on the development of the forelimb of Amblystoma, a self-differentiat- ing, equipotential system. J. Exp. Zool., 25:413- 462. 1921 On relations of symmetry in trans- planted limbs. J. Exp. Zool., 32:1-136. 1924 Some unexpected results of the heteroplastic transplantation of limbs. Proc. Nat. Acad. Sci., 10:69-74. 1925 The effect of reversing the medio- lateral or transverse axis of the fore-limb bud in the salamander embryo. Roux’ Arch. Entw.- mech., 106:469-502. Lewis, W. 1910 The relation of the myotomes to the ventro-lateral musculature and to the anterior limbs in Amblystoma. Anat. Rec., 4:183-190. Mollier, S. 1895 Die paarigen Extremitaten der Wirbeltiere. II. Das Cheiropterygium. Anat. Hefte, 5: T. 31-38, S. 435-529. Moscona, A. 1951 Cell suspensions from organ rudiments of chick embryos. Exp. Cell. Res., 3: 535-539. Nieuwkoop, P. D. 1946 Experimental investiga- 439 tions on the origin and determination of the germ cells, and on the development of the lateral plates and germ ridges in urodeles. Arch. Néerl. Zool., 8:1-205. Patten, B. M. 1951 Early Embryology of the Chick. Blakiston Co., Philadelphia. Peter, K. 1903 Mitteilungen zur Entwicklungs- geschichte der Eidechse. IV und V. Die Extrem- itatenscheitelleiste der Amnioten und die Anlage der Mitteldarmdriisen. Arch. Mikr. Anat., 67: 509-536. Romer, A. S. 1942 The development of tetrapod limb musculature—the thigh of Lacerta. J. Morph., 71:251-298. 1944 The development of tetrapod limb musculature—the shoulder region. of Lacerta. J. Morph., 74:1-41. Rudnick, D. 1945 Differentiation of prospective limb material from Creeper chick embryos in coelomic grafts. J. Exp. Zool., 700:1-17. Ruud, G. 1926 The symmetry relations of trans- planted limbs in Amblystoma tigrinum. J. Exp. Zool., 46:121-142. Saunders, J. W. Jr. 1948 The proximo-distal se- quence of origin of the parts of the chick wing and the role of the ectoderm. J. Exp. Zool., 108:363- 403. Schwind, J. L. 1931 MHeteroplastic experiments on the limb and shoulder girdle of Amblystoma. J. Exp. Zool., 59:265-295. Steiner, H. 1921 Hand und Fuss der Amphibien, ein Beitrag zur Extremitatenfrage. Anat. Anz., 53:513-542. Stultz, W. A. 1936 Relations of symmetry in the hind limb of Amblystoma punctatum. J. Exp. Zool., 72:317-367. Swett, F. H. 1923 The prospective significance of the cells contained in the four quadrants of the primitive limb disc of Amblystoma. J. Exp. Zool., 37:207-218. 1932 Reduplications in heteroplastic limb grafts. J. Exp. Zool., 67:129-148. 1945 The role of the peribrachial area in the control of reduplication in Amblystoma. J. Exp. Zool., 100:67-77. Vogt, W. 1929 Gestaltungsanalyse am Amphibi- enkeim mit 6rtlicher Vitalfarbung. II. Gastrula- tion und Mesodermbildung bei Urodelen und Anuren. Roux’ Arch. Entw.-mech., 720:384-706. Wiedersheim, R. 1889 Uber die Entwicklung des Schulter- und Beckengiirtels. Anat. Anz., 4:428- 441, Section VII CHAPTER. 5 Heart, Blood Vessels, Blood, and Entodermal Derivatives W. M. COPENHAVER HEART AND BLOOD VESSELS THE CARDIOVASCULAR SYSTEM attains functional importance relatively early in the life of an embryo and consequently the prefunctional phase of cardiovascular development is com- pleted rapidly. Furthermore, there is gener- ally a close correlation of morphogenesis and functional differentiation in the later stages of development. For example, as the heart chambers differentiate morphologically in a cephalocaudal direction (from conus toward sinus venosus) there is a corresponding shift in the location of the pacemaker and an ac- celeration in pulsation rate. On the other hand, experimental studies show a lack of correlation between functional and morpho- logical development in the earlier stages. Pre- sumptive heart mesoderm grown either as an explant or as a heterotopic transplant can de- velop pulsations without regard to the shape of the developing heart, and the pattern of the main blood vessels (aorta, gill vessels, etc.) can develop to a considerable extent in em- bryos experimentally deprived of circulation. These are a few of the points which require consideration under their respective headings below. Localization of the Presumptive Heart Rudi- ment. By using either vital dyes or surgical procedures, presumptive heart material can be located before there is any observable morpho- logical differentiation of a heart rudiment. In amphibian embryos, the position of the pre- sumptive material has been identified as early as the gastrula stage by tracing the fate of vitally stained regions (Vogt, ’29) and by iso- lation and extirpation procedures. The ap- proximate locations of heart-potency material in urodele amphibians at successive stages are shown in Figure 157. The boundaries of the presumptive heart material have not been studied so precisely for amphibians as they have been for the 440 chick by Rawles (43). This stems from the fact that the studies bearing on heart localiza- tion in early amphibians have not been de- vised primarily for ascertaining either the exact boundaries of, or the degree of potency of subdivisions of, the heart-forming area. For example, isolation experiments related to lo- calization have been concerned primarily either with the over-all distribution of organ potency areas in gastrula and neurula stages (Holtfreter, ’38; Mangold, ’37) or with spe- cific problems of heart self-differentiation (Ekman, ’21, ’27; Stohr, ’24a; Goerttler, ’28; Bacon, *45). The first visible indications of heart pri- mordia in urodele amphibians are found in early tail-bud stage embryos (Harrison’s stage 25). Figure 158 shows the amount of differentiation at stage 27. In anuran am- phibians, the heart primordia develop earlier than in urodeles. For example, the presump- tive heart material of Bombinator embryos completes the migration shown for urodeles in Figures 157 and 158 when the neural groove has just closed (Ekman, ’24). These differences should be remembered when one compares results of experiments on different forms. In birds and mammals, where considerable cardiac morphogenesis occurs before the lat- eral rudiments unite, the heart primordia are observable earlier than in amphibians. In man, a cardiogenic plate appears in the presomite stage (Davis, 27). In the chick, the pericardial coelom appears at the one or two somite stage (Rawles, ’43). The localiza- tion of presumptive heart material in chick embryos before the primordia are observable has been studied by transplanting small pieces of blastoderm either to tissue culture media (Olivo, ’28; Rudnick, ’38; Spratt, ’42) or to the chorioallantoic membrane (Willier and Rawles, °31; Hunt, ’32; Butler, °35; Rawles, ’43). Figure 159, after Rawles (’43), HEART, BLOOD VESSELS, BLOOD, AND ENTODERMAL DERIVATIVES 44] forming Mesoderm - Entoderm Presumptive Blood Cells Intestine Esophagus, Stomach Fig. 157. Location of presumptive heart material in Amblystoma. A, Organ-forming areas mapped on the ventral portion of an opened gastrula, dorsal view; B, gastrula, lateral view; C, neurula, stage 15, right lateral view; D, stage 22. Heart-forming areas on C and D are outlined by broken line. EF, transverse section at level of arrow in D; F, enlargement of ventral portion of same section. (A and B, redrawn after Holtfreter, °38; C, redrawn after Bacon, *45.) shows the location of bilateral areas of an early chick blastoderm which have the po- tency for differentiating into typical cardiac muscle when divided into small pieces and transplanted to the chorioallantoic mem- brane of older host embryos. The grafts which contained parts of the areas shown by the darkest shading in Figure 159 produced heart muscle with the greatest frequency. Heart-Forming Field and Induction. From the experiments cited above and from others which we will consider now it is evident that the heart resembles other organs in the sense that its field is more extensive than its presumptive material. It has been shown for both anurans (Ekman, ’21) and _ urodeles (Copenhaver, ’26) that a normal heart can develop. after the complete extirpation of a visibly demarcated heart primordium (Fig. 158). In Amblystoma, the heart-forming po- tency of the neighboring mesoderm is lost in late tail-bud stage embryos (after Harri- son’s stage 29); in Bombinator, it is lost earlier in correlation with the relatively earlier heart differentiation in anurans. The extent of mesoderm with heart-form- ing potency at the stage when the heart primordia are first visibly indicated has been studied by Ekman (’25). He found the meso- derm from the gill region can be induced to 442 take part in heart formation when it is transplanted into the mid-ventral area be- tween the bilateral heart primordia, but mesoderm from a more caudal region can not. Ekman considered this an example of true induction. His experiments pointed to C Presumptive SPECIAL VERTEBRATE ORGANOGENESIS tiation was not a factor in the results. The evidence indicates the existence of a heart inductor in the entoderm of the definitive heart region. In further support of this view, Bacon found that indifferent presumptive mesoderm from a beginning gastrula is in- ~ Pericardial Cavity Presumptive Endocardium Myocardium Fig. 158. Heart localization in a tail-bud, stage 27, Amblystoma punctatum embryo. A and B, ventral and lateral views respectively; C, ventral portion of a transverse section at level indicated on B. Extirpation of all mesodermal tissue over areas outlined on A and B, extending from X to Y in section C, prevents heart formation. Extirpation of presumptive heart material, extending from Z to W in section C, prevents heart development only when the wound is filled with foreign tissue. Otherwise, mesoderm in XZ and YW areas migrates ventrad and exhibits heart potency. the existence of an inductor in the heart primordium, but other experiments have shown that the inducing activity is not lim- ited to the heart itself. It has been shown by Bacon (745) that a normally structured heart can be organized when the heart rudiment of tail-bud stage Amblystoma is replaced with indifferent presumptive mesoderm from the marginal zone of a beginning gastrula (stage 10). The substituted tissue did not in- clude any presumptive heart material (it did not contain the lateral quadrants of the mar- ginal zone) and therefore heart self-differen- duced to form cardiac tissue when it is grown in cultures with archenteron floor from a late neurula. The organizing power of the heart region is unable to overcome the determina- tion present in most of the mesoderm after gastrulation—at least, it is unable to induce heart from presumptive somite mesoderm from neurula stages (Bacon, *45). On the other hand, mesoderm immediately adjacent to the heart region is still competent to form heart in tail-bud stage Amblystoma embryos, as indicated in Figure 158. Is the presumptive heart mesoderm itself HEART, BLOOD VESSELS, BLOOD, AND ENTODERMAL DERIVATIVES 443 dependent upon a secondary entodermal in- ductor for its differentiation? Nieuwkoop (46) found an absence of the heart in Triton embryos which developed after complete removal of the entoderm from early neurula. Bacon (personal communication) has ob- tained similar results for Amblystoma oper- ated upon at Harrison’s stages 15-18, and Balinsky (’39) has reported similar findings for neurula stage Tritons. One finds it diffi- cult to explain the lack of heart development in entodermless embryos in view of other evidence favoring ability of very early stages for cardiac self-differentiation. Heart Determination and _ Self-Differentia- tion. The first experimental evidence indi- cating the stage from which the presumptive heart-field can self-differentiate into an or- ganized heart was obtained by Ekman (’21). Using Bombinator embryos at the stage when the heart is first visibly indicated, he ex- planted the heart primordium in an ecto- dermal covering. Using similar methods and materials, Stohr (’24a) concluded that the heart differentiation in the explants was correlated with the presence of neigh- boring entodermal and mesodermal cells re- moved with the heart. Further evidence on the problem has been obtained from both explantations (Ekman, ’24, ’27, ’29; Bacon, ’45) and heterotopic transplantations (Stohr, ’24b; Copenhaver, ’26). The former method permits a study of heart formation free of confinement and association with other tis- sues; the latter method permits a study of the organ freed only from the structures with which it is normally associated but it gives more differentiation than the isolation method and enables one to identify the dif- ferent heart chambers with more assurance. Stohr’s results with this method reempha- sized his belief in the importance of “Endo- mesodermzellen” which were always present in successful transplants. On the other hand, heterotopic transplants on tail-bud stage Amblystoma embryos (Copenhaver, ’26) showed that well formed pulsating hearts can develop from presumptive heart meso- derm without the normally associated ento- mesodermal structures. Particularly strong evidence for the self-differentiating capacity of the heart has been obtained with the ex- plantation method by Bacon (’45), as shown in Figure 160. The explantation method has been em- ployed in studying the time of heart deter- mination in both anurans (Ekman, ’27, ’29) and urodeles (Bacon, °45). Bacon’s results indicate that the presumptive heart material of Amblystoma is determined and capable of self-differentiation into typical parts in the crescent blastopore stage (stage 11). But the explants at this stage consisted of a piece of gastrula wall which may have in- cluded neighboring entomesodermal cells. The evidence for self-differentiation is un- equivocal in the experiments done on medul- lary plate stages (stages 13-15) since only mesoderm was explanted. Thus the heart primordium is self-differentiating for a con- Fig. 159. Map of heart-specific areas on a head- process stage chick blastoderm. Numerals at right show distances in millimeters from the level of the primitive pit. (After Rawles, ’43.) siderable time before the environment of the cardiac region loses its ability to induce heart formation in indifferent mesoderm, as cited earlier. The heart primordium appears to have a “labile determination” for curvature at an early stage. Some studies have suggested that the heart possesses merely a tendency for curvature since the particular form ob- served under experimental conditions is vari- able and obviously influenced by the en- vironment (Spemann, ’06; Pressler, ’11; Ek- man, 724; Stéhr, ’24a, 25; Copenhaver, ’26). But the results obtained by Bacon (45) for hearts explanted without confinement show a determination for the characteristic S-type of curvature. This does not deny the view that the S-shape is suited to the environ- ment in which the heart normally develops nor does it contradict other evidence that the 444 curvature can be influenced by the shape of the pericardial cavity and by other mechani- cal factors. In birds, the cardiac primordia have greater potentialities for histological than for morphological self-differentiation. Small pieces of chick blastoderm will develop typi- cal cardiac muscle with rhythmical pulsa- tions in culture (Olivo, ’28), but they do not 26 Hrs. | Hr. 23 Hrs. SPECIAL VERTEBRATE ORGANOGENESIS amount of morphological self-differentiation for the rat heart, but it must be noted that neither in mammals nor in birds has morpho- logical differentiation been demonstrated for cardiac primordia completely isolated from other embryonic tissues. An abundance of evidence could be cited to show that structure and function are not dependent upon each other during the early 7S nts. 84 Hrs. B 5O Hrs. Fig. 160. Self-differentiation of presumptive heart material cultured in modified Holtfreter’s solution. A, Amblystoma punctatum, stage 11. Explant of gastrula wall from outlined area lateral to the blastopore differentiates as shown in B. Arrows show direction of contraction wave at 84 hours after explantation. C, differentiation of presumptive mesoderm explanted from an embryo of stage 15, late medullary plate stage. (Redrawn after Bacon, ’45.) show morphologically differentiated hearts comparable to those formed by amphibian explants. Chorioallantoic grafting of heart- forming areas of the chick blastoderm gives more cardiac morphogenesis than in vitro cultivation does, but morphogenesis is still inferior to histogenesis (Kumé, ’35; Rawles, 43). Rawles suggests that the atypical morphogenesis may be the result of abnormal mechanical conditions rather than a restric- tion of the determination process. In mammals, as in birds, the tendency for cardiac self-differentiation is greater physio- logically than morphologically, as shown by the behavior of whole blastocysts cultivated in vitro (Waddington and Waterman, °33; Nicholas and Rudnick, ’34; Goss, ’35). The studies by Goss have shown a considerable development of the heart. Bacon (45) finds explanted amphibian hearts may differen- tiate morphologically without developing pulsations, whereas contractions may occur in explanted rudiments lacking morphologi- cal differentiation. Numerous studies al- ready cited on_ self-differentiation have shown that a lack of blood circulation does not prevent the heart from developing well beyond the stage which it normally attains at the beginning of circulation. On the other hand, there is evidence that the blood stream has an influence on the later stages of heart differentiation in both amphibians and birds (Stohr, ’25; Bremer, ’31). In the chick heart, a valvelike action occurs at the a-v junction through an “endocardial mound” of cardiac jelly before the establishment of the pri- HEART, BLOOD VESSELS, BLOOD, AND ENTODERMAL DERIVATIVES 445 mordia for the leaflets of the valve (Patten, Kramer and Barry, °48). The authors sug- gest that the moulding effect of the spiral streams of blood, as described by Bremer, may act particularly on the cardiac jelly and thus “set the pattern followed by the endocardial cushion tissue.” Polarity. The determination of polarity in the heart rudiment has been studied for several species of amphibians with some- what difterent results. It is reported that 180 degree rotation of the anteroposterior axis of the heart rudiment in Bombinator does not interfere with normal development when the operation is made in embryos with a just-closed medullary tube, or earlier (Ek- man, “21; Stohr, °25). A similar operation at the tail-bud stage produces a reversed heart, indicating axial determination (Stohr, 25). In Rana fusca, the anteroposterior cardiac axis appears to be determined some- what earlier than in Bombinator (Ekman, 29). On the other hand, it is reported that the heart of R. nigromaculata still lacks anteroposterior axial determination when it is approaching a tubular stage (Ota, ’30). However, since Ota reports only three posi- tive cases out of 783 experiments one may well question whether the hearts of the positive cases developed from the rotated rudiments or whether they formed by re- generation from nonrotated neighboring tissue. In Amblystoma, the anteroposterior axis has been rotated 180 degrees either by a single operation at tail-bud stages or by a bilateral operation at stage 22, shortly after closure of the neural tube (Copenhaver, 26). Both types of operations produced atypical, reversed hearts. When one recalls that the heart of a stage 25 tail-bud Am- blystoma is differentiated only to a degree comparable to that of a neural fold Bom- binator, it appears that the anteroposterior axis of the heart is determined much earlier in the former species than in the latter. It is suggested that the determination of the anteroposterior cardiac axis should be re- studied to learn whether the differences de- scribed for Bombinator and Amblystoma are well founded and whether they exist for anurans and urodeles generally. Determination occurs later for the dorso- ventral and transverse axes than for the anteroposterior axis. When the ventral half of the heart rudiment of Bombinator is ro- tated dorsoventrally, up to beginning tail- bud stages, the parts unite to build a func- tioning heart (Ekman, ’29). Likewise when a right half of the heart rudiment of a tail- bud Amblystoma is replaced with a lett halt, thus changing the mediolateral axis of a half, the parts combine to form a function- ing heart (Copenhaver, ’26). ‘Yotipotency of the Heart Rudiment. There is considerable evidence that an entire heart can develop from a part of its rudi- ment and that each part is therefore equi- potential in its early developmental stages. One line of evidence comes from duplica- tions occurring in nature. A most remark- able case of this type was reported by Verocay (05), of a hen which had seven hearts of approximately equal size. It also became evident from some of the earliest studies in the field of experimental embryol- ogy that the heart rudiment is plastic and not a fixed mosaic. It was shown in chick embryos that the bilateral rudiments for one heart can develop into two hearts when they are prevented from uniting (Graper, 07). In frogs, it was found that the rudi- ments for two hearts exhibit various degrees of fusion when embryos are joined along the ventral region (Born, ’97). Extensive experimental studies showing totipotency in the amphibian heart rudiment have been made by Ekman (’21) on embryos of Bombinator. He found that a functioning heart will develop after removal of a lateral half and that two hearts, each with circu- lation, can develop from one rudiment split lengthwise. These studies were confirmed on tail-bud stage Amblystoma embryos (Copenhaver, ’26). They were also extended to show that a functioning heart can develop after removal of an anterior half rudiment or from combinations of two anterior or two posterior halves. On the other hand, Stohr (27) concluded from studies on Bombinator and Amblystoma that the heart does not fulfill the requirements of an equipoten- tial system. He found that a heart develop- ing from a lateral half rudiment generally has an atypical shape for one or more of its chambers but some of the illustrated cases approximate the normal to such de- gree that one may question whether they argue against equipotentiality of a half rudi- ment or whether they indicate merely a labile determination of form. Remarkably normal hearts from lateral half rudiments in Amblystoma have been described by Fales (46). Evidence for equipotentiality of anterior and posterior halves may be more question- able than that for lateral halves. In the lat- ter case, the four embryonic divisions are 446 represented in each half rudiment whereas in the former case a presumptive ventricle must form atrium or vice versa. It has been noted previously that the technique is more difficult for combinations of posterior halves than for lateral halves. In order to eliminate the possibility that a transplanted half is resorbed and superseded by host mesoderm, the transplantations should be made hetero- plastically on species with pigment differ- First Aortic Arch Ventricle Atrial Primordia SPECIAL VERTEBRATE ORGANOGENESIS a normal asymmetry whereas the right one frequently develops as a mirror image of the normal. This is seen particularly well in amphibians where experimentally pro- duced double hearts undergo complete functional differentiation, with circulation through each heart. Factors responsible for the situs inversus of the right member of a double heart are not clearly understood. Further questions on cardiac asymmetry will Fig. 161. Progressive fusion of paired primordia during morphogenesis of chick heart. A, at 9-somite (+ 29 hour) stage, when first contractions appear; B, at 16-somite (+ 38 hour) stage, when blood circula- tion begins. (Redrawn after Patten and Kramer, ’33.) ences in the embryonic tissues (e.g., T7riturus taeniatus and T. cristatus). Experiments by Goss (’35) show that some totipotency exists in the primordium of the mammalian heart. He found that double hearts develop in 9-day-old rat embryos in hanging drop cultures when the lateral primordia are prevented from uniting. The lateral rudiments self-differentiate beyond the stage when they unite in normal de- velopment. Available experimental methods do not maintain the growth of separated mammalian heart primordia for sufficient time to show whether a half can develop a fully differentiated mammalian heart with two atria and two ventricles. When double hearts develop from sepa- rate bilateral primordia, the left heart has be considered in relation to asymmetry of entodermal derivatives. Formation of the “Tubular” Heart and Functional Differentiation. Fusion of the bi- lateral heart primordia occurs in a cephalo- caudal sequence, with formation of the bulboventricular region first and the sinus venosus last. Space does not permit reference to all of the studies which have established this point for amphibians, birds and mam- mals, including man (Davis, ’27). Likewise, it has been demonstrated that functional dif- ferentiation progresses in a cephalocaudal direction. Sabin (’20) and Johnstone (725) appear to have been the first to note that the earliest contractions in the chick heart begin in the ventricle. Using a cinemato- graphic method, Patten and Kramer (733) HEART, BLOOD VESSELS, BLOOD, AND ENTODERMAL DERIVATIVES have extended these studies to give us a detailed account of the series of changes by which the sinus venosus is eventually established as the pacemaker. An excellent review of this work has been presented by Ventricle 447 birds, mammals exhibit less fusion and am- phibians more. In urodele amphibians, the so-called “tubular” heart of the beat initia- tion stage contains the future conus, ven- tricle and cephalic portion of the atrium Fig. 162. Heart differentiation in Amblystoma punctatum at stage 34, when first contractions appear. A, heart in situ after removal of ectoderm and ventral portion of parietal pericardium. B, enlargement of a portion of A. The outlined area on the bulboventricular region shows extent of contractions observed before fixation. X and Y indicate levels of sections illustrated in C and D, respectively. Note abundance of yolk droplets and lack of striations in myocardium. Patten (’49). Figure 161 shows some of the morphological stages in the cephalocaudal differentiation. Studies on the genesis of cardiac contrac- tions in mammals (Goss, ’38; Dwinnel, ’39) and in amphibians (Copenhaver, ’39a) have given results which differ in only a few respects from those described for the chick. The degree of fusion of the bilateral pri- mordia at the time of beat initiation varies for different animals; in comparison with (Fig. 162B). At this time, only the cephalic portion of the heart is invested completely by primordial myocardium. When dealing with amphibian development, some embry- ologists retain the mistaken impression that a “tubular” heart of four regions is formed all at once. It can be seen that this view is incorrect when Nile blue-stained areas of the early heart are identified in later de- velopment (Copenhaver, ’39a). Although the initiation of cardiac function 448 in the different classes of vertebrates appears to be timed appropriately to the needs of the embryo, it must be emphasized that the initial contractions are not dependent upon conditions within the embryo or upon the development of heart form. This has been shown by explantation, transplantation and tissue culture methods. Furthermore, the initial contractions do not appear to be dependent upon the presence of striations. Goss (738, °40, °42) has observed cardiac contractions in rat embryos before striations can be seen in the myocardial cells. Simi- lar conclusions have been given for Am- blystoma (Copenhaver, ’39a). In the chick heart, contractions have been observed at the 9-somite stage (Patten and Kramer, °33), whereas striations have not been observed be- fore the 10 somite stage (Lewis, 719). In this connection, it should be noted that studies with the electron microscope (Schmitt, ’45) have shown that the absence of histological signs of striations need not imply the ab- sence of molecular or micellar striations. In the preceding discussion, reference has been made to the fact that different parts of the embryonic heart have inherently dif- ferent rhythms. As early as 1890, Fano and Badano cut chick embryo hearts with refer- ence to definite anatomical levels and re- corded higher pulsation rates for the atrium than for the ventricle. These early observa- tions have been amply confirmed and ex- tended by later workers. Similar findings have been made for embryonic mammalian hearts (Hall, 51). That the part with the highest rate dominates the other regions was demonstrated clearly by Paff (36). He grew reversed parts of the chick heart in proximity to each other in culture media and observed the sinoatrial region imposing its rate on the slower beating anterior part when the two pieces became united by a bridge of myocardial tissue. Further evi- dence for sinus dominance is seen when parts of the embryonic heart are trans- planted between different species of am- phibians—a _ transplanted sinus venosus dominates the remainder of the heart of another species to the point of maintaining the donor species rate (Copenhaver, °45). Figures 163 and 164 summarize quanti- tative studies on the intrinsic rhythms of each of the cardiac regions at successive developmental stages in Amblystoma punc- tatum (Copenhaver, ’39a) and in the chick (Barry, °42). The rhythm of each part passes through a phase of rate acceleration which is generally followed by a period of decelera- SPECIAL VERTEBRATE ORGANOGENESIS tion. Eventually, some of the parts enter a phase in which they lack the ability for spontaneous contraction. The conus of the chick heart enters the latter phase very early and its rate is omitted from the figure; the conus of the Amblystoma heart exhibits spontaneous contractions intermittently for a much longer time than in the chick. An- other difference between the two species is found in the behavior of the sinus venosus which enters the deceleration phase earlier in Amblystoma. In the chick, there is a question whether the rhythm of the intact heart (and the rate of its sinus) shows any deceleration phase. At most, there is only a slight decrease in rate after about 17 days of incubation (Cohn and Wile, ’25) or merely a plateau in the rate curve at that time (Bogue, °33). Numerous studies have shown that the embryonic heart rate exhibits a progressive acceleration which is most pronounced in the early stages of cardiac function. This appears to be true in fish, amphibians, birds and mammals. Among the studies on chick heart rates. those by Cohn and Wile (725), Bogue (732) and Barry (’40) are particu- larly significant. In both the chick (Fig. 164) and Amblystoma (Fig. 163), one notes a particularly obvious acceleration in heart rate just after the establishment of atrial function, and again just after the initiation of sinus function. The available evidence supports the statement by Barry (’42) that the progressive acceleration in heart rate in the early stages of cardiac function is “due in great measure to the successive addition of new segments of myocardium of increasingly higher inherent automaticity.” The progressive acceleration in the chick heart rate in later stages has been correlated with (1) an increase in blood pressure (Barry, °41), or (2) a reduction in cardiac distention following the deflection of blood into newly formed arteries (Alexander and Glaser, °41.) These contradictory views are based on contradictory results obtained with exsanguination experiments on chick em- bryos—bleeding decreases the heart rate according to Barry, and increases it ac- cording to Alexander and Glaser. It seems that the effects of hemorrhage on embryonic heart rate could be profitably reexamined. The most significant studies correlating stages of heart development and changes in the electrocardiogram have been made on chick hearts by Hoff, Kramer, DuBois and Patten (’39). In the earliest electrocardio- gram that can be obtained, there is a sinu- HEART, BLOOD VESSELS, BLOOD, AND ENTODERMAL DERIVATIVES soidal pattern produced by a simple de- flection of the galvanometer. A deflection resembling the QRS ventricular complex is not obtained until about 38 hours incubation. At this time, the contraction is still pri- marily ventricular and only the cephalic 449 ward at first; later it becomes reversed. An approximately adult type of electrocardio- gram is established during the fourth day of incubation. Nervous Control of the Embryonic Heart. Since the work of His and Romberg (1890), Fig. 163. BEATS PER MINUTE AT 19.5°C. a4; S55 celeemnss “—. ——.— ——— ——t — 39 “40 41 #42 #43 «44 «45 STAGES OF DEVELOPMENT w Oo BEATS PER MINUTE 30 40 50 60 70 80 90 100 110 120 AGE IN HOURS Fig. 164. Figs. 163 and 164. Inherent rhythms of different regions of transected hearts in Amblystoma (Fig. 163) and in the chick (Fig. 164). The stages shown for Amblystoma cover a 12-day period extending from the time of beat initiation until the embryos begin to feed. S, A, V, and C, show rates for sinus, atrium, ventricle and conus, respectively. Each point on the Amblystoma curves represents a mean for all cases studied, and the difference shown for the sinus and the intact, normal heart (N.H.R.) is statistically insignificant. (Fig. 163 after Copenhaver, 39; Fig. 164 after Barry, ’42.) portion of the atrium is functioning. Ac- cording to Olivo et al. (46), the establish- ment of a QRS pattern is correlated with the differentiation of myofibrillae and may occur at a somewhat earlier stage than that just cited. The P wave (atrial) appears first at 42 hours incubation, soon after the beat becomes established in the caudal part of the atrium. The P deflection is down- it has been known that the embryonic heart functions preceding innervation. This indicates that the first heart beats are myo- genic in origin. Furthermore, Hooker (’11) observed rhythmical cardiac contractions in denervated amphibian embryos reared be- yond the stage when the heart of normal embryos becomes innervated. Further evi- dence for a myogenic origin of heart beat 450 is found in tissue culture experiments (Bur- rows, 11) and in some of the embryonic heart experiments cited in the earlier part of this discussion. The latter include the de- velopment of pulsating hearts from pre- sumptive heart mesoderm after explantation to cultures (Ekman, ’21; Stohr, ’24a; Bacon, 45); after transplantation to heterotopic positions on the embryo (Ekman, ’21; Stohr, 29: Copenhaver, ’26); and after transplan- tation to the chorioallantoic membrane (Kumé, ’35; Rawles, 43). Neither extrinsic nor intrinsic cardiac nerves were present in the experiments cited above but it may be argued that the evidence does not apply to the adult heart. Among the experiments sup- porting a myogenic origin of heart beat in the adult, the work reviewed by Haber- landt (’27) has particular significance. He used freezing and chemical agents for sepa- rating the neural from the muscular com- ponents in adult hearts. The embryonic heart does not become in- nervated until some time after the sinus venosus has become established as the pace- maker. Vagus ingrowth (parasympathetic innervation) generally precedes sympathetic innervation, and morphological innervation precedes functional control by a variable period in different species. Vagus ingrowth occurs: (1) in A. punctatum at Harrison’s stages 44-46 (about 10 to 12 days after beat initiation (Copenhaver, ’39b); (2) in the chick at 120 hours incubation according to Abel (712); (3) in man, at the beginning of the fifth week (His and Romberg, 1890). In Fundulus heteroclitus embryos, Arm- strong (’35) found that cardiac innervation can be demonstrated by vital staining with methylene blue on the eighth day, about 5 days after the onset of heart beat. He also found that acetylcholine in small amounts induces auricular diastolic arrest on the eighth day whereas the same drug in large amounts does not inhibit contractility pre- ceding innervation. Responses cannot be elicited by reflex vagus stimulation on the eighth day; apparently the reflex arc is not complete until about 36 hours after the vagus ingrowth. Brinley (’35) obtained adrenaline effects on F. heteroclitus embryos, indicating the presence of a sympathetic innervation in teleosts; earlier workers failed to find a cardiac sympathetic imnervation in this class of vertebrates. Functional changes correlated with vagus ingrowth have been studied most completely in Fundulus embryos (Armstrong, *31). Sev- SPECIAL VERTEBRATE ORGANOGENESIS eral physiological stages can be identified covering the period from the initial reflex vagus response on the ninth day until an adult type of response is attained on about day 12-13. The type of response elicited by vagus stimulation at each successive stage appears to be correlated with the progressive innervation of different parts of the em- bryonic heart. The amount of cardiac control normally exerted by the vagi (vagal tone) varies for different species (see review by Clark, ’27). In most amphibians, there is very little vagal tone. In this connection, it is interesting to note that embryonic hearts transplanted heteroplastically between A. tigrinum and A. punctatum maintain their donor species rhythms although they become morphologi- cally and functionally innervated by the host nerves (Copenhaver, ’30, ’39b). In these cases, the basic heart rate characteristic for the species is apparently myogenic. EMBRYOGENESIS OF BLOOD VESSELS The development of blood vascular endo- thelium falls into two stages: (1) differ- entiation in situ from mesenchymal cells; (2) formation by vascular sprouts from pre- viously formed vessels. An overlapping in the time of the two stages was noted by Sabin (717) when she observed that sprouting begins before in situ formation is everywhere complete. Once the embry- onic vascular system is fully established, the endothelium of new _ vessels arises only as an outgrowth from pre-existing vessels. Clark and Clark (39) have sum- marized and extended earlier studies on this point. The mesenchymal cells which form endo- thelium are usually designated as angio- blasts or vasoformative cells. Further studies are needed to show when the earliest vaso- formative cells are determined and when their formation from indifferent mesenchyme ceases. Lewis (731) suggested that endo- thelium and mesenchyme may be differenti- ated in the chick at the time when they leave the primitive streak. On the other hand, Sabin (720) described the formation of new angioblasts from mesoderm throughout the first two days of incubation. Since Sabin’s conclusions were based on cyto- logical differences between angioblasts and mesoderm, they do not eliminate the proba- bility that some of the mesodermal cells are determined as angioblasts before they can HEART, BLOOD VESSELS, BLOOD, AND ENTODERMAL DERIVATIVES 451 be identified as such by available technical methods. Since the vasoformative cells arise in diverse embryonic regions, it is difficult to devise experiments to test the time of their determination. However, it is not im- probable that some mesodermal cells are de- termined for endothelium by the end of gastrulation (at a time when other meso- dermal cells are determined for other pri- mordia such as heart, limbs, et cetera). If this view of an early determination is cor- rect, it becomes easier to harmonize angio- genesis in the lower vertebrates with that in mammals. In macaque and human chor- ions, endothelium and mesoderm differen- tiate simultaneously from cytotrophoblast, according to Hertig (35). Although the early differentiation of vascular endothelium in man and other mammals is well estab- lished, Hertig’s interpretation that endo- thelium arises directly from cytotrophoblast is not universally accepted (Bloom and Bartelmez, ’40). There is no experimental evidence for a view that the early intraembryonic endo- thelium arises by ingrowth from the extra- embryonic endothelium of the yolk region. Explants and heterotopic transplants of cardiac and other organ primordia develop endothelium in situ. Neither is there any evidence for the unusual view that endo- thelium of the branchial and head vessels arises by outgrowth from the heart. The chief embryonic vessels develop when heart formation is prevented by extirpation of pre- sumptive heart mesoderm in amphibians. Experimental evidence for in situ formation in mammals is seen in the development of blood vessels from mesenchyme in explants of the allantoic bud from rat embryos (Jolly, "40). The earliest vessels in union with the heart—the truncus arteriosus and the ce- phalic portions of the vitelline veins—arise in a manner resembling the formation of endocardium. Cells of mesodermal origin aggregate in the pathways of the future ves- sels and differentiate into primitive endo- thelium. As the vessels are traced progres- sively from the heart, the strands of vasoformative cells become progressively more irregular in their pattern and form elaborate capillary networks. The arteries and veins arise by enlargement and differentiation of pathways through the net- work. This mode of development is found even in parts of the aortae and cardinal veins (Evans, ’08; Sabin, ’17). An exception to the formation of vessels from a primitive net- work is found in the opossum brain (Wis- locki, °39). In this species, the cerebral arteries are non-anastomotic ‘‘end-vessels,” and from the time of their first appearance these arteries and their corresponding veins show their characteristic adult plan. The pattern of the vascular system is de- pendent upon several factors. Hereditary fac- tors supposedly play an important part in the formation of the earliest vessels (aorta and large veins) which develop before cir- culation begins. Many vessels will undergo a fairly extensive development when circu- lation is prevented—for example, after cardiac extirpation (Knower, 07; Clark, ’18; and others). It is obvious that the mechani- cal effects of circulation cannot apply to vascular development under these condi- tions. On the other hand, the inherent pat- tern of the earliest vessels may be dependent upon mechanical and chemical effects from other embryonic tissues. Chemical and mechanical factors associ- ated with blood flow (function) affect the further development of the vascular system after the embryonic circulation becomes established. Clark (718) suggested that the formation of new vascular sprouts is influ- enced by the amount of interchange through the walls of the vessels—interchange be- tween blood and surrounding tissues. Thus the growth and metabolism of outside tis- sues may control the outgrowth of new vessels. Streeter (’18, ’27) also stressed the importance of the endothelial environment and stated that the embryonic vessels do not have a ground plan of their own. In this connection, some _ observations by Scharrer (°39) are of particular interest. He found that the end-vessel pattern character- istic of the opossum brain will develop in dead brain tissue and he interpreted this finding as evidence for a factor inherent to the cerebral vessels themselves. On the other hand, Wislocki (739) believes the result can be explained by an environmental factor present even in the dead tissue. The amount of blood flow through a capil- lary, rather than the rate of flow, determines whether a given capillary within a mesh- work atrophies, remains a capillary, or en- larges to form an arteriole or venule (see Clark’s ’18 analysis of Thoma’s laws). Clark and Clark (40) find that the differentiation of adventitial cells into smooth muscle is influenced by blood pressure and thus they confirm the histomechanical principle of 452 Thoma that the thickness of the vessel wall is dependent upon blood pressure. BLOOD Experimental studies on amphibians have shown that the blood islands form the sole source of the primitive erythrocytes (see Fig. 165). Following extirpation of this re- gion, as first performed on anurans by Federici (26) and on urodeles by Goss (28), “the blood cells were always reduced in number and in some instances were en- tirery absent.” The maximum survival re- ported by Goss for an embryo lacking blood cells was 32 days. An unanswered question is whether definitive erythrocytes would develop in an embryo lacking the primitive blood cells provided the animal could be maintained until the formation of other hematopoietic regions—liver, spleen, etc. The progenitors of the red corpuscles are determined considerably in advance of their morphological differentiation in the blood island (considerably before they can _ be identified by the benzidine technique for hemoglobin as applied by Slonimski, °31). This is evident from the fact that the pre- sumptive blood cells of amphibians can self- differentiate following explantation from neurulae (Slonimski, 31) and from gastru- lae (Fernald, 47). Furthermore, presumptive erythrocytes of an amphibian gastrula are localized to a relatively small zone opposite the blastoporal groove, as shown in Figure 158 (Holtfreter, *38; Fernald, *47). The presumptive blood cells of the chick, as in amphibians, are determined in advance of their differentiation but they are dispersed over a relatively large region of the early blastoderm. Erythrocytes will differentiate in chorioallantoic grafts taken from all parts of the blastoderm posterior to the level of the anterior quarter of the primitive streak (Murray, ’32). Goss (28) has noted that blood cells and endothelial cells differ in their origin be- cause extirpation of the blood island has relatively little effect on the development of the vascular system whereas it eliminates blood formation. Cameron (’41) has observed that the red cell progenitors differ from the endothelial precursors in their susceptibility to x-rays in Amblystoma. Considerable disagreement exists concern- ing the capacity of endothelium for erythro- poiesis. It is held generally that the endo- thelium in the blood island region of the SPECIAL VERTEBRATE ORGANOGENESIS yolk sac retains erythrogenic potency for a short time after the development of the earliest blood vessels. An exceptional view attributes hemogenic potency to all endo- thelium in amphibians until the stage of metamorphosis (Storti, *31). Experimental evidence has been presented to show that the lining of intersinusoidal “capillaries” of avian and mammalian bone marrow is eryth- rogenic even in the adult (Sabin, ’22; and others) but the lining of these channels is probably not a true endothelium (see sum- Fig. 165. A, Ventral view of an axolotl, stage 31, showing blood island (dark) stained by benzidine. B, Amblystoma punctatum, stage 32, showing blood island (stippled) and area excised in experimental production of bloodless embryos. (A, after Slonim- ski, 31; B, after Goss, ’28). marization by McDonald, ’39). On the basis of their defect experiments on amphibians, Federici (’26), Goss (’28), Slonimski (’31) and Fernald (’47) do not attribute any eryth- rogenic potency to endothelium. Their ex- tirpations of blood-forming cells undoubtedly included endothelium-forming cells of the yolk region but approximately normal ves- sels developed within this region, presum- ably from cells migrating from the edge of the wound. Neither the endothelium of the regenerated vessels of the operated field nor that of any other region produced eryth- rocytes. Two views exist concerning the nature of the first free blood cells which develop in the blood islands: (1) the primitive cell is a hemocytoblast with potency to form primi- tive erythroblasts and leukocytes; or (2) the earliest free cell is a primitive erythro- blast (megaloblast). The controversy results HEART, BLOOD VESSELS, BLOOD, AND ENTODERMAL DERIVATIVES 453 chiefly from the difficulty of distinguishing morphologically between hemocytoblasts and primitive erythroblasts. The dry imprint technique indicates that the first circulating cells are megaloblasts (Kirschbaum and Downey, ’37). An experimental analysis by Block (’46) indicates that the earliest cells are hemocytoblasts. He finds that when the rat yolk sac is transplanted into the anterior chamber of the eye, the primitive cells pro- duce granulocytes as well as erythroblasts. This is interpreted as evidence that the primi- tive cell itself is not an erythroblast. Re- gardless of whether the primitive cell is an erythroblast or a hemocytoblast which later develops into an erythroblast, there is general agreement that the primitive eryth- roblasts and erythrocytes formed in the yolk sac represent a megaloblastic strain differing from the definitive erythrocytes (normoblas- tic series) characteristic of older embryos and adults. Environment apparently plays a part in determining whether a primitive stem cell forms an erythrocyte or a leukocyte. Evi- dence for this includes Dantschakoff’s (°24) experiments showing that when the chick yolk sac is transplanted to the chorioallan- toic membrane, the stem cells develop mainly into granular leukocytes whereas in their normal location they develop chiefly into primitive erythroblasts. Jordan and Speidel (23) suggested that the fundamental stimulus for erythropoiesis is “some product of cellular metabolism, probably carbon dioxide.” There is considerable evidence that a de- crease in oxygen stimulates the rate of re- generation of erythrocytes in adults. Studies by Grant and Root (747) on bone marrow blood following hemorrhage in dogs indicate that oxygen content and capacity, rather than oxygen tension and saturation, affect erythropoiesis. Their studies are particu- larly interesting in relation to the commonly held view that erythropoiesis in marrow is related to a sluggish blood flow and a lo- calized anoxia. They found that stagnant anoxia, following hemorrhage, does not per- sist any longer in marrow blood than in jugular vein blood. A discussion of the factors necessary for the maturation of the erythrocyte is beyond the scope of the present review. It is particu- larly interesting to note that vitamin B,», obtained from liver extract, appears to have the same effects as liver extract in pernicious anemia therapy (West, ’48). For a review of the nutritive factors and mechanisms in the regeneration of erythrocytes, reference may be made to Wintrobe (750). ENTODERMAL DERIVATIVES The localization of presumptive organ rudiments within the entoderm has been studied by the vital dye marking method (Vogt, ’29; Balinsky, ’47a) and by numerous experiments employing the methods of extir- pation and transplantation. In amphibians, the different regions of the entoderm appear to be capable of self-differentiation in the gastrula stage as evidenced by the behavior of transplants in the coelomic cavity and by explants in modified Ringer’s solution (Holt- freter, °25, ’38). On the other hand, evidence for alteration of the prospective fate of amphibian entoderm is seen in the occasional development of muscle from entodermal cells transplanted from gastrulae into the eye chamber of older larvae (Kusche, ’29). His- tological regulation in amphibian entoderm has been described for neurula stages also (Balinsky, ’38). However, studies on the tree frog by Kemp (46) are more in conformity with Holtfreter’s views and indicate that the entoderm in the neurula is unable to regulate histologically although it is capable of considerable morphological regulation. Extending the studies on amphibian ento- derm to earlier stages than those used by Holtfreter, Nicholas (’48) found that the entoderm already possesses some determina- tion in the late blastula stage. When the entire cover of an early gastrula is removed, the entoderm differentiates stomodeal and proctodeal pits. The factors for the formation of these two structures are apparently in- trinsic to the yolk entoderm at an early stage. An understanding of the formation of ento- dermal structures in the chick has been aided particularly by the method of chorio- allantoic grafting. Early blastoderms can be split into two layers: epiblast and ento- derm. Grafts have been made of entire layers, of parts of layers, and of small areas of total thickness blastoderms. The experiments indi- cate that: (1) the gut-forming potency is limited to the epiblast in prestreak and early streak blastoderms (Hunt, ’37a); (2) the definitive entoderm does not become deter- mined until just prior to the head-process stage (Hunt, °37a; Rudnick and Rawles, °37); (3) for a time there is an overlapping of gut potentialities in the epiblast and ento- dermal layers; (4) there is a progressive localization of primordia within the ento- 454 derm, occurring at different times for dif- ferent organs (Rudnick, °35). By means of vital dyes, a cellular migration from ecto- derm to entoderm has been demonstrated in correlation with the shift of gut potency from ectoderm to entoderm in the early blastoderm (Hunt, °37b), but there is “no evidence that the localization process itself is to be explained on the basis of a cellular migration” (Rudnick and Rawles, °37). SPECIAL VERTEBRATE ORGANOGENESIS Mouth Formation. Experiments on the stomodeal region of A. punctatum by Adams (24, °31) showed that mouth invagination is dependent upon contact of the ectoderm with the underlying entoderm. Experiments by Balinsky (47b) confirmed and extended the view that the entoderm acts as an in- ductor on the stomodeal ectoderm. Most of Balinsky’s experiments were done on T. taeniatus embryos in late gastrula and early Median Eye A ye Double Forelimb Fig. 166. A and B, sketch of operation for removal of whole entoderm at neurula stage. C, entoderm-free larva of Triturus alpestris 20 days after operation. (Redrawn after Nieuwkoop, ’46.) The prospective potencies of the different digestive tract primordia are greater than their individual prospective fates. This is shown by an overlapping of ectodermal and entodermal gut-potency stages (cited above) and by the extent of gut-potency tissue on the early blastoderm (Butler, °35; Rudnick and Rawles, °37). Complete extirpation of the entoderm from neurula stage amphibian embryos in- dicates that the entoderm has striking effects on the development of other tissues (Nieuw- koop, 46). Entoderm-free embryos frequently develop two pairs of forelimbs (Fig. 166), they show various degrees of microcephaly and cyclopia, they lack gills and, as described in a previous section, there is a failure of heart development. neurula stages. He found that ectodermal mouth invagination usually failed after com- plete removal of mouth entoderm, and that presumptive mouth ectoderm transplanted to abnormal locations produced a mouth in- vagination only when underlying mouth entoderm was transplanted along with the presumptive ectoderm. A regional factor in the stomodeal ectoderm was indicated by the following experiments: (1) mouth ento- derm transplanted to foreign positions did not induce mouth invagination in the over- lying foreign ectoderm; (2) mouth invagina- tion failed when mouth ectoderm was com- pletely replaced by foreign ectoderm. In the latter result, the experiments on Triton ap- pear to differ from those on Amblystoma by Adams (’24), where mouth parts were found HEART, BLOOD VESSELS, BLOOD, AND ENTODERMAL DERIVATIVES 455 to develop from foreign ectoderm trans- planted into the stomodeal region. Balinsky concludes that there is a regional factor in the presumptive mouth ectoderm but that determination occurs gradually. At the neurula stage, the presumptive mouth ecto- derm has only a labile determination and it will not differentiate without an inductive action from the underlying entoderm. Tooth development in some amphibians differs from that in mammals in regard to the origin of the enamel organs. Experiments by Adams (724, ’31) showed that in A. punctatum the enamel organs of some tooth germs arise from ectoderm, others from entoderm. On the other hand, Woerdeman failed to find proof for the formation of teeth from entoderm in xenoplastic grafts between TJ. taeniatus and the axolotl (Woerdeman and Raven, *46). Studies on Amblystoma by Adams (24, 31) supported the view that the pulp of the tooth germ is mesectodermal (neural crest) in origin. Her experiments also raised the question whether the entoderm or mesecto- derm initiates tooth formation. Among those who have studied this question further, Woerdeman and Raven (746) concluded that there is a reciprocal inductive influence—that the papilla “induces the epithelium to form a tooth-germ, whilst the size of the papilla de- pends on the enamel-organ.” Horstadius (750) demonstrated the reciprocal relationships more specifically by experiments in which tissues from the head were grafted in differ- ent combinations into the trunk. He found a failure of tooth development in the following grafts: oral entoderm plus oral ectoderm, oral entoderm plus neural crest, and oral ectoderm plus neural crest. Teeth developed in grafts of oral entoderm plus neural crest plus oral ectoderm. Studies on the man- dibular arch by Levy and Detwiler (751, and personal communication) are essentially in agreement with the views of Horstadius. Likewise, in vitro studies by Wilde (751) show that it is necessary to explant cranial neuroepithelium along with stomodeal ecto- derm and foregut entoderm for teeth to develop. Role of Entoderm in Gill Formation. Har- rison (’21) studied the respective roles of branchial ectoderm and mesoderm in gill formation by extirpation and transplantation experiments. He established the following facts: (1) the branchial ectoderm of A. punctatum is specifically gill forming but an ectodermal region surrounding the gills also has the potency to form gills in a diminishing intensity as the distance from the gills increases; (2) the specific gill pat- tern is not laid down in the ectoderm; (3) the specific pattern must be determined by the deeper layers. In the absence of experi- ments on the entoderm, the respective roles of mesoderm and entoderm were not clear but it was suggested that the entoderm did not play merely a passive role. Heteroplastic gill transplantations made by Harrison (729) showed that gills resembling those of the donor species occurred only when the whole gill complex was transplanted, entoderm included. The importance of the branchial entoderm was demonstrated in experiments on tail-bud stage Amblystoma embryos by Severinghaus (730) in which it was found that a gill fails to develop after removal of its entoderm. It is well established that neural crest cells contribute to the formation of the visceral arch skeleton (Landacre, ’21; Stone, ’22; and others). It has also been shown that the entoderm of the oral region acts as an in- ductor on the migrating neural crest cells for the formation of cartilage. Following ex- perimentally produced defects in the oral region, Balinsky (47b) found a correlation between the arrangement, number and size of the visceral cartilages and the presence of entoderm of the pharyngeal pouches. In transplantations of head tissues to the trunk, Horstadius (50) found that neural crest transplanted with oral entoderm will pro- duce cartilage whereas neural crest grafted with oral ectoderm alone will not. Mid-gut Derivatives—Stomach, Intestine, Liver and Pancreas. In amphibians, presump- tive entodermal organ rudiments appear to be determined relatively early. The hetero- topic transplantation experiments on Bombi- nator by Holtfreter (25) showed that the liver and pancreatic primordia are capable of self-differentiation in the late gastrula stage. He found that the liver primordium itself is not a mosaic—it is not regionally determined—and a gallbladder can develop in grafts from either cephalic or caudal por- tions of the liver rudiment of tail-bud em- bryos. Whether one presumptive organ rudiment (e.g., stomach) can regulate to form another (eg., liver) is controversial. According to one view (Holtfreter, ’25, °38; Kemp, °46, °51; and others) considerable regulation may occur in gross morphology but not in histological structure. According to another view (Kusche, ’29; Balinsky, °38, ’47b), histological regulation can occur. In xenoplastic grafts between Triturus and the 456 axolotl, Balinsky found that presumptive liver tissue grafted into the stomach region could differentiate into stomach tissue. Like- wise, presumptive stomach differentiated into liver cells and pancreatic acini when grafted into the regions of the latter tissues. These experiments suggest that the stomach, liver and pancreas have only a labile type of determination in neurula stage Triturus. In the pancreas, there is an interesting differ- ence between the potencies of the dorsal and ventral primordia in that only the former is capable of differentiating islands of Langer- hans (Wolf-Heidegger, °36). The influence of mesodermal derivatives on the form of the digestive organs has been studied particularly by Kemp (751). Em- bryos of R. pipiens at tail-bud stages were transected at different levels and reared in Holtfreter’s solution. In other experiments, dorsal and ventral embryonic regions were excised. The following conclusions were drawn: development of a normal pattern of intestinal coiling in anurans is dependent upon the establishment of the vitelline cir- culation, upon the regulation of hydrostatic pressure within the digestive tract and coelom, and upon the restricted space of the coelomic cavity. In the chick, chorioallantoic grafts indi- cate that liver potency tissue exists in both prospective dorsal and prospective ventral embryonic regions until sometime between the pre-somite and third somite stages; after this, the liver is segregated in so far as the potentialities for its development are elimi- nated from the dorsal region (Rudnick, 35). Segregation of a similar nature occurs at a somewhat later stage for the pancreas. Liver differentiation in the chick appears to depend upon an inducing action from the heart rudiment, since liver is rarely present in grafts lacking heart tissue whereas the converse occurs quite often (Willier and Rawles, °31). The inducing action occurs particularly in grafts containing prospective liver material; the heart inductor action seems unnecessary when the liver develops from grafted material of some other prospec- tive value, e.g., from prospective dorsal re- gions (Rudnick, ’35). In amphibians, liver differentiation is not dependent upon heart development since liver can differentiate in the absence of the heart in explants from the gastrula stage (Holtfreter, ’25). Embryos survive complete liver extirpa- tion for a longer period than do larvae or adults of the same species. When the opera- SPECIAL VERTEBRATE ORGANOGENESIS tion is done on embryos in the gill-formation stage, survival is about two weeks for anuran amphibians (Yamada, *33) and about four weeks for urodeles (Copenhaver, ’43). Devel- opmental defects following embryonic hepa- tectomy indicate that liver development aft- fects the formation of other embryonic struc- tures through (a) a morphological relation- ship and (6) a functional relationship. Ex- amples of the first type include modifications in the ventral mesentery and alterations in the course of venous drainage from the in- testine. These changes can be seen best when the growth of a hepatectomized embryo is maintained either by parabiosis with a nor- mal embryo (Yamada, *33) or by liver tissue implanted to the tail (Copenhaver, °43). Defects resulting from functional changes following liver extirpation include anemia and retarded growth of the spleen (Copen- haver, 43). The heart may be abnormally small, particularly in anemic animals. Other defects arising either directly or indirectly from functional changes following hepatec- tomy are distention of the pronephric canals (Holtfreter, °25), hypertrophy of the pro- nephros (Yamada, °33), and edema (Holt- freter, ’25; Yamada, 33). Numerous studies have established the fact that there is a rapid restoration of liver tissue following partial hepatectomy in mammals. Higgins and Anderson (731) have shown that the liver of rats can regain its normal size within two to three weeks after a 70 per cent hepatectomy. According to Newman and Grossman (751), nucleic acid supplements in the diet accelerate the rate of regeneration. Higgins and Anderson noted that restoration of mammalian liver tissue differs from regeneration of the tadpole’s tail; in the former, there is no blastema of re- generating tissue at the level of the cut. When a lobe of the liver is removed, there is a proliferation of cells and formation of new lobules throughout the remaining lobes but the extirpated lobe itself is not restored. Considering the fact that most organs have a much greater regenerative capacity in amphibians than in mammals, it is surprising to find that an opposite result has been re- ported for the liver. Experiments on Bombi- nator and on R. esculenta by Holtfreter (’25) have confirmed the lack of liver regeneration reported by Banchi (06) for Bufo. Studies on the salamander, Triturus viridescens, by Jordan and Beams (’30), showed compensa- tory hypertrophy but no proliferative activity after semihepatectomy. After excision of HEART, BLOOD VESSELS, BLOOD, AND ENTODERMAL DERIVATIVES 457 more than one-half of the liver, the lobular remnants were apparently incapable of ef- fective restoration. Onset of Liver Function. Functional dif- ferentiation of liver tissue has been studied more completely for the chick than for other species. Bile secretion appears on about the sixth day of incubation (Sandstrom, 7°34). On about the seventh day, appreciable amounts of glycogen are found in the liver although the pancreatic islands do not ap- pear before the eleventh day (Dalton, ’37b). From the developmental sequence just cited and from the fact that glycogen is stored by embryonic liver cells cultivated in vitro (Nordmann, ’29), it is obvious that the syn- thesis and storage of glycogen in the embry- onic liver is not dependent upon insulin. Fowever, one notes a more adult type of e'ycogen distribution in the liver after the islands of Langerhans appear. Data on the excretion of uric acid indicate that the liver begins to function in protein metabolism on about the seventh day (Fiske and Boyden, 26). Cholesterol is normally not evident in liver tissue until the eleventh day of incubation, but an earlier capacity for lipid metabolism is indicated by the fact that cholesterol appears on the seventh day in embryonic hepatic cells grafted to the chorio- allantois of older embryos (Dalton, 37a). Thus the available evidence indicates that most of the multiple functions of the liver cell appear at about the same time. Although the hepatic cell is capable of function at an early age, it does not assume its full degree of function until relatively late. In the chick, glycogen storage in the extraembryonic tissues exceeds that in the liver until the seventeenth day of incubation, Le., until 82 per cent of the incubation period has elapsed (Needham, ’31). In mammals, the glycogenic function occurs chiefly in the placenta until relatively late in development. For example, the glycogen content of the rat liver first exceeds that of the placenta after the elapse of 75 per cent of the total gestation period (Corey, °35). Asymmetry and Situs Inversus. The present discussion will be confined to asymmetry of the heart and viscera; for an analysis of other aspects of symmetry in the developing embryo, reference may be made to Harrison (45). Questions related to the development of normal visceral asymmetry and to situs in- versus have been studied extensively by the experimental method, particularly in am- phibians. Spemann, (’06) showed that a 180 degree rotation of the middle part of the medullary plate and its underlying meso- derm and entoderm frequently produced situs inversus of the gut and heart. Numer- ous investigators have confirmed and ex- tended this finding. Experiments by von Woellwarth (50) are particularly significant. In a control series of animals with a part of the medullary plate and its underlying tissue reimplanted with normal orientation, he found visceral situs inversus in about the same frequency as in another series of animals with a similar block of tissue rotated through 180 degrees. Working on the hypoth- esis that the unexpected results in the non- rotated reimplantations were caused by tis- sue damage, von Woellwarth made unilateral defects in the different germ layers in gas- trula and neurula stage embryos. The follow- ing results were obtained from neurula stage operations: defects in the left side of the mesoderm gave weak inversion effects (9 per cent), defects of the right mesoderm gave no inversion effects, defects in ento- derm gave strong inversion effects (about equal on the two sides—22 and 21 per cent, respectively, and defects in mesoderm and entoderm gave a strong effect for the left side (41 per cent) and a medium effect for the right (20 per cent). In gastrula stage operations, defects in the presumptive meso- derm of the left side gave strong inversion effects (50 per cent), defects of right meso- derm gave weak effects (8 per cent) and defects of presumptive entoderm of both sides gave weak effects (6 per cent). The experiments of von Woellwarth show that situs inversus occurs with the greatest frequency after defects to the mesoderm of the left side at the gastrula stage. He believes that the results indicate an asymmetry of the organizer. In this connection, he notes that Goerttler (28) argued for a physio- logical asymmetry of the mesoderm. The latter reported that explants of presumptive heart mesoderm from the left side of medul- lary plate stage urodeles developed pulsa- tions within a few days whereas explants from the right side did not. On the other hand, Holtfreter (733) found that pulsations began at about the same time in right and left heart anlage explanted from neurula stage axolotls. Nevertheless, there is support for the idea of an asymmetry in the two sides of the heart anlage. Bacon (745) found that both sides of the bilateral heart rudi- ment of medullary plate stage Amblystoma 458 are capable of self-differentiation but ex- plants from the left side give a higher per- centage of positive cases and an earlier initia- tion of pulsations. Rawles (43) obtained essentially similar results for chorioallantoic grafts of heart-forming areas from the early chick blastoderm. Situs inversus of the heart and gut fre- quently occur together although either one may appear without the other. This is true both for the cases seen in nature and for those produced experimentally. The most commonly described situs inversus cordis in the absence of situs inversus viscerum is that which occurs in the right member of double hearts produced by prevention of union of the bilateral heart primordia. In these cases, left hearts have normal asymmetry; right hearts frequently show situs inversus. The - asymmetry of the right member of double hearts is probably correlated with an en- vironmental influence already discussed by Ekman (’24, ’25)—the union of the heart with the blood vessels and the pressure ef- fects within the pericardial cavity favor normal curvature on the left side and situs inversus on the right side when the bilateral heart primordia develop separately. It is not surprising that the right member of double hearts does not develop situs inversus in- variably, that both hearts occasionally show the same asymmetry (Fales, *46). Without the environmental effects just cited, normal asymmetry of the right member of double hearts might occur more often than it does. Theoretically, one might expect right heart primordia freed from the influence of the left side to develop with normal asymmetry and with situs inversus in about eaual num- bers, similar to the results obtained by Ruud and Spemann (723) for right half blastulae. They found that when blastulae are com- pletely constricted, left halves develop into small animals with normal asymmetry while right halves develop with situs inversus and with normal asymmetry in about equal numbers. For further details on the relation of situs inversus to localized defects in the germ layers and the probable relationship between normal visceral asymmetry and asymmetry in the embryonic “organizer,” reference should be made to von Woellwarth (oD). REFERENCES Abel, W. 1912 Further observations on the de- velopment of the sympathetic nervous system in the chick. J. Anat. and Physiol., 47:35-72. Adams, A. E. 1924 An experimental study of the SPECIAL VERTEBRATE ORGANOGENESIS development of the mouth in the amphibian em- bryo. J. Exp. Zool., 40:311-379. 1931 Some effects of removal of endoderm from the mouth region of early Amblystoma punctatum embryos. J. Exp. Zool., 58:147-163. Alexander, R. S. and Glaser,O. 1941 Progressive acceleration in embryonic hearts. J. Exp. Zool., 87:17-30. Armstrong, P. B. 1931 Functional changes in the embryonic heart accompanying the ingrowth and development of the vagus innervation. J. Exp. Zool., 58:43-67. 1935 The role of the nerves in the action of acetylcholine on the embryonic heart. J. Physiol., 84:20-32. Bacon, R. L. 1945 Self-differentiation and induc- tion in the heart of Amblystoma. J. Exp. Zool., 98:87-125. Balinsky, B. I. 1938 On the determination of entodermal organs in amphibia. Compt. Rend. Acad. Sci. U.RB.S.S., 20:215-217. 1939 Experiments on total extirpation of the whole entoderm in Triton embryos. Compt. Rend. Acad. Sci. U.R.S.S., 23:196-198. 1947a Kinematik des entodermalen Ma- terials bei der Gestaltung der wichtigsten Teile des Darmkanals bei den Amphibien. Roux’ Arch. Entw.-mech., 743:126-166. 1947b Korrelationen in der Entwicklung der Mund- und Kiemenregion und des Darmkan- als bei Amphibien. Roux’ Arch. Entw.-mech., 143:365-395. Banchi, A. 1906 Sulla rigenerazione degli abozzi del fegato e del pancreas. Arch. ital. di Anat. e di Embriolo., 5:507-532. Barry, A. 1940 Age changes in the pulsation fre- quency of the embryonic chick heart. J. Exp. Zool., 85:157-170. 1941 The effect of exsanguination on the heart rate of the embryonic chick. J. Exp. Zool., 1942 The intrinsic pulsation rates of frag- ments of the embryonic chick heart. J. Exp. Zool., 91:119-130. Block, M. 1946 An experimental analysis of hematopoiesis in the rat yolk sac. Anat. Rec., 96: 289-312. Bloom, W., and Bartelmez, G. 1940 Hemato- poiesis in young human embryos. Am. J. Anat., 67:21-44, Bogue, J. Y. 1932 The heart rate of the develop- ing chick. J. Exp. Biol., 9:351-358. Born, G. 1897 Uber Verwachsungsversuche mit Amphibienlarven. Roux’ Arch. Entw.-mech., 4: 349-465. Bremer, J. L. 1931 The presence and influence of two spiral streams in the heart of the chick em- bryo. Am. J. Anat., 49:409-440. Brinley, F. J. 1935 Evidence of a sympathetic innervation of the teleost heart, with a note on a method of transplanting the heart of Fundulus embryos. Physiol. Zool., 8:360-373. Burrows, M. T. 1911 The growth of tissues of the chick embryo outside the animal body, with special reference to the nervous system. J, Exp. Zool., 10:63-84. . HEART, BLOOD VESSELS, BLOOD, AND ENTODERMAL DERIVATIVES 459 Butler, E. 1935 The developmental capacity of regions of the unincubated chick blastoderm as tested in chorio-allantoic grafts. J. Exp. Zool., 70: 357-395. Cameron, J. A. 1941 Primitive blood-cell gen- erations in Amblystoma. J. Morph., 68:231—237. Clark, A. J. 1927 Comparative Physiology of the Heart. Cambridge University Press, Cambridge, England. Clark, E.R. 1918 Studies on the growth of blood- vessels in the tail of the frog larva—by observa- tion and experiment on the living animal. Am. J. Anat., 23:37-88. , and Clark, E. L. 1939 Microscopic ob- servations on the growth of blood capillaries in the living mammal. Am. J. Anat., 64:251-301. , and Clark, E. L. 1940 Microscopic ob- servations on the extra-endothelial cells of living mammalian blood vessels. Am. J. Anat., 66:1-49. Cohn, A. E., and Wile, E. L. 1925 Physiological ontogeny. A. Chicken embryos. V. On the rate of the heart beat during the development of chicken embryos. J. Exp. Med., 42:291-297. Copenhaver, W. M. 1926 Experiments on the development of the heart of Amblystoma puncta- tum. J. Exp. Zool., 43:321-371. 1930 Results of heteroplastic transplanta- tion of anterior and posterior parts of the heart rudiment in Amblystoma embryos. J. Exp. Zool., 55:293-318. 1939a Initiation of beat and intrinsic contraction rates in the different parts of the Amblystoma heart. J. Exp. Zool., 80:193-224. 1939b Some observations on the growth and function of heteroplastic heart grafts. J. Exp. Zool., 82:239-271. 1943 Liver extirpation and implantation in Amblystoma embryos with particular refer- ence to blood formation. Am. J. Anat., 73:81-105. 1945 Heteroplastic transplantation of the sinus venosus between two species of Amblys- toma. J. Exp. Zool., 700:203-216. Corey, E. L. 1935 Growth and glycogen content of the fetal liver and placenta. Am. J. Physiol., 112:263-267. Dalton, A. J. 1937a Cholesterol storage and bile secretion in chorio-allantoic grafts of liver. Anat. Rec., 67:431-439. 1937b The functional differentiation of the hepatic cells of the chick embryo. Anat. Rec., 68:393-409. Dantschakoff, V. 1924 Wachstum transplantier- ter embryonaler Gewebe in der Allantois. Zeit. Anat. Entwk., 74:401-431. Davis, C. L. 1927 Development of the human heart from its first appearance to the stage found in embryos of twenty paired somites. Carnegie Contrib. to Embryol., 19:245-284. Dwinnell, L. A. 1939 Physiological contraction of double hearts in rabbit embryos. Proc. Soc. Exp. Biol. & Med., 42:264-267. Ekman, G. 1921 Experimentelle Beitrage zur Entwicklung des Bombinatorherzens. Oversikt av. Finska Vetenskapssocietetens Forhandlingar, 63: 1-37. 1924 Neue experimentelle Beitrage zur friihesten Entwicklung des Amphibienherzens. Soc. Scient. Fenn. Comm. Biol., I.9:1-37. 1925 Experimentelle Beitrage zur Herz- entwicklung der Amphibien. Roux’ Arch. Entw.- mech., 106:320-352. 1927 Ejinige experimentelle Beitrage zur friihesten Herzentwicklung bei Rana fusca. Ann. Acad. Scient. Fenn., Ser. A. 27:1-26. 1929 Experimentelle Untersuchungen uber die friiheste Herzentwicklung bei Rana fusca. Roux’ Arch. Entw.-mech., 116:327-347. Evans, H. M. 1908 On the development of the aortae, cardinal veins and umbilical veins, and the other blood vessels of vertebrate embryos from capillaries. Anat. Rec., 3:498-518. Fales, D. E. 1946 A study of double hearts pro- duced experimentally in embryos of Amblystoma punctatum. J. Exp. Zool., 101:281-298. Fano, G., and Badano, F. 1890 Etude physiolo- gique des premiers stades de développment du coeur embryonnaire du poulet. Arch. Ital. Biol., 73:387-422. Federici, H. 1926 Recherches expérimentales sur les potentialitiés de Vilot sanguin chez l’embryon de Rana fusca. Arch. de Biol., 36:466—487. Fernald, R. L. 1947 The origin and development of the blood island of Hyla regilla. Univ. Calif. Publ. Zool., 57:129-147. Fiske, C. H., and Boyden, E. A. 1926 Nitrogen metabolism of the chick embryo. J. Biol. Chem., 70:535-556. Goerttler, K. 1928 Die Bedeutung der ventro- lateralen Mesodermbezirke fiir die Herzanlage der Amphibienkeime. Anat. Anzeig. Erganzungs- heft, 66:132-139. Goss, C. M. 1928 Experimental removal of the blood island of Amblystoma punctatum embryos. J. Exp. Zool., 52:45-61. 1935 Double hearts produced experi- mentally in rat embryos. J. Exp. Zool., 72:33-49. 1938 The first contractions of the heart in rat embryos. Anat. Rec., 70:505-524. 1940 First contractions of the heart with- out cytological differentiation. Anat. Rec., 76: 19-27. 1942 The physiology of the embryonic mammalian heart before circulation. Am. J. Phys- iol., 137:146-152. Grant, W. C., and Root, W.S. 1947 The relation of Oz in bone marrow blood to post-hemorrhagic erythropoiesis. Am. J. Physiol., 750:618-627. Graper, L. 1907 Untersuchungen iiber die Herz- bildung der Végel. Roux’ Arch. Entw.-mech., 24: 375-410. Haberlandt, L. 1927 Das Hormon der Herzbe- wegung. Urban and Schwarzenberg, Berlin. Hall, E. K. 1951 Intrinsic contractility im the embryonic rat heart. Anat. Rec., 177:381- 400. Harrison, R. G. 1921 Experiments on the devel- opment of the gills in the amphibian embryo. Biol. Bull., 47:156-170. 1929 Heteroplastic transplantations in amphibian embryos. X¢ Congrés Int. d. Zoolo- gie, Budapest, Part I: 642-650. 1945 Relations of symmetry in the de- 460 veloping embryo. Trans. Conn. Acad. Arts & Sciences, 36:277—330. Hertig, A. T. 1935 Angiogenesis in the early human chorion and in the primary placenta of the macaque monkey. Carnegie Contrib. to Em- bryol., 25:37-81. Higgins, G. M., and Anderson, R. M. 1931 Ex- perimental pathology of the liver. Arch. Path., 12:186-202. His, W., Jr., and Romberg, E. 1890 Beitrage zur Herzinnervation. Fortschritte der Medizin, 8:374— 380. Horstadius, S. 1950 The Neural Crest. Oxford University Press, Oxford, England. Hoff, E. C., Kramer, T. C., DuBois, D., and Patten, B. M. 1939 The development of the electro- cardiogram of the embryonic heart. Am. Heart J., 17:470-488. Holtfreter, J. 1925 Defekt- und Transplantations- versuche an der Anlage von Leber und Pancreas jiimgster Amphibienkeime. Roux’ Arch. Entw.- mech., 105:330-383. 1933 Die totale Exogastrulation, eine Selbstablésung des Ektoderms vom Entomeso- derm. Roux’ Arch. Entw.-mech., 129:669-793. 1938 Differenzierungspotenzen isolierter Teile der Urodelengastrula. Roux’ Arch. Entw.- mech., 138:522-656. Hooker, D. 1911 The development and function of voluntary and cardiac muscle in embryos with- out nerves. J. Exp. Zool., 77:159-186. Hunt, T. E. 1932 Potencies of transverse levels of the chick blastoderm in the definitive-streak stage. Anat. Rec., 55:41-69. 1937a The development of gut and its derivatives from the mesectoderm and mesento- derm of early chick blastoderms. Anat. Rec., 68: 349-369. 1937b The origin of entodermal cells from the primitive streak of the chick embryo. Anat. Rec., 68:449-460. Johnstone, P. N. 1925 Studies on the physiolog- ical anatomy of the embryonic heart. II. An in- quiry into the development of the heart beat in chick embryos including the development of ir- ritability to electrical stimulation. Bull. Johns Hopkins Hosp., 36:299-311. Jolly, J. 1940 Recherches sur la formation du systeme vasculaire de l’embryo. Arch. d’anat. microscop., 35:295-361. Jordan, H. E., and Beams, H. W. 1930 Hepatec- tomy in the salamander with special reference to hemopoiesis and cytology of the liver remnant. Proc. Soc. Exp. Biol. & Med., 28:181-184. ,and Speidel,C.C. 1923 The fundamental erythrocytopoietic stimulus. Proc. Soc. Exp. Biol. & Med., 21:399-404. Kemp, N. E. 1946 Regulation in the entoderm of the tree frog Hyla regilla. Univ. Calif. Publ. Zool., 51:159-182. 1951 Development of intestinal coiling in anuran larvae. J. Exp. Zool., 116:259-287. Kirschbaum, A., and Downey, H. 1937 A com- parison of some of the methods used in the studies of hematopoietic tissues. Anat. Rec., 68:227-231. Knower, H. McE. 1907 Effects of early removal of the heart and arrest of the circulation on the SPECIAL VERTEBRATE ORGANOGENESIS eee of frog embryos. Anat. Rec., 7:161- 165. Kume, M. 1935 The differentiating capacity of various regions of the heart rudiment of the chick as studied in chorio-allantoic grafts. Physiol. Zool., 8:73—90. Kusche, W. 1929 Interplantation umschriebener Zellbezirke aus der Blastula und der Gastrula von Amphibien. Roux’ Arch. Entw.-mech., 720:192- 7k Landacre, F.L. 1921 The fate of the neural crest in the head of urodeles. J. Comp. Neur., 33:1-43. Levy, B. M., and Detwiler, S. R. 1951 Experi- mental studies on the development of the mandib- ular arch in Amblystoma punctatum. J. Dental Research, 30:1-12. Lewis, M. R. 1919 The development of cross striation in the heart muscle of the chick embryo. Bull. Johns Hopkins Hosp., 30:176-181. Lewis, W. H. 1931 The outgrowth of endothe- lium and capillaries in tissue culture. Bull. Johns Hopkins Hosp., 48:242-253. Mangold, O. 1937 Isolationsversuche zur An- alyse der Entwicklung der Gehoér-, Kiemen- und Extremitatenregion bei Urodelen. Acta Soc. pro Fauna et Flora Fenn., 60:3-44. McDonald, J. G. 1939 Avian bone marrow with particular reference to red cell development. Am. J. Anat., 65:291-308. Murray, P.D. F. 1932 The development in vitro of the blood of the early chick embryo. Proc. Roy. Soc. Lond. B., 171:497-521. Needham, J. 1931 Chemical Embryology, Vol. II. Cambridge University Press, Cambridge, Eng- land. Newman, E. A., and Grossman, M. I. 1951 Ef- fect of nucleic acid supplements in the diet on rate of regeneration of liver in rats. Am. J. Physiol., 164:251-253. Nicholas, J. S. 1948 Form changes during pre- gastrular development. Ann. N.Y. Acad. Sci., 49:801-817. , and Rudnick, D. 1934 The develop- ment of rat embryos in tissue culture. Proc. Nat. Acad. Sci., 20:656-658. Nieuwkoop, P. D. 1946 Experimental investiga- tions on the origin and determination of the germ cells, and on the development of the lateral plates and germ ridges in urodeles. Arch. Néerland. Zool., 8:1-205. Nordmann, M. 1929 Wachstum und Stoffwech- sel der Leberzellen in der Gewebskultur. Arch. exp. Zellf., 8:371-414. Olivo, O. M. 1928 Uber die friihzeitige Deter- minierung der Herzanlage beim Hiihnerembryo und deren Differenzierung in vitro. Anat. Anz., Erg. Heft, 66:108-118. , Petralia, S., and Ricamo, R. 1946 Elet- trocardiogramma e miofibrille nelle colture in vitro di miocardio embrionale. Boll. Soc. Ital. Biol. _ Sperim., 22/7:911—913. Ota, T. 1930 Experimentelle Studien iber den Herzbildungsbezirk an den Amphibienlarven. Jap. J. Med. Sci., 2:235-242. Paff, G. H. 1936 Transplantation of the sino- atrium to the conus in the embryonic heart. Am. J. Physiol., 117:313-317. HEART, BLOOD VESSELS, BLOOD, AND ENTODERMAL DERIVATIVES 461 Patten, B. M. 1949 Initiation and early changes in the character of the heart beat in vertebrate embryos. Physiol. Rev., 29:31—47. =—— and Kramer, 0. G. “943.5 dine initiation of contraction in the embryonic chick heart. Am. J. Anat., 53:349-375. . Kramer. 0. °C: and: Barry, DyA- 1948 Valvular action in the embryonic chick heart by localized apposition of endocardial masses. Anat. Rec., 102:297-312. Pickering, J. W. 1893 Observations on the phys- iology of the embryonic heart. J. Physiol., 74; 383-466. Pressler, K. 1911 Beobachtungen und Versuche tuber den normalen und inversen Situs viscerum et cordis bei Anurenlarven. Roux’ Arch. Entw.- mech., 32:1-35. Rawles, M. E. 1943 The heart-forming areas of the early chick blastoderm. Physiol. Zool., 76: 99-45. Rudnick, D. 1935 Regional restrictions of po- tencies in the chick during embryogenesis. J. Exp. Zool., 77:83-99. 1938 Differentiation in culture of pieces of the early chick blastoderm. I. The definitive primitive streak and head-process stages. Anat. Rec., 70:351-368. , and Rawles, M. E. 1937 Differentiation of the gut in chorio-allantoic grafts from chick blastoderms. Physiol. Zool., 70:381-395. Ruud, G., and Spemann, H. 1923 Die Entwick- lung isolierter dorsaler und lateraler Gastrula- halften von Triton taeniatus und alpestris, ihre Regulation und Postgeneration. Roux’ Arch. Entw.-mech., 52:95-166. Sabin, F. R. 1917 Origin and development of the primitive vessels of the chick and of the pig. Car- negie Contr). to Embryol., 6:61—-124. 1920 Studies on the origin of blood-ves- sels and of red blood-corpuscles as seen in the living blastoderm of the chick during the second day of incubation. Carnegie Contrib. to Embryol., 9:21 3-259. 1922 On the origin of the cells of the blood. Physiol. Rev., 2:38-69. Sandstrom, R. H. 1934 The differentiation of hepatic and pancreatic tissues of the chick embryo in chorio-allantoic grafts. Physiol. Zool., 7:226— 246. Scharrer, E. 1939 The regeneration of end-ar- teries in the opossum brain. J. Comp. Neur., 70: 69-76. Schmitt, F.O. 1945 Ultrastructure and the prob- lem of cellular organization. The Harvey Lec- tures, 40:249-268. Severinghaus, A. E. 1930 Gill development in Amblystoma punctatum. J. Exp. Zool., 56:1-29. Slonimski, P. 1931 Recherches expérimentales sur la géneése du sang chez les Amphibiens. Arch. Biol., 42:415-477. Spemann, H. 1906 Uber embryonale Trans- plantation. Verhandl. Ges. deutscher Naturfor- scher u. Arzte, 78:189-201. Spratt, N. T., Jr. 1942 Location of organ-specific regions and their relationship to the development of the primitive-streak in the early chick blasto- derm. J. Exp. Zool., 89:69-102. Stdhr, Ph., Jr. 1924a Experimentelle Studien an embryonalen Amphibienherzen. I. Uber Ex- plantation embryonaler Amphibienherzen. Roux’ Arch. Entw.-mech., 702:426-451. 1924b Experimentelle Studien an em- bryonalen Amphibienherzen. II. Uber Trans- plantation embryonaler Amphibienherzen. Roux’ Arch. Entw.-mech., 103:555-592. 1925 Experimentelle Studien an embry- onalen Amphibienherzen. III. Uber die Entste- hung der Herzform. Roux’ Arch. Entw.-mech., 106:409-455. 1927 Experimentelle Studien an embry- onalen Amphibienherzen. IV. Roux’ Arch. Entw.- mech., 712:696-738. Stone, L. S. 1922 Experiments on the develop- ment of cranial ganglia and the lateral-line sense organs in Amblystoma. J. Exp. Zool., 35:421- 496. Storti, E. 1931 Sulla capacita ematopoietica dell’ endotelio nelle larve degli anfibi. Boll. Soc. Ital. Biol. Sperim., 6:97—99. Streeter, G. L. 1918 The developmental altera- tions in the vascular system of the brain of the human embryo. Carnegie Contrib. to Embryol., 8:5-38. 1927 Archetypes and symbolism. Sci- ence, 65:405—412. Verocay 1905 Multiplicitas cordis (Heptocardia) bei einem Huhn. Verhandl. der Deutschen pathol. Ges., Erg. Heft, 76:192-198. Vogt, W. 1929 Gestaltungsanalyse am Amphib- ienkeim mit Ortlicher Vitalfarbung. Roux’ Arch. Entw.-mech., 720:384—706. Waddington, C. H., and Waterman, A. J. 1933 The development ‘‘in vitro” of young rabbit em- bryos. J. Anat., 67:355-370. West, R. 1948 Activity of vitamin Byz2 in Ad- disonian pernicious anemia. Science, 107:398. Wilde, C. E., Jr. 1951 An in vitro study of the urodele neural crest. Anat. Rec., 777:92. Willier, B. H., and Rawles, M. E. 1931 Develop- mental relations of the heart and liver in the chorio-allantoic grafts of whole chick blastoderms. Anat. Rec., 48:277-302. Wintrobe, M. M. 1950 Factors and mechanisms in the production of red corpuscles. The Harvey Lectures, Ser. 45. Wislocki, G.B. 1939 The unusual mode of devel- opment of the blood vessels of the opossum’s brain. Anat. Rec., 74:409-428. Woellwarth, C. von 1950 Experimentelle Unter- suchungen iiber den Situs Inversus der Einge- weide und der Habenula des Zwischenhirns bei Amphibien. Roux’ Arch. Entw.-mech., 144:178- 256. Woerdeman, M. W., and Raven, C.P. 1946 Mon- ographs on the Progress of Research in Holland during the War. Elsevier Publishing Co., New York. Wolf-Heidegger, G. 1936 Experimentelle Stud- ien zur Genese der Langerhansschen Inseln des Pankreas. Roux’ Arch. Entw.-mech., 735:114 134. Yamada, H. 1933 Uber die Elimination der Le- ber bei den parabiotischen Bufolarven. Folia Anat. Japonica, 77:191-211. Section VII CHAPTER 6 Urinogenital System R. K. BURNS THE URINARY and genital systems of verte- brates are related in the adult only in having certain external passages in common; in development, however, most of the internal organs of reproduction are derived from parts of the primitive nephric system. The history of the remarkable transformations involved has long been familiar from com- parative morphological studies, but only in recent years have advances in theory and experimental techniques permitted rapid progress in analyzing the mechanisms of control and integration. As the morphological precursor of many genital structures, the nephric system must be given first considera- tion. THE NEPHRIC SYSTEM In many vertebrates the nephric system develops as three distinct entities—proneph- ros, mesenephros and metanephros—which appear successively in a regular temporo- spatial order (Fig. 167). However, in certain cyclostomes and primitive amphibians the system is essentially continuous, with little regional specialization (for a recent review see Fraser, 50). The units of the system at all levels develop from the intermediate meso- derm. The pronephric and more anterior mesonephric tubules have a simple meta- meric disposition, arising from _ discrete nephrotomes; posteriorly the tubules differ- entiate within an unsegmented cord of nephric material—secondary and _ tertiary elements, etc., developing as buds from the primary units, or from residual nephrogenic tissue in association with diverticula from the nephric duct. [For general accounts see Hall (04); Gray (32, ’36); Hamilton (52). For the relationship between meso- nephros and “definitive kidney” in amphib- ians see Gray (32) and Fraser (’50).] The metanephros of amniotes develops entirely in the latter fashion. Its nephrons differenti- ate within a blastema, more or less continu- 462 ous with the nephrogenic cord, in relation with the ureteric diverticulum. Thus, the tubules at all levels appear as independent primordia and only secondarily unite with the duct system. The nephric duct is laid down as the duct of the pronephros, serves subsequently as mesonephric duct, and as such gives rise to the ureter. The problems presented are concerned with (1) the history and progressive localization of the nephro- genic materials prior to the appearance of definitive nephric primordia, and (2) the nature of the integrative forces which co- ordinate the later development of the various parts of the system and relate them to the regional environment. THE TOPOGRAPHY OF NEPHROGENIC AREAS IN EARLY STAGES OF DEVELOPMENT The first problem to be considered con- cerns the topographic localization of the pro- spective nephrogenic materials at successive stages of development, leading up to the appearance of discrete nephric primordia; the related and concurrent problem of the progressive determination of these materials is dealt with elsewhere (Holtfreter and Ham- burger, see Section VI, Chapter 1). In early development most embryonic organ systems are not precisely localized, and after removal of prospective organ-forming areas extensive reorganizations are possible. Division of the amphibian egg may be followed by essen- tially normal development of its parts, in which nephric structures show the same regulative capacities as other systems. B1- lateral regulation of the nephric system may even occur in dwarf embryos derived from lateral pieces of the early gastrula (Holtfreter, ’38). At the beginning of gas- trulation, however, there is a definite con- centration of the prospective nephrogenic material in the posterior region of the mar- URINOGENITAL SYSTEM ginal zone, as shown by the consistent results of many vital-staining experiments (for reviews see Pasteels, °42; Nieuwkoop, °47). At this stage a rather sharply defined area representing the pronephros lies somewhat ventrolateral to the dorsal lip of the blasto- pore (Pasteels, °42). When this area is marked with a vital stain the color is later found to be confined almost entirely to the pronephros and its duct (Fig. 168). The area ~ ND NGC A MES< Fig. 167. Plan of development of the nephric sys- tem of vertebrates. A, Origin and mode of develop- ment of the nephric duct; B and C, its relations to the other parts of the system; PR, pronephric units; ND, nephric duct; NGC, nephrogenic cord; G, gonad; MES, mesonephric units; MET, metane- phros; UR, ureter; CL, cloaca. in question lies at the future cephalic end of a band of nephrogenic tissue which, in the course of gastrulation, passes inward around the lateral lip of the blastopore toward its final position in the lateral trunk region. In salamander embryos final topographic localization of the pronephric material oc- curs from late gastrula to middle neurula stages. In the early neurula the material has been invaginated and lies in the lateral body wall, anterior to the blastopore and below prospective somites 4 and 5 (Fig. 169, after Yamada, °37; see also Muchmore, 51). After this stage its position relative to adjacent regional structures (somites, limb area, gills, etc.) does not essentially change. By the 463 middle neurula stage, however, as a result of posterior elongation of the body axis, the location is relatively further forward (Fales, 35; see also Yamada, °37; Nieuwkoop, °47; Muchmore, °51). At this stage transplants of a limited area (Fig. 171) always produce pronephros, and contiguous areas have no pronephric potency (Fales). Nevertheless, Fig. 168. A, Position of the prospective pro- nephric area in the early gastrula of the axolotl, according to Pasteels (°42); B, later distribution of the stain. Fig. 169. Position of the pronephric area in the early neurula stage of the salamander, according to Yamada (737). Prn., Pronephric area; numerals in- dicate the position of the future somites. Bl., blasto- pore. Fig. 170. Position of the mesonephric material (stippled) and the lateral plate material (lined) in a salamander. A, Middle gastrula stage; B, tail-bud stage (after Nieuwkoop, *47). CH, Chorda; S, somite area; YP, yolk plug. the definitive pronephros-forming area is still not completely autonomous; the character of its final differentiation, as respects the type of tubule produced, and the details of tubule structure, depend on its regional environ- ment—a matter for later consideration (see p. 465). The topographic localization of the pro- spective mesonephric tissue has also been established by vital staining. In the middle gastrula stage of a salamander egg it les ventral and ventrolateral to the blastopore, in the form of an open collar (Fig. 170A, after Nieuwkoop, ’47). The sides of the collar are 464 undergoing involution around the lateral lips. (At this stage the pronephric material has for the most part passed in.) Mesonephric potency, however, is not restricted to this band. After extirpation there is extensive reconstitution of the nephrogenic and lateral plate material (Nieuwkoop). In neural plate stages the mesonephric material is still in process of involution. Two positions have been demonstrated independ- ently (Fig. 171). One lies just lateral to the neural fold, a little anterior to the blasto- pore (Fales, °35; Nieuwkoop, ’47); but at a slightly earlier stage an area within the posterior neural plate also produces meso- nephros (see Spofford, *45). The latter area Fig. 171. Position of pronephric and mesonephric areas in the middle neurula stage of Amblystoma, modified from Fales (735). The pronephric area is stippled; the adjacent areas indicated lack nephric potency. M1, position of prospective mesonephric material in the hinder part of the medullary plate, according to Spofford (’45). M2, position of mesone- phric material already invaginated and_ shifted laterally toward its definitive position (Fales, ’35; Nieuwkoop, ’47). evidently represents in part the uninvoluted portion of the nephrogenic band; the former consists of material which has been invagin- ated. After involution is complete, elonga- tion of the body axis shifts the mesonephric material forward to its definitive position (Fig. 170B). At this stage 90 per cent of grafts of the intermediate mesoderm posterior to the ninth somite give rise to well devel- oped mesonephric tubules (Humphrey, ’28a). In chick blastoderms of the head-process stage (comparable to the early neurula of amphibians) mesonephric potency is found in a limited area which includes the anterior part of the primitive streak and the node (Fig. 172A, after Rawles, ’36; see also Willier and Rawles, ’35). When portions of this area are transplanted to the chorio-allantoic mem- brane, mesonephros develops (in association with gonad and adrenal cortex). Nephro- genic capacity presumably resides in the mesodermal constituent; however, all three germ layers are included in such trans- plants. Somewhat earlier, in the definitive streak stage, a prospective nephric area lies SPECIAL VERTEBRATE ORGANOGENESIS close to the mid-line, a little behind the node (Fig. 172B; for a review and analysis see Rudnick, °44). - ge — _— ple eal) | A 8 Fig. 172. A, Location of the prospective mesone- phros- and gonad-forming areas in the chick blasto- derm of the head-process stage, adapted from Rawles; the area within the broken line represents mesonephros, the cross-hatched area gonad and adrenal. B, Location of nephrogenic material (pre- sumably mesonephros) at the stage of the definitive primitive streak, after Rudnick (’44); left side: epi- blast—the material is moving toward the primitive groove; right side: mesoblast, showing position of material already involuted and moving laterally. THE EXISTENCE OF NEPHRIC FIELDS Although topographically restricted areas representing pronephros or mesonephros can be defined in amphibian eggs as early as gastrula stages, nephrogenic potency is not limited to such areas until much later. In fact, capacity to produce nephric tubules is widely distributed in the gastrula, in keeping with the pluripotency of both ectoderm and mesoderm at this stage (for a discussion see Holtfreter and Hamburger, Section VI, Chap- ter 1). Almost any part of the ectoderm, epidermal or neural, may produce nephric tubules when transplanted into the _pro- nephric region (Holtfreter, ’33; Mangold, 24) or when exposed to the influence of “trunk organizer” (Holtfreter °36); and nephrogenic tissue inserted into the blasto- coele may induce nephric tubules in any overlying ectoderm (Holtfreter, ’33). Nephric tissue has been obtained also from various regions of the mesoderm which normally give rise to other tissues, for examples, the archen- teric roof (Holtfreter, °36) and somite ma- terial (Yamada, °37; Muchmore, 51). The fact that various tissues not normally nephrogenic become so when placed in a nephrogenic region, or brought in contact with nephrogenic tissue, led to the concept URINOGENITAL SYSTEM of a nephric field (see Holtfreter, °33). Systematic experimentation showed that gas- trula ectoderm may produce nephric tubules when implanted into any part of a lateral zone extending from the level of the gills to the cloaca (Fig. 173). However, as with embryonic fields in general, inductive power is not uniform throughout this zone but is strongest near the site of the pronephros. This was shown by the relative frequency of neph- ric differentiation in grafts placed at various levels in the field. Furthermore, grafts of the prospective pronephric area diminish both in size and in degree of differentiation as they are moved posteriorly from the normal posi- tion (Fales, ’35). Evidence of field effects may be adduced with respect to other nephric regions. Gas- trula ectoderm implanted in the mesonephric area of the posterior neural plate (Fig. 171) produces mesonephric tubules (Spofford, 48); on the other hand, mesonephric ma- terial moved forward into the pronephric area forms pronephric tubules, and its rate of differentiation is correspondingly accele- rated (Machemer, ’29). Finally, tubules of mesonephric type may be induced to develop precociously in the metanephric blastema of the chick by the action of abnormal in- ductors (Gruenwald, °42, °43). From such results it may be concluded that: (1) a variety of foreign tissues, meso- dermal and ectodermal, acquire nephrogenic potencies if introduced into a nephrogenic field or exposed to an appropriate inductor; (2) in the differentiation of such materials both the frequency and the special character (pronephric or mesonephric) of the differen- tiation are modified by the position of the material in the field. It has been suggested (Machemer, Gruenwald, op. cit.) that capa- city to produce nonspecific tubular struc- tures is essentially of the same order at all levels of the system, the specific character of the tubule depending on local factors. The appearance of sub-regions within a wider and more generalized nephric field is to be re- garded as an aspect of regional differentia- tion in the embryo as a whole. It must be recognized, however, that the inductive and integrative activities attributed to the field may function to a remarkable degree in relative isolation. Highly differen- tiated pronephric structures may be obtained from disorganized cell aggregates* differenti- * Taken from the dorsolateral lip of the blasto- pore. The material may possess original pronephric significance in part, or may acquire potency through induction by materials from the organization center, 465 ating in ectopic situations, or in vitro (Holt- freter, °44). Aggregation and reintegration of cell types on the basis of specific affinities, and assimilation of other types by induction, are postulated. In such tissue complexes it is likely that the essential constituents of the normal field environment are _ repre- sented. A R (0 Dd Fig. 173. Influence of the host field in determin- ing the differentiation of pronephros from gastrula ectoderm (Holtfreter, 33). A, Sources of donor ecto- derm taken from prospective neural (7) or epider- mal (2) areas of the early gastrula; B, C, and D, zones within which pronephric structures may dif- ferentiate from such ectoderm, when transplanted into host embryos of different ages. THE DIFFERENTIATION OF THE PRONEPHROS With respect to the pronephros it has been shown that during the gastrula stage exten- sive regulations are possible, and that vari- ous non-nephrogenic materials may be as- similated with essentially normal results (p. 464). However, by middle or late neurula stages (corresponding approximately with final restriction of potency; see p. 463) ir- reversible determinations have occurred. [There has been no analysis of pattern and axiation in the pronephros comparable to the analysis of the limb disc (see Section VII, Chapter 4).] In the frog Discoglossus, the dorsoventral axis is apparently fixed by the middle neurula stage, when inversion of the primordium results in an inverted pro- nephros (Tung, °35). In the late neurula. division of the pronephric area is followed by development of partial structures which correspond with the plane of section (Holt- freter, °44), and combinations of primordia in middle neurula stages result in mosaic organs with extra parts (Fales, ’35). The pronephric primordium is now essentially a mosaic, and problems of a different kind are introduced. To what extent are the individual] parts of the complex interdependent in final development; and what is the role of the field in their final differentiation? 466 The several parts (glomus, nephrostomal canals, tubules, common duct) have distinct origins (see Field, °91; Fales, °35; Cambar, °49). The glomus arises from the coelomic wall and may develop independently after excision of the pronephric primordium (How- land, ’21; Miura, ’30a; Fales, ’35), but an accessory glomus may also develop in hetero- topic positions in response to transplanted tubules (Fales). Nephrostomal canals de- EXPERIMENT AUTHOR SPECIAL VERTEBRATE ORGANOGENESIS is increasingly impaired (Fales, °35). When older donors are used, however, developmeni is better at all levels. Grafts to the body cavity yield only disorganized tubules, but again the quality of differentiation varies somewhat with the age of the primordium. Apparently two major factors are operative at this period: an organization in the pri- mordium, expressed by capacity for autonom- ous differentiation, which increases with FORM USED EFFECT ON NEPHRIC DUCT Ectopic pronephros — transplanted or induced Holtfreter Triton Ducts grew out from pronephros Excision of entire pronephric Primordium Burns Amblystoma Duct absent on operated side Excision of pronephric primordium Waddington Rana Obstruction of duct primordium O'Connor Amblystoma Pleurodeles Excision or obstruction of duct primordium Tung ond Ku Bufo Rana Obstruction of duct primordium Van Geertruyden Rana Excision of entire pronephric Pprimordium Nieuwkoop Triton Interference with duct primordium Holtfreter Triton by cutting posterior to cut Amblystoma ue uM Staining of primordium of duct O'Connor (somites 5-7) Triton Entire duct contains stain Amblystoma Staining prospective mesonephric material Spofford Amblystoma Mesonephros stained - duct clear Tip of duct destroyed by cautery Boyden 1932 Chick Absent caudal to site of operation = Tip of duct obstructed by incision Waddington 1938 . " D " ceo " 4 Grunwald 1937 : - " " i an " Tip of duct destroyed by cautery Tip of duct blocked by graft Gruenwald Fig. 174. Table showing effects of various experimental procedures on the development of the nephric duct, with respect to its origin and manner of formation. Staining experiments (O’Connor, ’38) indicate a local origin from a primordium adjacent to and continuous with the pronephros. Asterisks indicate cases in which a short segment of duct was found posteriorly, in communication with the cloaca. According to O’Connor (739, ’40) such pieces are of cloacal origin. velop from the somatopleure underlying the tubule primordia, and are said to remain after removal of the primordium (Howland, 21); but this is also denied (Dalcq, 742; Cambar, °49). However, primordia develop- ing in ectopic locations, or in the coelom, usually produce tubules without nephros- tomes (Fales). In later development there is marked interdependence of parts, probably mediated by functional influences. For ex- ample, removal of the glomus, or obliteration of nephrostomal canals, leads to atrophy of the tubules; and reducing the number of tubules results in shrinkage of the common duct (Miura, *30a; Shimasaki, ’30a; Fales, ap) In the middle neurula stage transplanta- tion of the pronephric primordium in the orthotopic position results in normal develop- ment, but at posterior levels differentiation age, and regulative influences essential for normal growth and development, exerted by the regional environment. ORIGIN AND DEVELOPMENT OF THE NEPHRIC DUCT The classic view derives the nephric duct from a union of the ends of the anterior pronephric tubules (Fig. 167; for reviews of normal development see Goodrich, °30; Fraser, 50). In many amphibians, however, the tubules do not appear as discrete units but arise from a continuous primordium, the posterior extremity of which represents the duct (e.g., Field, 91). A vital stain applied to the pronephric swelling of a salamander ventral to somites 3 and 4 appears subse- quently only in the tubules; but if applied ventral to somites 5-7 the color is later con- URINOGENITAL SYSTEM fined to the duct (O’Connor, °38), which presumably has a distinct origin from inter- mediate mesoderm at this level* (see also Fig. 168; Pasteels, 42). Transections through the pronephric region at the neurula stage indicate approximately the same localization (Holtfreter, ’44). More difficult to determine is the manner in which the duct reaches the cloaca. On this point conflicting views have long existed. The classic theory envisages independent caudal growth of a bud of tissue arising at the pronephric level (Fig. 167), but an alternative view (Field, °91, and others) postulates formation in situ by incorporation of local materials at each successive level (hence the term ‘segmental duct”). The problem has been investigated experiment- ally by three methods: (1) excision of the pronephric primordium before outgrowth of the duct; (2) interference with the growth of the duct at various levels; (3) the use of vital staining to identify the duct materials. The results are summarized in Fig. 174. In brief, it appears that removal of the pri- mordium, or interference with the tip of the duct, prevents development posterior to the point of intervention (Fig. 175); but if the growing tip is left intact, development continues. Stain applied to the duct rudiment at the point of origin (as noted previously) appears subsequently throughout its length, but not in any adjacent tissue (O’Connor, 38). Conversely, if the prospective meso- nephric material is stained prior to the for- mation of the duct, the latter is subsequently found to be unstained (Spofford, ’48). The evidence supports the view that the duct develops by caudal growth of an original primordium rather than by local accretion. An exception must be made with respect to the posterior end of the duct. A short piece— often less than one somite in length—is sometimes found in contact with the cloaca, in the absence of the main part of the duct (Machemer, 729; O’Connor, ’39, °40; Holt- freter, 44; Tung and Ku, *44; van Geertruy- den, 46). This piece is considered by O’Con- nor to be of cloacal origin. However, the territory traversed by the duct is not without influence on its develop- ment. In normal development the duct grows backward in a narrow groove between the ventral edges of the somites and the lateral mesoderm, and covered by ectoderm. Ob- structions placed across this pathway usually *TIn anuran embryos the duct primordium lies at the level of somite 5 according to Dalcq (742) van Geertruyden (742), and Cambar (749). 467 prevent further development (see Fig. 174); or if the tip of the duct is rotated it usually fails to grow far into strange territory (Tung and Ku, °44). On the other hand, if the pathway is reoriented or dislocated in various ways, the duct may deviate considerable distances in order to reenter and traverse it (Holtfreter, "44; Tung and Ku). The duct is able to traverse the path in either direction, or two ducts may be made to do so in oppo- site directions (Tung and Ku). When seg- ments of the duct are excised, regeneration Fig. 175. A, Absence of the nephric duct in a chick embryo (left side) after preventing the back- ward growth of the primordium by transection at the 12-somite stage; B, failure of the mesonephric tubule to develop on the operated (right) side im the absence of the nephric duct—only a formless blastema is present (after Waddington, *38); ND, nephric duct; BL, blastema; T, mesonephric tubule. proceeds along the pathway from either or both ends, (Howland, ’26; Tung and Ku). Evidently no irreversible polarizations are involved, either of the pathway or the duct. The properties of the path favoring develop- ment of the duct are perhaps in large part mechanical, but the importance of physio- logical factors has also been stressed (Holt- freter, ’44). In amphibians the nephric ducts terminate in a pair of cloacal diverticula (cloacal horns) which are histologically distinct from the duct (Field, ’91; O’Connor, °40); in avian embryos union is with the urodeum. Studies of this relationship have yielded dif- ferent results. In amphibians cloacal horns develop normally in the absence of the duct, and may even elongate, replacing its hinder end (O’Connor, ’39, ’40). Elimination of the duct in chick embryos results in reduction and abnormal differentiation of the urodeum (Boyden, ’24). The duct readily unites with grafts of cloacal tissue placed along its path 468 (O’Connor, 40), but may also join other regions of the gut (O’Connor, ’39; Holtfreter, 44). THE DEVELOPMENT OF THE MESONEPHROS Up to the middle neurula stage at least, while the mesonephric material is still in process of involution, determination within SPECIAL VERTEBRATE ORGANOGENESIS act as an inductor (for a discussion see Griin- wald, °37). Experiments in which develop- ment of the duct was prevented in the meso- nephric region provide a direct test of this hypothesis (Fig. 176). The results, however, vary greatly and are inconclusive. In most cases only rudiments of tubules, or mere cel- lular condensations, appear in the absence of the duct (Fig. 175B); and in one species no visible blastema develops. But if the duct is EXPERIMENT AUTHOR FORM USED | STATE OF MESONEPHRIC TUBULES —_—_—_—_+ SF ESS Grafts of mesonephros _no duct present Humphrey 1928 Amblystoma Well-differentiated tubules Excision of nephric duct Miura 1930b Rana, Bufo No development of tubules Excision of nephric duct Shimasaki 1930b | Bufo Irregular differentiation of tubules Excision of pronephros and duct Burns 1938 Amblystoma Irregular differentiation of tubules = at —_ : —— sos = Excision of pronephros and duct Waddington 1938 Rana Local condensations of cells Obstruction of duct primordium O'Connor 1939 Rana Local| condensationsiof cells | Pleurodetes No condensations present = = — - eee = jae UAE! — Obstruct { ing duct Holtfret 1944 ten Local condensat f cell Dstruction of growing duc oltfreter Amblystoma | ocal condensations of cells Excision or obstruction of duct Van Geertruyden 1946 Rana | Cellular blastema forms —disappears } { = : = +8 a ee Removal of duct after formation Gambar 1948 Rana Local condensations of cells Partial removal or displacement of duct Alytes Tubules develop only in close proximity to duct Ti fiducvidestrovedb ut Bowden 1927. 1932 Chick Local condensations of cells Pores S y bf AIS 7 oY z which later disappeor Tip of duct obstructed by incision Waddington 1938 Chick Local condensations of cells Tip of duct destroyed by cautery Grunwald 1937 Chick No tubules caudal to lesion Tip of duct blocked by graft Gruenwald 1942 Chick Irregular differentiation of tubules es | Fig. 176. Table showing the effects of absence of the nephric duct on the differentiation of the tubules of the mesonephros. In most cases tubules fail to differentiate, although local cellular condensations may appear; in some species, however, variable development of tubules may occur, subject to later degeneration. the area is labile to the extent that foreign material is still readily assimilated (Spof- ford, 48). Subsequently two degrees of auton- omy can be demonstrated. In later neurula stages mesonephric tissue placed in ectopic situations (coelom, eye chamber) is capable of producing unspecific tubular structures, but is still plastic to a degree—at the site of the pronephros it precociously forms tubules of pronephric type (Machemer, ’29). In this phase, rate of development and final character of morphogenesis are still modified by external forces. After the tail-bud stage determination is fixed. The same material produces typical mesonephros in any loca- tion (Humphrey, ’28a); and nephrogenic potency is restricted to the prospective meso- nephric material (e.g., van Geertruyden, "46: Cambar, °48; cf. also Gruenwald, 42). The mesonephric tubules originate from nephrotomes or as condensations in the un- segmented nephrogenic cord (p. 462) long after the nephric duct has developed—a fact which long ago suggested that the duct might removed after brief contact with the nephro- genic tissue, delayed but fairly complete dif- ferentiation occurs (O’Connor, ’39). Again, if the duct is displaced, tubules develop only where nephrogenic tissue remains in close proximity to the duct (Cambar, ’48). In such cases the role of the duct seems clear. In other cases, however, the tubules behave in an irregular fashion, showing considerable autonomy. All stages of differentiation may be realized in the absence of the duct, even in the same species or individual (e.g., Shi- masaki, °30b; Humphrey, ’28a,b; Burns, 38; Gruenwald, *42: Nieuwkoop, ’48). The im- portance of the duct may well vary in differ- ent species, or other inductors may be in- volved (see Gruenwald, ’42, ’43; van Geer- truyden, ’46). Account must be taken also of the fact that all tubule primordia (even in the same region are not of the same morpho- logical order (e.g., Hall, ’04; Gray, ’32) and may possess different capacities for self- differentiation. These questions require further investigation. URINOGENITAL SYSTEM THE DEVELOPMENT OF THE METANEPHROS The development of the metanephros pre- sents problems of a similar nature. The em- bryonic ureter is an outgrowth of the nephric duct, while the kidney tubules differentiate from the adjacent metanephric blastema. Here the analysis rests mainly on experi- ments with chick embryos. Interference with the growth of the nephric duct at anterior levels results in absence of its hinder part (see Fig. 174) and in consequence the ureter is also lacking. In this case the metanephric tubules fail to differentiate (Boyden, 27; Griimwald, °37) although a blastema appears and may persist for a time. This result is consistent with the generally accepted inter- pretation of agenesis of the kidney (Boyden, 32; Grunwald, ’37,’38; Auer, 47). Congenital absence of a kidney is always accompanied in males by absence of the ductus deferens (nephric duct) and ureter on that side. Furthermore, development of double ureters (a well known anomaly) results in the formation of two kidneys, which may show various degrees of fusion according to pro- pinquity. In an early stage of this anomaly each ureter has its separate blastema (e.g., Wharton, °49). Evidently the appearance of a blastema and the subsequent differentiation of tubules are distinct phases. The first is independent of the ureter (although if two ureters are present twin blastemas may be formed; cf. the eye-forming area of the neural plate), but the ureter must be present for the differ- entiation of tubules. This conclusion accords with the fact that metanephric potency has not been demonstrated in parts of the chick blastoderm, but development readily occurs in explants of the metanephric region after the ureter has formed (see Seevers, ’32). The ureter is held to be essential. It is reported that tissues other than the ureter (nephric duct, nervous tissue) may induce differentiation in the metanephric blastema, but the tubules formed are atypical and of mesonephric type (Gruenwald, 42, "43; see p. 465). THE ORIGIN AND DEVELOPMENT OF THE OVIDUCT The embryonic oviduct (Miillerian duct) develops much later than the nephric duct; however, the two structures are closely re- lated in origin and mode of development. In certain fishes and amphibians the ostium 469 is considered to arise from one or more persistent pronephric tubules, or their ne- phrostomes. The duct reaches the cloaca by the backward growth of a cord of cells from the ostial primordium (for reviews consult Goodrich, ’30; Brachet, °35; Willier, ’39; Gallien, ’44). In amniotes the ostium is said to be formed by invagination of a special area of the coelomic epithelium, near the anterior pole of the mesonephros, and the duct is again formed by backward growth. The resemblances are obvious but in amni- otes the role of pre-existing nephric rudi- ments is not generally admitted. Neverthe- less, origin of the ostium from pronephric remains, or from funnel-like structures re- sembling nephrostomes, has been reported in certain mammals (see von Winiwarter, ’10; Brambell, 27; Burns, *41). Recently, also the old problem of the relationship of the developing oviduct to the nephric duct has been re-examined. In both avian and mam- malian embryos the growing tip of the duct is almost inseparably applied to the wall of the nephric duct, from which it may derive material (Gruenwald, *41). Altogether, a nephric origin for the Miillerian duct is probable. The main problems in the development of the oviduct are (1) the origin and nature of the ostial primordium and (2) the mode of extension of the duct to the cloaca. On the first point experimental evidence is limited*; with respect to the second it is clear that development of the oviduct depends on the presence of the nephric duct. In chick em- bryos in which the posterior part of the nephric duct is lacking as a result of opera- tion at anterior levels (Griinwald, ’37) the ostium is present, plus a segment of the oviduct corresponding exactly to the surviv- ing portion of nephric duct. Although devel- opment had been initiated it did not proceed beyond the termination of the nephric duct. Further evidence is found again in studies of renal agenesis. Absence or partial develop- ment of the nephric duct is accompanied by corresponding deficiencies in the female gen- ital tract (Grunwald, ’38, ’41; Auer, 47). The nature of the dependence is not estab- lished. The nephric duct may be merely a guide—an essential feature of the path along * According to unpublished observations of the author, early extirpation of the pronephros (see Fig. 176), with absence of the nephric duct, is fol- lowed long afterward by failure of either oviduct or ostium to appear on the operated side. Formation of the ostium seems to depend in some way on the presence of the pronephros, cf. Gallien (744). 470 which the oviduct grows—or it may also contribute material for its formation. THE GENITAL SYSTEM Most of the structures which make up the embryonic genital system have been taken over from other systems, and their readapta- tion to genital functions is a secondary and relatively late phase in their development. The early differentiation of such structures is therefore independent of sexuality. Also, each embryo is at first morphologically bi- sexual, possessing all necessary structures for the differentiation of either sex. The differ- entiation of one set of sex primordia and the gradual involution of the other is normally determined by the sex type of the gonad. The initial problems, then, concern the con- stitution of the gonad primordium and the factors which direct its evolution into an ovary or a tesiis. STRUCTURE AND ORGANIZATION OF THE EMBRYONIC GONAD The sexually undifferentiated gonad is a composite structure. Male and female po- UNDIFFERENTIATED GONAD TESTIS Fig. 177. Diagram illustrating the role of medul- lary (stippled) and cortical components of the un- differentiated amphibian gonad in the differentia- tion of ovary and testis. Small arrows indicate the antagonistic inhibitory action exerted by each com- ponent against the other until final dominance of one is established. tentialities are represented by specific histo- logical elements, medulla and cortex, which have alternative roles in gonadogenesis (Fig. 177). Normal differentiation involves the gradual predominance of one component, while in various types of intersexuality the recessive element persists in varying degree. The extent to which the recessive component develops in the embryo, and the duration of SPECIAL VERTEBRATE ORGANOGENESIS the bisexual phase, differ widely from group to group and greatly influence reversibility under experimental conditions. In amphib- ians, in which both components are well developed over a relatively long period, transformation occurs readily; in amniote embryos, however, development is rapid and the recessive component tends to be vestigial and transient. Under such conditions the state of differentiation at the moment of experimental intervention may be a decisive factor in the result. The medullary and cortical components of the indifferent gonad are laid down through the activities of the somatic or non- germinal constituents of the genital ridge, conveniently called the structural elements. In the early testis of an amniote embryo (Fig. 178) the medullary component consists of primary sex cords, the cortical element is represented by the regressing germinal epi- thelium; however, as long as this epithelium is present, the testis is potentially bisexual. Conversely, in an ovary the male component is represented by more or less rudimentary medullary cords (primary sex cords) and rete elements. All attempts to control sex differ- entiation experimentally undertake by vari- ous means to influence the development of one sex component at the expense of the other. The germinal elements of the gonad are the primordial germ cells, which are often recognizable long before the gonad primor- dium appears. Nevertheless, as will be shown, they are apparently not essential either for the formation of a genital ridge or for the differentiation of specific histological structure. Eventually their fate as gametes depends upon the differentiation of the struc- tural elements and their role in morpho- genesis is secondary. EARLY TOPOGRAPHIC LOCALIZATION OF THE GONAD CONSTITUENTS In most vertebrates the germ cells and the structural elements of the gonad have dis- tinct and sometimes widely separated origins. Their history, prior to final localization in the genital ridge, may be for the most part considered separately. The Structural Elements. In embryos of urodele amphibians, from yolk-plug to mid- dle neurula stages, both structural elements and germ cells are found together in the region of the prospective lateral plate. Heter- oplastic exchange of material from this area at the yolk-plug stage (Fig. 170A) results in URINOGENITAL SYSTEM composite gonads containing both germ cells and structural elements derived from the eraft (Nieuwkoop, °47). In tail-bud stages the two elements are still closely associated in the intermediate mesoderm (Humphrey, ’28a,b). Transplantation of the gonad-meso- nephros complex at this stage shows it to be Coelomic epithelium Glomerular capsule Germinal epithelium A Tunica albuginea TESS aves Rete cords Glomerular capsule Germinal epithelium 471 of the blastoderm (Fig. 172A; see Rawles, 36). (At this stage the germ cells are pre- sumed to be located peripherally, in the “germ cell crescent”; p. 472.) Transplants which include the anterior end of the primi- tive streak and node produce gonad tissue containing testis cords (in association with Primary sex cords Rete cords |) Glomerular Germinal epithelium Medullary cords (primary sex cords) Rudimentary tunica albuginea \ Lees VAS? -- Glomerular capsule SI Cortical cords Germinal epithelium (secondary sex cords) Fig. 178. Diagrams illustrating the origin of medullary and cortical components in the histologically more complex gonads of mammals. A, Genital ridge, with incipient primary sex cords originating from the germinal epithelium. B, Indifferent gonad, in which the medullary component is present as the primary sex cords, while the cortex is potentially represented by the germinal epithelium. C, Development of the testis, with disappearance of the germinal epithelium. D, Development of the ovary, with the production of dominant secondary sex cords and cortex. capable of autonomous differentiation. Gon- ads of specific sex are produced, the fre- quency increasing with age. In the tail-bud stage of anuran embryos, on the other hand, only the structural ele- ments are found in the mesoderm; the germ cells lie in the “‘yolk ridge.” When the inter- mediate mesoderm is exchanged between em- bryos of different sex, the gonads which develop always correspond to the sex of the donor embryo (Humphrey, °33). Thus the capacity to produce a gonad of specific sex resides in the structural elements. In the chick a localized gonad-forming area exists as early as the head-process stage mesonephros and adrenal cortex), but such gonads are sterile (Willier, ’37, °50). The structural elements of the gonad are thus localized topographically long before the formation of a gonad primordium. In the embryos of mammals little is known of the localization of the structural elements before the stage of the genital ridge, but close association with the nephrogenic ma- terial is assumed. The Primordial Germ Cells. The history of the primordial germ cells of vertebrates has long been a controversial field. The classic theory of early segregation, as opposed to a later and more local origin, has been sup- 472 ported in many groups and as often denied. For a survey of the general status of the germ cell theory see Heys C3i)5 formmore recent discussions see Willier (39), Nieuw- koop (’47, 49), Gillman (’48), Witschi (48) and Johnston (751). There is as yet no ade- quate experimental analysis for the early stages of development. But if the ultimate source of the germ cells in most cases re- mains in doubt, it is generally accepted that they originate outside the gonad-forming area. Source of the Primordial Germ Cells. In only one case is there significant evidence indicating early segregation of a germinal material in a vertebrate. In the egg of a frog a cytoplasmic substance in the yolk near the vegetal pole has been traced to the “volk ridge” of the gut (Bounoure, °34), the site where the germ cells are generally identified in anurans. Irradiation of the vegetal pole results in gonads which are sterile (see Bounoure, 50). In urodele eggs, on the other hand, elimination or addition of polar yolk material does not seriously affect the de- velopment of germ cells (Nieuwkoop, °47, ONE In most vertebrates the germ cells as such have first been identified in various parts of the entoderm: the “yolk ridge” of anuran am- phibians, the entoderm of the yolk sac in rep- tiles, the splanchnopleure of the proamnion region of the avian blastoderm, and the hind gut or adjacent yolk sac entoderm in mam- malian embryos. Urodele amphibians are a notable exception. In this group most in- vestigators have first identified the germ cells in the “intermediate mesoderm” of the pos- terior trunk (see Humphrey, 725, ’29). A recent experimental analysis (Nieuwkoop, 47, °50) shows that in late yolk-plug and neurula stages they occupy a corresponding position in the prospective lateral mesoderm, ventral and ventrolateral to the blastopore (Fig. 170A). Removal of this area in the early neurula results in sterility. Neverthe- less, the entoderm is in some way involved in their differentiation as germ cells. In its ab- sence germ cells fail to appear in the gonad, or are greatly diminished in number. It is concluded that precursor cells in the prospec- tive lateral mesoderm are able to differen- tiate into germ cells only after contact with the dorsocaudal entoderm (Nieuwkoop). In avian embryos the germ cells are usually said to lie in the extra-embryonic splanchnopleure, near the germ wall In the chick they tend to be concentrated anteriorly in the “germ cell crescent” (see Swift, 714). SPECIAL VERTEBRATE ORGANOGENESIS Destruction of the crescent or isolation of the gonad-forming area of the early blastoderm both result in genital ridges or gonads which lack germ cells (for a discussion see Willer, 39, °50). The evidence supports the view that the cells in question are indeed pri- mordial germ cells but it is hardly crucial. Migration of the Primordial Germ Cells. If the germ cells originate outside the gonad region, sometimes at a considerable distance, there is a problem as to how they reach their destination in the genital ridge. In urodele amphibians the situation is fairly simple. From the earliest known stages, the germ cells lie with the other gonad constituents in the lateral plate mesoderm, and appear to move passively with the lateral plate ma- terials to a point medial to the Wolffian duct (Humphrey, ’25; Nieuwkoop, ’47). For entry into the genital ridge, independent “‘amoe- boid movement” directed by an influence emanating from the ridge has been sug- gested. In anurans, according to Bounoure (above), the germinal material is carried from the vegetal pole to the “yolk ridge” by the movements of gastrulation and dorsal closure of the gut. The “yolk ridge” is then separated from the gut by the growth forces which produce the mesentery, after which only a slight lateral shift to the genital ridge is necessary. For this, independent migration and a directive influence from the ridge are again postulated. In amniote embryos migration from the yolk sac or gut entoderm involves more difficulty. In most cases independent migra- tion through the splanchnopleural mesen- chyme is assumed, based on the distribution of germ cells from stage to stage, and the appearance of “pseudopodial” form (e.g., see Witschi, °48). However, passive transport may again be a factor locally, as when germ cells embedded in the coelomic epithelium (prospective germinal epithelium) are shifted around the dorsal angle of the body cavity (see Willier, °39). The most unusual ‘“‘migra- tion” is described in bird embryos, where the germ cells presumably move from the “germ cell crescent” to the gonad region. According to the classic theory of Swift (714) they are enveloped by extra-embryonic blood vessels and enter the embryo, some escaping in the gonad region. Definite proof of this hypothesis is lacking. Destruction of the crescent area, or early isolation of the gonad- forming region, both result in sterility, indicating that the germ cells have been eliminated or excluded. However, this re- sult sheds little light on the pathway or URINOGENITAL SYSTEM manner of migration (for a discussion see Willier, °39). FORMATION OF THE GENITAL RIDGE In a strict sense gonadogenesis begins with the formation of the genital ridge. Two problems are presented: (1) the role of ex- trinsic or regional factors in the origin of the ridge, and (2) the relative importance of the primary gonad constituents: germ cells vs. structural elements. The first problem is inseparable from the larger question of re- gional organization and has not been com- pletely analyzed. A recent study indicates that general as well as local factors are in- volved in the production of the ridge (Nieuw- koop, *47, °50). In the urodele, Triton, dif- ferentiation depends first on early contact between lateral plate material (future peri- toneum) and the entoderm. Later, folds of coelomic epithelium, closely resembling gen- ital ridges, may be induced independently by notochord, Wolffian duct, and probably by mesonephric tissue. In later development gonads may be present in the absence of mesonephros (e.g., Humphrey, ’28a,b; Griin- wald, ’37). It is well established that the germ cells are not essential for the origin of a gonad. In amphibian embryos sterile genital ridges may develop after removal of the germ cells (e.g., Humphrey, ’27, ’28a; Nieuwkoop, °47); and elimination of the primordial germ cells of the chick before they reach the gonad region does not prevent the formation of gonads (for a summary see Willier, °39). Furthermore, germ cells in ectopic situations are unable to initiate gonad formation (Humphrey, ’28b; Willier, 33; Witschi, ’34). These results are opposed to the view that the primordial germ cells act as inductors (Dantschakoff, ’32) or are otherwise essen- tial for the origin of a gonad. The evidence indicates that (1) the local appearance of a genital ridge is conditioned by regional influences; (2) primordial germ cells alone cannot induce a genital ridge and are not essential for its origin; and (3) the formation of the ridge is an activity of the structural elements. THE ROLE OF GERM CELLS VS. STRUCTURAL ELEMENTS IN SEX DIFFERENTIATION The conclusion last stated may be ex- tended to cover the period of sex differentia- tion. The characteristic structure of testis or 473 ovary may be developed in the absence of germ cells. Irradiation of the vegetal pole of the frog’s egg results later in well dif- ferentiated testes and ovaries which are completely sterile (Bounoure, °37, ’50), and tadpoles that develop from overripe eggs eat ere Fig. 179. Sterile gonad, of testicular structure, in a chorio-allantoic graft derived from a small piece of blastoderm contaming Hensen’s node, taken from an embryo of the head-process stage (from Wiullier, *39). Fig. 180. Diagram of an experiment demonstrat- ing the role of the structural elements (intermediate mesoderm) in determining the sex-type of the gonad (after Humphrey, ’33). A, Frog embryo, showing the area of intermediate mesoderm (broken lines) containing the gonad-forming elements. When this area is transplanted to a corresponding position in an embryo of different sex, a situation is set up as in B, in which the structural elements of one sex (area A-B) are associated with the primordial germ cells (GC) of the other. The sex-type of the gonad which develops is always determined by the struc- tural elements. have defective gonads which are typically sterile (Witschi, 51a). When the gonad- forming area of the chick blastoderm is iso- lated, in the absence of germ cells, sterile testicular structures are produced (Fig. 179; Willier, ’37, 50); and gonads of both sexes differentiate after destruction of the germ cells by irradiation (Salzgeber, *50; Dul- becco, *46). It has been shown further that the struc- 474 tural elements establish the sex-type of the gonad, irrespective of the genetic constitution of the included germ cells. If the intermedi- ate mesoderm of a frog embryo is trans- planted orthotopically to an embryo of dif- ferent sex, the germ cells of one sex are combined with the structural elements of the other (Fig. 180). It is the differentiation of the latter that determines the sex type of the gonad (Humphrey, ’33). This principle applies, moreover, in all cases of functional sex transformation, which end in the pro- duction of gametes of the opposite sex (pp. 475-476). Even with respect to gamete forma- tion the germ cells are indifferent or bi- potential. APPEARANCE OF SEX-SPECIFIC ORGANIZATION IN THE GONAD The time at which the gonad primordium becomes organized with respect to its future sex has been determined by testing its ca- pacity for self-differentiation at successive stages of development. In amphibian em- bryos transplantation of the gonad-forming area shows that sex-specific organization exists long before the appearance of a genital ridge (Humphrey, 28a), but in these experi- ments mesonephros and other regional ele- ments are also present in the graft. The gonad-forming area of the chick blastoderm (p. 471) likewise shows a degree of sex- specific organization, subject, however, to the same qualification. A certain number of isolates produced gonad-like bodies, with sex cords of male type and the structure of rudimentary testes (Fig. 179). Ovaries were not formed. These experiments are not strict tests of the capacity of the gonad for in- dependent differentiation. During the formation of the genital ridge, analysis shows that the capacity for specific differentiation appears and increases from stage to stage (see Willier, ’39). After a distinct germinal epithelium is present, transplants of the ridge yield gonads of specific sex with much higher frequency than before the germinal epithelium is recognizable; at earlier stages (31 somites or less) only gonads of indeterminate sex are formed. It may be concluded that during formation of the genital ridge in the chick, morphogenetic changes occur which gradu- ally determine the sex-type of the gonad. In the duck embryo, as in the chick, the his- tologically undifferentiated gonad is strongly organized with respect to sex and, in the SPECIAL VERTEBRATE ORGANOGENESIS female, to laterality as well. Gonads removed before the beginning of morphological sex differentiation, and cultivated in vitro, pro- duce structurally normal testes, or right and left ovaries. Such gonads are sometimes sterile (Wolff and Haffen, *52). The organization of the gonad primordium of the rat has also been analyzed by trans- plantation in various ways. Mesonephric bodies isolated before a genital ridge is pres- ent do not produce gonads, but in the ridge stage male organization is well established (Torrey, 50). When sexually undifferenti- ated primordia are transplanted to the adult kidney (Buyse, °35) there is a striking sex difference in capacity for self-differentiation. (This study was carried out to test the pos- sibility that the differentiation of the grafted gonads might be modified by the hormones of the host. From this standpoint the results were negative.) Testes develop normally but prospective ovaries lack stable organization and give rise to the following types: (1) retarded ovaries; (2) ovotestes, in which both gonad components are developed but more or less rudimentary; (3) rudimentary testes, in which only the medullary component of the ovary has survived; and (4) grafts of indeterminate sex. Evidently, testis organiza- tion is strongly fixed in the indifferent primordium (cf. Torrey, *50) but the pro- spective ovary is extremely labile, the medul- lary component developing with about the same frequency as the cortical. Similar results were obtained after transplantation to other sites (Moore and Price, ’42;* Holy- oke, ’49), with indications that the lability of the ovary is related to its age at transplanta- tion. Further analysis of the age factor (Torrey, 50) shows that the organization of the testis is fixed as early as the 12-day pri- mordium (early gonadal blastema stage). The critical period for the ovary comes much later, at 15 to 16 days (cf. Moore and Price), when grafts may give rise to ovaries, ovo- testes or rudimentary testes. By the seven- teenth day (when a cortex is histologically present) all grafts produce typical ovaries. Evidently the sex-specific organization of the gonad primordium is gradually acquired and is realized much earlier in the testis than in the ovary. This fact corresponds with the normal order of development, in which the medullary component (primary sex cords) has priority of origin in both sexes. It is in * These authors did not describe distinct ovotestes but “transformed” and hypertrophied medullary cords were frequently found. URINOGENITAL SYSTEM contrast with conditions in amphibians (Humphrey, ’28a) and birds (Willer, °39; Wolff and Haffen, 52), in which the indif- ferent gonad primordium in both sexes shows sex-specific organization. THE DIFFERENTIATION OF SEX IN THE GONAD The lability of gonad organization during the stages when the sex components are be- ing laid down has both a physiological and a structural basis. Development of ovotestes or rudimentary testes from ovaries may be favored by conditions in the graft environ- ment which interfere with cortical differen- tiation; but the immediate basis for reversal lies in the presence of a differentiated medul- lary component. Differentiation is essentially a competitive process which may be upset by any condition which effectively inhibits the dominant component—a view which is consistent with the fact that reversal of sex can be induced by various and apparently unrelated experimental procedures. Effect of Eliminating the Dominant Gonad Component. The simplest test of this con- cept is direct elimination of the dominant gonad component. In early development this procedure is difficult, but two classic experi- ments may be cited which are based on survival of a recessive component long be- yond the period of sex differentiation. In adult male toads a part of the embryonic cortex survives as the organ of Bidder. After removal of the testes this structure slowly develops into an ovary, capable of producing fertile eggs (Ponse, ’24). A parallel case is the removal of the left ovary of the recently hatched female chick. After this operation the rudimentary right gonad, composed chiefly of medullary tissue, frequently de- velops into a testis which may be fertile (see Domm, 7°39). In each case a long dormant recessive component retains its capacity to produce a gonad of opposite sex when re- leased from inhibition by removal of the functional gonad. Sex Reversal Induced by Unspecific or Environmental Agencies. It is well known that in many animals transformation of sex may be induced by various physiological conditions such as nutritional level, the effects of parasitism, disease, or even by fac- tors in the physical environment. A familiar case is the influence of temperature on the differentiation of the gonads in certain amphibians. High temperatures favor dif- ferentiation of the medulla, and ovaries are gradually transformed into testes (Witschi, 475 29; Piquet, °30). The primary effect, how- ever, seems to be a degenerative change in the cortex, which is followed by medullary differentiation. Low temperatures, on the other hand, promote cortical development by retarding the differentiation of the me- dulla. Temperature therefore exerts its ef- fects by interference with the dominant component of the gonad, and the result is again consistent with the concept of compet- itive interaction between medulla and cor- tex. A comparable effect of temperature on embryonic mammalian gonads has_ been shown (Torrey, ’50). A B Fig. 181. Methods of inducing sex reversal in am- phibians by combining whole embryos or gonads of different sex. A, Parabiosis, in which embryos (of the same or of different species) develop with a com- mon circulation; 8B, orthotopic transplantation, bringing together gonads of different sex in the same individual. Hormonal Control of Gonad Differentiation. If a local interaction between medulla and cortex is the basic mechanism in the differ- entiation of the gonad,t it is also well known that a gonad is able to influence the differentiation of another at a distance (and to control the development of accessory sex structures as well). From this fact arose the theory of hormones as differentiating agents, which has inspired most of the modern work on sex differentiation. The hormone theory of sex differentiation was first developed with special reference to the freemartin problem by Lillie (16, 17) and Keller and Tandler (16). The theory has been tested chiefly by two methods: grafting, during the differen- + The view that the interactions between medulla and cortex are mediated by special inductor sub- stances, of a different chemical order and with a different mode of action from hormones, will be considered later, 476 tiation period, of gonads or gonad tissue in various ways (including the union of whole embryos—parabiosis); and treatment of the embryo with hormones. For technical reasons grafting methods have been largely confined to embryos of amphibians and birds. Grafting Experiments. In amphibians the results of grafting the gonad primordium and of parabiosis (Fig. 181) are essentially the same. In combinations of the same _ sex, gonad differentiation is normal, but in un- like combinations one sex undergoes inhibi- tion or reversal. Typically the male gonad is dominant, but with marked disparity in size or rate of development (as in certain heteroplastic combinations) the ovary may SPECIAL VERTEBRATE ORGANOGENESIS hormones in avian development. Gonads, or gonad tissue, grafted to the chorio-allantois failed to modify the sex structures of host embryos, or to be themselves modified (for reviews see Willier, ’39, 52). It was eventu- ally shown, however, that grafts to the em- bryo itself are effective (Wolff, ’46). Ovarian grafts induce cortical differentiation on the embryonic testes of the host, but host ovaries are not modified by testis grafts. Evidently in birds the ovary is the dominant gonad. Administration of Hormones. Reversal of differentiation by treatment with pure hor- mones follows a similar pattern. Amphibian larvae treated during the differentiation period develop varying degrees of intersexu- Fig. 182. Sections, in anterior-posterior sequence, through the testis of a salamander, joined in parabiosis with a much larger female of another species. Various stages are shown in the conversion of a testis to an ovary by degeneration of the medulla and development of the cortex. (From Burns, ’35.) predominate. Histologically, reversal involves inhibition and gradual involution of the genetically dominant component, accom- panied by differentiation of the recessive (Fig. 182). Reversal may end in complete functional transformation (see Humphrey, 44, °45, °48), or incomplete dominance may lead to prolonged intersexuality. Removal of the graft causing reversal may then permit reversion to the original sex (for reviews see Willier, ’39; Witschi, °39; Humphrey, °42). The results depend in detail on various ex- perimental conditions: the relative size and rate of development of the gonads (or grafts), and the time relations of the experiment. In addition to such variables, pronounced spe- cies and race differences are known (see Witschi, ’34, °39) which are reflected in the composition of the gonads, with respect to the representation of medulla and cortex, or the intensity and timing of humoral activity. These differences evidently have a constitutional basis. In birds the problem has been studied mainly in the chick embryo. For a long time there was serious doubt as to the role of ality, according to experimental conditions and species. Two methods have been used: direct injection during the differentiation period, and immersing the subjects in aque- ous solutions of the hormone. As in grafting experiments, the histological picture in trans- forming gonads shows involution of one gonad component and gradual emergence of the other, according to the sex type of the hormone. Early treatment frequently in- duces complete transformation in either sex (for reviews see Gallien, °44, ’50). In some cases, however, large doses may have exactly opposite effects; for example, a female hor- mone may completely masculinize female gonads (see Padoa, 36, 738, ’42; Gallien, 41, 44). Such “paradoxical effects” do not occur, however, at low dosages; and it has been shown further that the same hormone, under the same conditions, may have a feminizing effect at low dosages, produce intersexes at intermediate levels, and have only a mas- culinizing effect at high dosages (e.g., Padoa. 38, °42; Witschi, *51b). Other examples of paradoxical effects will be found and their significance will be considered later, URINOGENITAL SYSTEM In bird embryos, hormones introduced into the egg produce typical reversal effects on the gonads (see Wolff and Ginglinger, ’35; Willier, Gallagher, and Koch, ’35, ’37). Fe- male hormone transforms embryonic testes into ovotestes or ovaries, by reduction of the medullary zone and differentiation of cortex (Fig. 183). Male hormones transform ovaries by repression of the cortex with medullary hypertrophy. The relative value of constitu- tional vs. hormonal factors is illustrated clearly in consequence of the interesting 477 have not been regarded as significant. An ex- ception has recently been found in the effects of estrogen on the testes of new-born opossum embryos (Burns, 50). Extensive persistence of the germinal epithelium was found, and in a few cases typical ovotestes were obtained, except that the induced cortex was sterile (Fig. 184). Histological changes followed the usual sequence: repression of testicular development, followed by reactivation of the germinal epithelium to produce a cortex. This result suggests that transformation of Fig. 183. Ovotestis developed from the left gonad of a genetic male chick by treatment with female hor- mone. The testicular part of the gonad occupies a central or hilar position; the highly developed cortex is peripheral. (From Willer, ’39.) lateral differences in organization in birds. The left testis (with its incipient cortical element) reacts at much lower dosages than the right, and with increasing doses is more completely transformed (see Willer, °39). Notwithstanding the fact that the gonads of freemartins are often greatly modified (Willier, ’21), exposure of mammalian em- bryos to relatively large doses of hormones* has usually had but slight effect on the gonads, although accessory genital struc- tures may be completely transformed. In some cases male hormones produce a limited hypertrophy of the medullary cords of the ovary (Jost, 47a; Wells and Van Wagenen, 54); and persistense of localized areas of ger- minal epithelium on the testis has been re- ported after the use of either type of hormone (e.g., Raynaud, *42; Jost, 47a). Such changes * Administered to the mother during pregnancy or, in the case of marsupial embryos, directly to the pouch young. mammalian gonads may prove feasible if proper experimental conditions can be real- ized. There is substantial evidence that the gonads are potentially bisexual. In addition to the freemartin, the reversals described in rat gonads (p. 474) must be noted, and the anomalous occurrence of /hermaphroditis- mus verus, characterized by well differen- tiated ovotestes, is well known in many mam- mals. DIFFERENTIATION OF THE ACCESSORY GENITAL STRUCTURES On the basis of embryological origin the accessory genital structures of vertebrates comprise three main groups: (1) the sex ducts and associated structures, of nephric origin; (2) derivatives of the cloaca or the urinogenital sinus; and (3) copulatory struc- tures, which develop from the genital tuber- cle. For a detailed account of the origin and 478 early development of these structures see Willier (’39). The differentiation of the accessory sex structures has been studied experimentally in various ways. As in the case of the gonads, different forms of grafting were first em- largely ployed. Eventually grafting was SPECIAL VERTEBRATE ORGANOGENESIS throughout life in the males of many species as complete if somewhat rudimentary struc- tures. An ideal basis for sex reversal is thus provided. In male urodeles elaborate cloacal glands develop, which in females are absent, rudimentary, or (in some species) differently specialized. The dimorphism of these struc- Fig. 184. Stages in the development of ovotestes in opossum embryos after treatment from birth with female hormone. A, Persistence of a thick germinal epithelium on a greatly retarded testis of 14 days (x 500). B, Ovotestis of a male aged 30 days (x 150), in which the medullary zone is separated from a sterile cortex by a thick, fibrous tunica albuginea. (From Burns, ’50.) superseded by administration of pure hor- mones. The development in recent years of methods of castrating embryos has provided crucial evidence, and tests of the self- differentiating capacities of sex primordia in physiological isolation have lately been utilized. Results of Grafting Methods. Grafting tech- niques have been mainly confined to amphib- ians and birds. The structures to be con- sidered in amphibians are the sex ducts and the glands of the cloaca. In many amphibians both gonaducts are retained indefinitely; the Wolffian ducts function as excretory ducts in both sexes, and Miillerian ducts persist tures is controlled by the gonads. After cas- tration in larval life both ducts remain in a sexually indifferent condition, and cloacal differentiation does not occur (de Beaumont, 33). Gonads grafted into castrates induce differentiation of the appropriate sex duct, and testis grafts cause development of the cloacal glands. When the gonads are trans- formed experimentally (p. 476) the sub- sequent differentiation of gonaducts and cloaca conforms (see Humphrey, ’42). In bird embryos grafting methods long gave negative results, and the role of hor- mones in sex differentiation was in doubt (see p. 476). It now appears that these fail- URINOGENITAL SYSTEM ures were due merely to a quantitative in- sufficiency of hormone, since gonads grafted directly into the embryo are effective. Grafted ovaries induce partial retention of Millerian ducts in male embryos, while testes stimulate Wolffian ducts and completely inhibit Miil- lerian ducts in both sexes (Wolff, 46). Sim- ilar results have recently been obtained by multiple testis grafts on the chorio-allantois (Huijbers, °51). 479 consult Colloques Internationaux, Centre Na- tional de la Recherche Scientifique (Paris, 1951); La Différenciation Sexuelle chez les Vertébrés. | It is not possible to consider all findings in detail, but as a basis for discussion the prin- cipal results may be summarized as follows: 1. Sex hormones have sex-specific effects on the development of genital structures, either by direct action on the individual primordia, Fig. 185. Development of the sex ducts in ee embryos after treatment with sex hormones (from Willier, 39). A, Normal female embryo incubated 18 days; B, male embryo treated with female hormone—both oviducts are present and greatly hypertrophied; C, normal male embryo of 17 days showing complete ab- sence of oviducts; D, female embryo treated with male hormone—the oviducts are absent except for small fragments anteriorly, and the Wolffian ducts are greatly hypertrophied. Gonad grafting has only recently been utilized in mammalian embryos, chiefly with a view of counteracting the effects of castra- tion. The results are positive, and will be mentioned in that connection. Administration of Hormones. The production of steroid hormones in pure form enor- mously accelerated the experimental study of sex differentiation. For the first time it became possible to attack the problem in mammals by administration of effective doses during pregnancy. The different experi- mental conditions under which hormones have been used, in many different groups and species, have inevitably led to variable and sometimes inconsistent results. [For re- views and analyses of a large literature see Wolff (38, °47); Willier (39); Humphrey (42); Raynaud (42); Mintz (47); Moore (47); Price (47); Jost (48); Ponse (’49); Burns (749). For a recent synopsis of the field or by acting indirectly, as in release from inhibition, or via other endocrine agencies. 2. In general, male hormones accelerate differentiation of male structures and inhibit certain female structures, in embryos of both sexes; they also induce certain male pri- mordia in females. Conversely, female hor- mones stimulate the development of female structures and inhibit certain male _pri- mordia. 3. In addition to their usual effects, how- ever, both types of hormone in many cases paradoxically stimulate structures of the other sex. If the effects of hormones as outlined re are in general consistent with theory, the occurrence of so-called “paradoxical effects” is discordant, and has been differently inter- preted. It is denied by some that the effects of steroid hormones are in any sense spe- cific, and the validity of the hormone theory 480 in its entirety has been questioned (Moore, "44, °47). It is known, however, that some of the most obvious paradoxical effects either are exerted indirectly, or depend on spe- cial experimental conditions. This subject is better discussed after some of the evi- dence has been presented. The Effects of Hormones on the Sex Ducts. Although the early differentiation of the gonaducts is independent of sex, they are SPECIAL VERTEBRATE ORGANOGENESIS regression (such as normally occurs in fe- males) is initiated (eg., Raynaud, 42; Greene, °42). Paradoxically, it sometimes causes retention of the ducts in females (Greene). Male hormone induces retention and pre- cocious hypertrophy of the Wolffian ducts in both sexes. In mammalian embryos the epi- didymis also hypertrophies, and epididy- mides and seminal vesicles develop in fe- GONAD — OVARY VAS DEFERENS MESONEPHROS = > 4a) NECK OF WOLFFIAN DUCT BLADDER | | MULLERIAN EPIDIDYMIS DUCT ~ | Ry SS ZZ, F ony ASS UG NECK OF Ser iC }+—TESTIS steissatelaiy VAGINAL CANAL SEAS ale URINOGENITAL : lke . As SiN URETHRA CAURD PROSTATE PENS “*—— BULBAR GLAND PHALLUS URETHRA POUCH YOUNG — + IODAYS URINOGENITAL SIN US BULBAR GLAND ma Wie HCALA CULL FEMALE & MALE-+35DAYS Fig. 186. The normal differentiation of the genital tracts in pouch young of the opossum (from Burns, 49). A, Sexually undifferentiated embryo of 10 days, showing male and female sex ducts, urinogenital sinus and phallus; B, sexually differentiated genital tracts in young at 35 days, female characterized by vaginal canals and absence of prostate, male showing absence of vaginal structures and numerous prostatic glands. capable of responding early to adequate doses of sex hormones. Female hormone in- duces hypertrophy of Miillerian ducts in fe- male embryos, and retention with hyper- trophy in males. The effects in bird embryos are particularly striking (Fig. 185). In males both oviducts persist and hypertrophy, as does also the right duct of females. However, the period of reactivity is limited; there is a “critical period” for effective action. Re- tention and development of the ducts in male embryos is secured by injections of female hormone up to the tenth day (“stabilization effect”); later treatment is useless (Wolff, 38). Female hormone typically has ne ef- fect on Wolffian ducts, but in some mammals males (e.g., Greene, 42; Burns, *42; Ray- naud, ’42; Wells and van Wagenen, 54). The effects of male hormones on the Miillerian ducts are more complex, depending especially on timing and dosage, and on species differ- ences. In larval amphibians and in chick embryos, treatment during the formative period largely or entirely suppresses the ducts (Fig. 185). [See, eg., Burns (’39), Foote (’41), Hanaoka (741) for amphibians; Wolff (38, °50), Gaarenstroom (739), Stoll (48), Huijbers (751) for the chick.] This effect is also produced by grafts of the em- bryonic testis (Wolff, ’46; Huijbers, ’51). For complete suppression the hormone must act early—in chick embryos before the sixth URINOGENITAL SYSTEM or seventh day; later treatment has no effect (Stoll, ’48; cf. the stabilizing effect of female hormone on male Miillerian ducts). Again there is a critical period after which suscep- tibility is lost. In larval amphibians the Miillerian ducts develop slowly. Treatment with male hor- mone during the backward growth of the duct suppresses the unformed portion, but 481 ferentiation of the glands in larvae of either sex (see Burns, ’39; Mintz, ’47) and a male type of cloaca. The urinogenital sinus of mammals is de- rived embryologically from the cloaca. In its primitive condition it receives the sex ducts near the neck of the bladder, and opens externally at the base of the genital tubercle. Differentiation in females is characterized by Fig. 187. Development of the sinus region and prostate in young opossums aged 50 days, after treatment from birth with male hormone (from Burns, ’49). A, Section of normal male sinus showing prostatic glands; B, male receiving hormone, showing enormous hypertrophy of the prostate; C, female showing great de- velopment of the prostate induced by male hormone; D, normal female, characterized by vaginal canals and complete absence of prostatic glands; V, terminus of vaginal canal. the part already present persists, and with large doses paradoxically hypertrophies (see Mintz, °47). In mammalian embryos com- plete suppression of the Miillerian ducts is not obtained, but the vaginal region may be inhibited, e.g., in opossums (Burns, °42) and in mice (Raynaud, ’42). On the contrary, paradoxical stimulation occurs readily (e.g., Moore, *47; Burns, *49), but is not found with low dosages (Burns, ’42, ’45a). Reactions of Cloaca and Urinogenital Sinus to Hormones. The development of cloacal glands in amphibians is conditioned by male hormone; the response of the cloaca to castra- tion and to testis grafts has been noted (p. 478). Male hormones induce precocious dif- formation of the vagina, while the male ducts retrogress. In the male, regression of the Miillerian ducts and development of the complex of prostatic glands are the chief features (Fig. 186). Male hormones produce a male type of sinus in both sexes (for de- tails see Greene, °42; Raynaud, 42, °50; Moore, ’47; Burns, ’49; Jost, °50; Wells and van Wagenen, °54). The differentiation of males is accelerated; females undergo trans- formation—vaginal development is partly or entirely suppressed and prostatic glands de- velop (Fig. 187). Conversely, fermale hor- mone suppresses prostatic differentiation and produces a sinus of female form, with a vagi- nal type of epithelium. Permanent suppres- 482 sion of the prostate may be induced by a single dose of hormone administered just be- fore the prostatic buds should appear (Burns, 42); and prostatic glands induced in female embryos continue to develop without further treatment (see Moore, ’47). Thus, bud forma- tion represents a critical stage in which de- velopment or permanent suppression of the prostate is determined by presence of the proper hormone. Reactions of the Copulatory Structures to Hormones. In both birds and mammals, hor- SPECIAL VERTEBRATE ORGANOGENESIS ferentiation is provided by castration of the embryo. The effects of castration in larval amphibians have been mentioned (p. 478), and successful techniques have recently been developed for birds and mammals (Fig. 189). To be decisive the operation must be per- formed early. After early castration the Miil- lerian ducts of male embryos, instead of retrogressing, continue to develop (as does also the abortive right oviduct in female chicks). Presumably in normal development regression is conditioned by the testis—in Scrotum Fig. 188. The effects of male and female sex hormones on the development of the external genital struc- tures in young opossums, treated from birth to an age of 20 days (from Burns, 49). A, Precocious develop- ment and hypertrophy of the phallus (male type) in a male embryo receiving male hormone; B, genital structures of female type produced in another male embryo by female hormone. When both embryos are female the effects are the same. mones control development of the copula- tory structures, producing a complete change in morphology in many species, e.g., the duck (Wolff and Wolff, ’48), the opossum, (Moore, °41; Burns, 42, °45b), and various placental mammals (Greene, *42; Raynaud, ’42; Jost, 47a; Wells and van Wagenen, ’54). The embryological basis for transformation has been studied in the opossum (Burns, ’45b). It consists in specific responses of the histological constituents of the embryonic phallus to a given hormone. The erectile tissues, which largely determine the form of the penis, are stimulated by male and virtually suppressed by female hormone, which in turn produces marked hyperplasia of the periurethral and vulvar connective tis- sues (Fig. 188). The Effects of Castration or Isolation on the Development of Sex Primordia. The most direct test of the role of hormones in sex dif- the case of the right oviduct of the chick by the ovaries (Wolff and Wolff, 51). Is de- velopment of Miillerian ducts after castra- tion due solely to release from inhibition, or does some positive humoral factor intervene? [In various species of mammals estrogens are known to be present in the placenta and fetal fluids. For a review see Price (’47).] This question has been approached by test- ing the developmental capacity of the ducts in isolation. After explantation of the em- bryonic genital tracts in vitro (Jost and Bergerard, 49; Wolff, 50; Jost and Bosic, 51), or transplantation to the chorio-al- lantois (Wolff, 50), the Miillerian ducts of both sexes persist and continue to differ- entiate. When explanted after the eighth day of incubation, however, those of male chicks degenerate (Wolff). At this stage regression has been finally determined. The evidence indicates that in the absence of testes the URINOGENITAL SYSTEM Miillerian ducts differentiate autonomously. It should be recalled in this connection that the regression of the Miillerian ducts in male embryos is prevented by female hor- mone (p. 480). Their retention in such case might be due initially to inhibition of the testes, but when there is also marked hyper- trophy of the ducts a direct stimulation must be involved. The effects of castration on the male sex ducts vary in different forms (Fig. 189). In larval amphibians and in chick embryos the Wolffian ducts, which serve as nephric ducts, remain after castration in a sexually undif- ferentiated condition (de Beaumont, 733; 483 tance of the time factor. Castration on the twenty-third or twenty-fourth day is fol- lowed by almost normal development; if per- formed somewhat earlier the prostatic buds are arrested and differentiate no further. Castration before the twentieth day sup- presses all development—the buds do not appear. The effects of castration are entirely prevented by male hormone or by grafts of the embryonic testis (Jost, ’50; Wells, ’50). The copulatory structures are also highly sensitive to castration. Castrated mammalian embryos of both sexes develop external geni- talia of female type (see Jost, °47b, 750; Raynaud and Frilley, ’47; Raynaud, *50), but CASTRATION EFFECTS IN EMBRYOS MAMMALS Absent o& Castrates | Developed | | Mtlerian Ducts | Wolffian Ducts | UG.Sinus | Prostate} Ext. Genitalia 9 Castrates | Developed Absent * Absent Absent Mammary Glands ? type ? type BIRDS er ete Millerian Ducts ¢o Castrates | Developed 9 Castrates Wolffian Ducts Present Developed Present (on both sides) Gen tubercle g type ¢ type 3 type do type Fig. 189. Table summarizing the effects of castration in embryos of birds and mammals, based on the reports of Wolff (750) and Wolff and Wolff (51) in the chick and duck; Huijbers (51) in the chick; Jost (47b, °50) in the rabbit; Raynaud and Frilley (’47) and Raynaud (’50) in the mouse. Similar but less severe effects have been obtained in the rat (Wells, 50; Wells and Fralick, 51); but Moore (’47) reports little effect in young opossums castrated at the age of 22 days. The asterisk indicates that partial persistence of Wolffian ducts may occur in castrate female mice (see Raynaud, ’50). Wolff, ’50). In mammals their retention and sexual differentiation in the male appear to depend upon the testis; they regress in cas- trate rabbit fetuses, as in the normal female (Jost, 47b, *50). [There is an indication in mice that the Wolffian ducts may persist in part in castrate females, suggesting that the ovary may have a role in the regression of the ducts in normal females (see Raynaud, 60) =] Castration prevents development of the male urinogenital sinus and accessory glands, but female castrates are virtually normal (Fig. 189). In males, development of the vagina (in conjunction with retention of the Miillerian ducts) gives the sinus a female form. Prostatic differentiation is suppressed (Jost, °47b), and in mice coagulating glands are also absent (Raynaud, °50). The effects in the rat embryo are less severe, depending perhaps on time of operation (Wells and Fralick, *51). With respect to the prostate, the results in the rabbit are especially clear (Jost, °47c), illustrating the great impor- the effects of castration are prevented by male hormones. As in the case of the duct systems and the sinus structures, normal male development evidently depends on the testis, while the female pattern develops without hormonal conditioning, in a_ so- matic or asexual manner. In bird embryos, however, the situation is reversed, the male sex representing the asexual type (Wolff and Wolff, *51). SOME SPECIAL PROBLEMS AND CONSIDERATIONS Patterns of Hormonal Control. The hormone theory of sex differentiation was developed by Lillie and his associates with special ref- erence to the case of the freemartin. The dominance of the male twin was explained on the grounds that the testis produces its hormone before the ovary; in fact, no sug- gestion of endocrine activity in the ovary was seen until late in fetal life. The domi- nant role of the male hormone was thus 484 emphasized, without positive commitment as to the role of the ovary in female de- velopment. (The suggestion was made that in cases of early transfusion the first effect of the male hormone is to inhibit the ovarian cortex, in effect eliminating the fetal ovary as a factor in later development.) The re- sults of castration in mammalian embryos fit into the above picture in a remarkable way. Absence of the testis and its hormone arrests development of all male characters— TESTIS FEMALE HORMONE FEMALE SEX CHARACTER Fig. 190. Diagram illustrating possible modes of hormone action on a sex character. Development of a female structure may result from inhibition of the male component of the gonad (A), thus removing an inhibitory influence, as in castration; or after inhibition of the medulla the structure may respond to the differentiating cortex (B). Finally, it may be stimulated directly asim C. A male character may be adversely affected by any of the above modes of action. Whenever the response of a structure ex- ceeds its normal rate of development, direct stimu- lation, as in C, is indicated. ducts, sinus derivatives and external genitalia are uniformly affected (see Fig. 189)—while a full complement of female characters ap- pears in castrates of both sexes. It is clear, then, that the male hormone is necessary for development of the male system, and also for inhibition of female development. On the other hand, no essential role for a female hormone is indicated, although the capacity of female primordia to respond to sufficient doses of female hormone is well established (pp. 480-482). The situation con- forms closely with Wiesner’s ““monhormonic theory,” based on the results of castration and hormone treatments in early postnatal life (Wiesner, °34, °35). In amphibians and birds, on the contrary, both gonads participate in the differentiation process. In bird embryos, Miillerian ducts persist and develop after castration in both sexes, as in Mammals, regression in the male depending on the testis. But differentiation of the genital tubercle (and syrinx) is con- SPECIAL VERTEBRATE ORGANOGENESIS ditioned by the ovary, castration producing the male type in both sexes. Administration of hormones produces corresponding effects. Male hormone inhibits Miillerian ducts and produces a genital tubercle of male type, overruling the ovary; female hormone stabi- lizes the Miillerian ducts in the males and inhibits the tubercle. In larval amphibians, of either sex, both gonaducts are retained after castration in a sexually undifferenti- ated condition. Grafting of a gonad then induces development of the proper duct and other sexual characters. Sex-reversed gonads have the same effects. Thus both hormones participate in differentiation. It is difficult to escape the impression that the mam- malian pattern, in which the male hormone is all important, has evolved in adaptation to the conditions of intra-uterine develop- ment. Mode of Action of Hormones. In the inter- pretation of hormone action there is always the problem as to how the responses of embryonic primordia are elicited—whether by direct action of a stimulatory or inhibi- tory nature, or indirectly, by depression or elimination of the normal control mecha- nism. The ability of various primordia to develop autonomously after castration indi- cates the importance of the latter alternative. The situation in an intact embryo is com- plex. It is possible for an administered hor- mone to act directly on the individual parts of the genital apparatus, or indirectly, by way of the gonad. Direct action on one gonad component may indirectly influence the other, and through it the accessory struc- tures (Fig. 190). In modified gonads _his- tological appearances suggest that the ge- netically dominant component is directly inhibited by a heterotypic hormone, allow- ing development of the recessive component (p. 477, see also Willier, ’39). This inter- pretation is supported by experiments (p. 475) in which removal or depression of the dominant component is followed by recovery and development of the recessive. It is pos- sible, however, that in some cases a hormone may directly stimulate the homotypic gonad component (Vannini, 46). In interpreting the actions of hormones on accessory sex structures, the status of the individual primordia must be considered. The results of castration show that some (e.g., sex ducts, female sinus derivatives and genitalia) are capable of a high degree of autonomous differentiation; others (e.g., the prostate) are entirely conditioned by hor- mones. Structures of the first type may de- URINOGENITAL SYSTEM velop as a result of indirect action on the gonad, depressing or abolishing an in- hibitory mechanism. When, however (as frequently happens under experimental con- ditions), the parts in question are differ- entiated precociously, or show marked hyper- trophy in comparison with normal struc- tures, direct stimulation by the introduced hormone is indicated (Fig. 190). Direct ac- tion is also indicated when castration effects are prevented by hormones (p. 483). The differentiation of structures of the second class probably always involves direct action. Specificity of Hormone Action. Evidence has been presented showing that, on the whole, the actions of hormones are specific in the sense that a given type of hormone has a stimulatory action on structures of the proper sex, while inhibiting various hetero- typic structures. Specificity of action has been denied, however, largely on the basis of paradoxical effects, various examples of which have been cited. [For a summary of the paradoxical effects of various steroid hormones see Wolff, Strudel and Wolff (48).] To be valid this argument must as- sume that such effects are the result of direct action of the hormone on heterotypic sex primordia. But in some of the best known cases of paradoxical effects it is probable that the action is mediated indirectly, and in many instances excessive dosage is the determining factor. It must be recalled that under certain conditions gonads also exert paradoxical effects in which the action in- volves production of the hormone of the other sex (for examples see Ponse, °48; Hill, ’50). The paradoxical effects of high dosages are well known. For example, in opossum em- bryos receiving large doses of male hormone extreme stimulation of male structures is also accompanied by hypertrophy of Miil- lerian duct derivatives in both sexes. The latter effect declines rapidly with dosage, and below a certain level disappears com- pletely, although male structures are still stimulated (Burns, °42, 45a). A similar situation appears in many experiments with amphibians. [For paradoxical effects on the gonads see Padoa (736, ’38, °42); Gallien C41, °44); Vannini (746); and for the Miil- lerian ducts, Mintz (’47).] Large doses of female hormone have a strong masculinizing effect on gonads, but low doses are feminiz- ing only. Indeed, the same substance (es- tradiol), in the same dosage, may have opposite effects depending on the solvent. In aqueous solution there is a masculinizing effect and in oil an orthodox action (Gallien, 485 41). The rate of utilization—in effect a difference in dosage—is evidently the deter- mining factor. The question remains as to how the para- doxical effects of large dosages are exerted. If, for example, female primordia can re- spond directly to a sufficient concentration of male hormone, only a difference in threshold between male and female _pri- mordia is involved, and to this extent the hormone is “ambisexual.” But when the dose required to elicit a paradoxical response so far exceeds physiological limits, the result has little significance for normal differentia- tion. There is, however, much evidence sug- gesting that paradoxical effects are exerted indirectly. Masculinization of gonads by large doses of female hormone is known in some cases to be associated with strong hy- perplasia of the interrenal (adrenal cortical) tissue (e.g., Padoa, ’38, °42; Witschi, ’51b), a potential source of androgen; and in vari- ous mammals (see Burrows, ’45) treatment with large doses of male hormone is followed by excretion of estrogens, in both intact and castrate individuals. This phenomenon has not as yet been demonstrated in embryos but the possibility must always be consid- ered. On the whole, the evidence provided by paradoxical effects seems at present in- conclusive and greatly outweighed by the many unqualified examples of specific action. Embryonic versus Adult Hormones. The re- sults of parabiosis and various types of gonad grafting in amphibian and bird embryos demonstrate that embryonic gonads produce sex-differentiating substances which are transported and act in the manner of hor- mones (cf. also the freemartin). The results of embryonic castration confirm this view in the strongest possible manner. A question as to the nature of embryonic hormones, and their relation to adult sex hormones, is nat- urally posed. The appearance of paradoxical effects, and the fact that steroid hormones do not in all cases produce complete and inte- grated transformations, have led to the view that embryonic hormones are chemically and physiologically different. But when the re- sponses of sex primordia to steroid hormones are compared with those produced by em- bryonic gonads in various types of grafting experiments, no essential differences appear. There is in most cases a high degree of specificity in the responses to steroid hor- mones, and the mechanism of action (stim- ulation-inhibition) seems identical for many structures. The significance of paradoxical effects has been discussed. 486 It has been shown further that crystalline sex hormones prevent or repair the effects of embryonic castration (e.g., Jost, 50; Wells, 50) and, conversely, that embryonic gonads are able to repair castration effects in the adult (Jost and Colonge, °49). Moreover, substances have been obtained from em- bryonic or fetal testes which modify various adult male sex characters (e.g., Leroy, *48; see also Jost, 48). In view of the many other variables which enter into experimental re- sults there seems no necessity at present for assigning embryonic hormones to a different category. There is also no necessity or justi- fication for assuming that they are identical molecularly with any particular steroid hor- mones. Embryonic Hormones and Inductor Sub- stances. A similar problem exists with respect to embryonic hormones and the inductor substances (corticin, medullarin) which, as mediators of the interactions between cortex and medulla, have been postulated to control differentiation of the gonads (see, e.g., Witschi, ’34, ’39, °50). The inductor sub- stances may be characterized as follows: (1) as the primary effectors in sex differentia- tion, they exert their effects early, before embryonic hormones are present; (2) their action is exerted locally, ordinarily within the confines of the gonad; and (3) transport is by diffusion through the tissues rather than by the blood. Thus inductors are hu- moral in nature but not hormones. It is sug- gested further that the great variability in the reactions of gonads under experimental conditions, particularly as regards the many race and species differences, presupposes such specificity in organization as to indicate that the inductor substances are probably protein in nature. The theory of inductor substances has be- come difficult to maintain in the face of growing evidence that hormones are capable of producing most of the effects ascribed to the inductors. The mechanism of action in the gonad is apparently the same. When gonads undergo transformations (whether by administered hormones or by gonads acting from a distance, as in parabiosis) the his- tological picture shows a change in the ratio of cortex to medulla. This change is appar- ently produced by inhibition of the heter- otypic gonad component—that is to say, the effect of the hormone is to redirect the mech- anism of normal differentiation. Crystalline hormones, male or female, are also capable of controlling the primary dif- ferentiation of gonads, so that all embryos SPECIAL VERTEBRATE ORGANOGENESIS from the beginning develop as one sex (see Padoa, °38, °42; Gallien, “44, °50). The dosages required may be minute (see Mintz, °48). If inductors are present their actions are overruled, the hormone assuming the role of the genetically recessive inductor. It may be urged that hormones work by controlling the inductor systems; but, in the absence of di- rect evidence of the existence of inductor substances, it seems unnecessary to assume two agencies. (For fuller discussions see Willier, 39; Wolff, °47; Jost, ’48; Ponse, ’49; Burns, °49; Gallien, 50.) The actual evidence seems insufficient to prove that the postu- lated differences between inductor and hor- mone effects are more than quantitative, depending on such factors as intensity (con- centration) and timing, in conjunction with species differences in the sensitivity of re- actor systems. Under proper conditions the local action of inductors is closely imitated by hormones. Only a few examples may be cited. In am- phibians (under normal or experimental con- ditions) the first responses of the gonaducts to hormones often take place in close prox- imity to the gonads (normal or grafted); and in chick embryos, gonads and sex ducts both show highly localized reactions to an adjacent graft (Wolff, ’46). Finally, the ex- tremely localized action of an embryonic testis in repairing castration atrophy in the adult seminal vesicle (Jost and Colonge, ’49) is a case in point. REFERENCES Auer, J. 1947 Bilateral renal agenesia. Anat. Rec., 97:283-292. Bounoure, L. 1934 Recherches sur la lignée germinale chez la Grenouille rousse aux premiers stades du développement. Ann. Sc. Nat. Zool., 17: 69-248. 1937 La constitution des glandes géni- tales chez la Grenouille rousse aprés destruction étendue de la lignée germinale par l’action des rayons ultra-violets sur loeuf. Compt. Rend. Acad. Sc., 204:1957-1959. 1950 Sur le développement sexuel des glandes génitales de la Grenouille en l’absence de gonocytes. Arch. d’Anat. micr. et de Morph. exp., 39:247-254. Boyden, E. A. 1924 An experimental study of the development of the avian cloaca with special reference to a mechanical factor in the growth of the allantois. J. Exp. Zool., 40:437-471. 1927 Experimental obstruction of the mesonephric ducts. Proc. Soc. Exp. Biol. & Med., 24:572-576. 1932 Congenital absence of the kidney: an interpretation based on a 10-mm. human em- URINOGENITAL SYSTEM bryo exhibiting unilateral renal agenesis. Anat. Rec., 52:325-339. Brachet, A. 1935 Traité d’Embryologie des Ver- tébrés. 2d ed. Masson et Cie., Paris. Brambell, F. W. R. 1927 The development and morphology of the gonads of the mouse. Part II. The development of the Wolffian body and ducts. Proc. Roy. Soc., Series B., 702:206-221. Burns, R.K. 1935 The process of sex transforma- tion in parabiotic Amblystoma. III. Conversion of testis to ovary in heteroplastic pairs of A. tigri- num and A. punctatum. Anat. Rec., 63:101-129. 1938 Development of the mesonephros in Amblystoma after early extirpation of the duct. Proc. Soc. Exp. Biol. & Med., 39:111-113. 1939 The effects of crystalline sex hor- mones on sex differentiation in Amblystoma. II. Testosterone propionate. Anat. Rec., 73:73-93. 1941 The origin of the rete apparatus in the opossum. Science, 94:142-144. 1942 Hormones and experimental mod- ification of sex in the opossum. Biological Sym- posia, 9:125-146. 1945a_ Bisexual differentiation of the sex ducts in opossums as a result of treatment with androgen. J. Exp. Zool., 700:119-140. 1945b The differentiation of the phallus in the opossum and its reactions to sex hormones. Contr. Embryology, Carnegie Inst. Washington, 31:147-162. 1949 Hormones and the differentiation of sex. Survey of Biological Progress, vol. I. Academic Press, Inc., New York. 1950 Sex transformation in the opossum: some new results and a retrospect. Arch. d’Anat. micr. et de Morph. exp., 39:467—481. Burrows, H. 1945 Biological Actions of Sex Hor- mones. Parts III and IV. Cambridge University Press, Cambridge, England. Buyse, A. 1935 The differentiation of trans- planted mammalian gonad primordia. J. Exp. Zool., 70:1—-41. Cambar, R. 1948 Recherches expérimentales sur les facteurs de la morphogénése du mésonéphros chez les amphibiens anoures. Bull. Biol. France et Belgique, 82:214-285. 1949 Données récentes sur le développe- ment du systeme pronéphrétique chez les Am- phibiens (Anoures en particulier). Ann. Biol., 25: 115-130. Colloques Internationaux du Centre National de la Recherche Scientifique. XXXI: La Différencia- tion sexuelle chez les Vertébrés. Paris, 1950. Dalcq, A. 1942 Contribution a ]’étude du poten- tial morphogénétique chez les Anoures. III. Opérations visant |’ébauche pronéphrétique au seuil de la gastrulation. Arch. Biol., 53:1-124. Dantschakoff, W. 1932 Keimzelle und Gonade. Die entodermale Wanderzelle als Stammzelle in der Keimbahn. Experimentelle Beweise. Zeit. Zellforsch. Mikr. Anat., 74:376-384. de Beaumont, J. 1933 La différenciation sexu- elle dans l'appareil uro-génital du Triton et son déterminisme. Roux’ Arch. Entw.-mech., 129: 120-178. Domm, L. 1939 Modifications in sex and second- 487 ary sexual characters in birds; in Sex and In- ternal Secretions. 2d ed. Williams & Wilkins Co., Baltimore. Dulbecco, R. 1946 Sviluppo di gonadi in assenza di cellule sessuali nell’ embrione di pollo. Steriliz- zazione completa mediante esposizione a raggi y allo stadio di linea primitiva. Rend. Acc. Naz. Lincei, ser. VIII, 7:1211-1213. Fales, D. E. 1935 Experiments on the develop- ment of the pronephros of Amblystoma puncta- tum. J. Exp. Zool., 72:147-173. Field, H. H. 1891 The development of the pro- nephros and segmental duct in amphibia. Bull. Mus. Comp. Zool., Harvard College, 27:201-340. Foote, C. L. 1941 Modification of sex develop- ment in the marbled salamander by administra- tion of synthetic sex hormones. J. Exp. Zool., 86: 291-319. Fraser, E. A. 1950 The development of the ver- tebrate excretory system. Biol. Rev., 25:159-187. Gaarenstroom, J. H. 1939 Action of sex hor- mones on the development of the Miillerian duct of the chick embryo. 1. Exp. Zool., 82:31-46. Gallien, L. 1941 Recherches expérimentales sur Vaction amphisexuelle de lhormone femelle (oestradiol), dans la différenciation du sexe chez Rana temporaria 1. Bull. Biol. France et Bel- gique, 75:369-397. 1944 Recherches expérimentales sur Vorganogenése sexuelle chez les Batraciens Anoures. Bull. Biol. France et Belgique, 75:257- 359. 1950 Les hormones sexuelles dans la dif- férenciation du sexe chez les Amphibiens. Arch. d’Anat. micr. et de Morph. exp., 39:337-360. Gillman, J. 1948 The development of the gonads in man with a consideration of the role of fetal endocrines and the histogenesis of ovarian tu- mors. Contr. Embryology, Carnegie Inst. Wash- ington, 32:81-131. Goodrich, E. S. 1930 Studies on the Structure and Development of Vertebrates. The Macmillan Co., London. Gray, P. 1932 The development of the amphib- ian kidney. II. The development of the kidney of Triton vulgaris and a comparison of this form with Rana temporaria. Quart. J. Micr. Sci., 75: 425-465. 1936 The development of the amphib- ian kidney. III. The post-metamorphic develop- ment of the kidney, and the development of the vasa efferentia and seminal vesicles in Rana temporaria. Quart. J. Micr. Sci., 78:445-473. Greene, R. R. 1942 Hormonal factors in sex in- version: the effects of sex hormones on embryonic sexual structures of the rat. Biological Symposia, 9:105-123. Griinwald, P. 1937 Zur Entwickelungsmechanik des Urogenitalsystems beim Huhn. Roux’ Arch. Entw.-mech., 736:786-813. 1938 Entwicklungsmechanische Unter- suchungen iber die Genese einiger Fehlbild- ungen des Urogenitalsystems. Beitr. z. path. Anat. und z. allg. Path., 700:309-322. Gruenwald, P. 1941 The relation of the growing tip of the Miillerian duct to the Wolffian duct and 488 its importance for the genesis of malformations. Anat. Rec., 87:1-19. Gruenwald, P. 1942 Experiments on distribution and activation of nephrogenic potency in the em- bryonic mesenchyme. Physiol. Zool., 75:396—407. 1943 Stimulation of nephrogenic tissue by normal and abnormal inductors. Anat. Rec., 86:321-339. Hall, R. W. 1904 The development of the mesonephros and the Miillerian ducts in Am- phibia. Bull. Mus. Comp. Zool., Harvard College, 45:29-125. Hamilton, H. 1952 Lillie’s Development of the Chick. Henry Holt, New York. Hanaoka, K. 1941 The effect of testosterone- propionate upon the sex differentiation in Hyno- bius retardatus. J. Fac. Sci., Hokkaido Imperial Univ., Series VI, Zool., 7:413-419. Heys, F. 1931 The problem of the origin of germ cells. Quart. Rev. Biol., 6:1—45. Hill, R. T. 1950 Multiplicity of ovarian func- tions in the mouse. Arch. d’Anat. micr. et de Morph. exp., 39:634-642. Holtfreter, J. 1933 Der Einfluss von Wirtsalter und verschiedenen Organbezirken auf die Dif- ferenzierung von angelagertem Gastrulaekto- derm. Roux’ Arch. Entw.-mech., 127:619-775. 1934 Uber die Verbreitung induzier- ender Substanzen und ihre Leistungen im Triton- Keim. Roux’ Arch. Entw.-mech., 732:307-383. 1936 Regionale Induktionen in xeno- plastische zusammengesetzten Explantaten. Roux’ Arch. Entw.-mech., 134:466-550. 1938 Differenzierungspotenzen isolierter Teile der Urodelengastrula. Roux’ Arch. Entw.- mech., 738:522-656. 1944 Experimental studies on the de- velopment of the pronephros. Rev. Canad. de Biol., 3:220-250. Holyoke, E. A. 1949 The differentiation of em- bryonic gonads transplanted to the adult omen- tum in the albino rat. Anat. Rec., 103:675-699. Howland, R. B. 1921 Experiments on the effect of removal of the pronephros of Amblystoma punctatum. J. Exp. Zool., 32:355-395. 1926 Regeneration of the segmental duct and experimental acceleration of growth of the mesonephros in Amblystoma punctatum. J. Exp. Zool., 44:327-353. Huiybers, M. 1951 The influence of the gonads on the development of the reproductive system in the chick embryo. (Summary of Thesis, Anatomy and Embryology Laboratory, University of Am- sterdam. ) Humphrey, R. R. 1925 The primordial germ cells of Hemidactylium and other Amphibia. J. Morph. Physiol., 47:1-43. 1927 Extirpation of the primordial germ cells of Amblystoma: its effect upon the develop- ment of the gonad. J. Exp. Zool., 49:363-399. 1928a The developmental potencies of the intermediate mesoderm of Amblystoma when transplanted into ventrolateral sites in other em- bryos: the primordial germ cells of such grafts and their role in the development of a gonad. Anat. Rec., 40:67-101. SPECIAL VERTEBRATE ORGANOGENESIS 1928b Sex differentiation in the gonads developed from transplants of the intermediate mesoderm of Amblystoma. Biol. Bull., 55:317- 339. 1929 The early position of the primordial germ cells in urodeles: evidence from experi- mental studies. Anat. Rec., 42:301-314. 1933 The development and sex differ- entiation of the gonad in the wood frog (Rana sylvatica) following extirpation or orthotopic implantation of the intermediate segment and adjacent mesoderm. J. Exp. Zool., 65:243- 264. 1942 Sex inversion in the Amphibia. Bio- logical Symposia, 9:81—104. 1944 The functional capacities of hetero- plastic gonadal grafts in the Mexican axolotl, and some hybrid offspring of grafted animals. Am. J. Anat., 75:263-287. 1945 Sex determination in Ambystomid salamanders: a study of the progeny of females experimentally converted to males. Am. J. Anat., 7 6:33-66. 1948 Reversal of sex in females of geno- type WW in the axolotl (Siredon or Ambystoma mexicanum) and its bearing upon the role of the Z chromosome in the development of the testis. J. Exp. Zool., 109:171-185. Johnston, P. M. 1951 The embryonic history of the germ cells of the large-mouth black bass, Mi- cropterus salmoides (Lacépéde). J. Morph., 88: 471-542. Jost, A. 1947a Recherches sur la différenciation de l’embryon de lapin. 2. Action des androgénes synthése sur l’histogénése génitale. Arch. d’ Anat. micr. et de Morph. exp., 36:242-270. 1947b Recherches sur la différenciation de ’embryon de lapin. 3. Réle des gonades foetales dans la différenciation sexuelle somatique. Arch. d’Anat. micr. et de Morph. exp., 36:271-315. 1947c The age factor in the castration of male rabbit fetuses. Proc. Soc. Fxp. Biol. & Med., 66:302-303. 1948 Le contréle hormonal de la différ- enciation du sexe. Biol. Rev., 23:201—236. 1950 Sur le contréle hormonal de la dif- férenciation sexuelle du lapin. Arch. d’Anat. micr. et de Morph. exp., 39:577-598. , and Bergerard, Y. 1949 Culture in vitro d’ébauches du tractus génital du foetus de rat. Compt. Rend. Soc. Biol., 743:608. , and Bosic, B. 1951 Données sur la dif- férenciation des conduits génitaux du foetus de rat étudiée in vitro. Compt. Rend. Soc. Biol., 745: 647-650. , and Colonge, R. A. 1949 Greffe de testi- cule foetal de Rat sur l’adulte castré et hypophys- ectomisé. Remarques sur la physiologie du tes- ticule foetal de Rat. Compt. Rend. Soc. Biol., 143; 140. Keller, K., and Tandler, J. 1916 Uber das Ver- halten der Eihaute bei der Zwillingsstrachtigkeit des Rindes. Untersuchungen iiber die Entste- hungsursache der geschlechtlichen Unterent- wicklung von weiblichen Zwillingskalbern, welche neben einem mannlichen Kalbe zur Ent- URINOGENITAL SYSTEM wicklung gelangen. Wiener Tierarztl. Wochen- schrift, 3:513. Leroy, P. 1948 Effet androgéne d’extraits em- bryonnaires de Poulet sur la créte du Chapon. Compt. Rend. Acad. Sci., Paris, 226:520. Lillie, F. R. 1916 The theory of the free-martin. Science, 43:611. 1917 The free-martin: a study of the ac- tion of sex hormones in the foetal life of cattle. J. Exp. Zool., 23:371—452. Machemer, H. 1929 Differenzierungsfahigkeit der Urnierenanlage von Triton alpestris. Roux’ Arch. Entw.-mech., 778:200-251. Mangold, O. 1924 Transplantationsversuche zur Frage der Spezifitat und der Bildung der Keim- blatter. Arch. mikr. Anat. u. Entw., 700:198-301. Mintz, B. 1947 Effects of testosterone propionate on sex development in female Ambystoma lar- vae. Physiol. Zool., 20:355-373. 1948 Testosterone propionate minimum for induction of male development in Anurans; comparative data from other vertebrates. Proc. Soc. Exp. Biol. & Med., 69:358-361. Miura, K. 1930a Uber die Einfliisse der totalen Extirpation des ausseren Glomerulus auf die Vorniere bei Froschlarven. Jap. J. Med. Sci., I: Anat., 2:125-133. 1930b Experimentelle Untersuchungen liber die genetische Beziehung zwischen dem Wolffschen Gang und der Urniere bei Frozchlar- ven. Jap. J. Med. Sci., I: Anat., 2:105-124. Moore, C. R. 1941 On the role of sex hormones in sex differentiation in the opossum (Didelphys virginiana). Physiol. Zool., 14:1-45. 1944. Gonad hormones and sex differ- entiation. Amer. Nat., 78:97-130. 1947 Embryonic Sex Hormones and Sex- ual Differentiation. Charles C Thomas, Spring- field, Ilinois. ,and Price, D. 1942 Differentiation of em- bryonic reproductive tissues of the rat after trans- plantation into post-natal hosts. J. Exp. Zool., 90: 299-2965. Muchmore, W. B. 1951 Differentiation of the trunk mesoderm in Amblystoma maculatum. J. Exp. Zool., 778:137-180. Nieuwkoop, P. D. 1947 Experimental investiga- tions on the origin and determination of the germ cells, and on the development of the lateral plates and germ ridges in urodeles. Arch. Néerl. Zool., 8:1-205. 1948 Some further data concerning the determination of the mesonephros. Experientia, 4:391-394. 1949 The present state of the problem of the “Keimbahn” in the vertebrates. Experientia, 5:308-312. 1950 Causal analysis of the early devel- opment of the primordial germ cells and the germ ridges in urodeles. Arch. d’Anat. micr. et de Morph. exp., 39:257-268. O’Connor, R. J. 1938 Experiments on the devel- opment of the pronephric duct. J. Anat., 73:145- 154. 1939 Experiments on the development of the amphibian mesonephros. J. Anat., 74:34-44. 489 1940 An experimental study of the de- velopment of the amphibian cloaca. J. Anat., 74: 301-308. Padoa, E. 1936 Effetto paradossale (mascoliniz- zazione) sulla differenziazione sessuale di girini di Rana esculenta trattati con ormone follicolare. Mon. Zool. Ital., 47:285. 1938 La differenziazione del sesso in- vertita mediante la somministrazione di ormoni sessuali. Ricerche con follicolina in Rana escu- lenta. Arch. Ital. Anat. e Embr., 40:122-172. 1942 Il differenziamento del sesso inver- tita mediante la somministrazione di ormoni ses- suali e corticosurrenali. Ricerche con diidrofolli- colina, progesterone e acetato di desossicortico- sterone, in Rana esculenta. Pubbl. Staz. Zool. Napoli, 79:185-223. Pasteels, J. 1942 New observations concerning the maps of presumptive areas of the young am- phibian gastrula (Amblystoma and Discoglos- sus). J. Exp. Zool., 89:255-281. Piquet, J. 1930 Détermination du sexe chez les Batraciens en fonction de la température. Rev. suisse Zool., 37:173-281. Ponse, K. 1924 L’organe de Bidder et le déter- minisme des caractéres sexuels secondaires du Crapaud (Bufo vulgaris L.) Rev. suisse ‘Zool., 37: 177-336. 1948 Actions paradoxales des _ glandes géenitales. Rev. suisse Zool., 55:477-531. 1949 La différenciation du sexe et l’in- tersexualité chez les Vertébrés. F. Rouge, Lau- sanne. Price, D. 1947 An analysis of the factors influ- encing growth and development of the mam- malian reproductive tract. Physiol. Zool., 20:213- 247, Rawles, M. E. 1936 A study in the localization of organ-forming areas in the chick blastoderm of the head-process stage. J. Exp. Zool., 72:271- 315. Raynaud, A. 1942 Modification expérimentale de la différenciation sexuelle des embryons de sourls, par action des hormones androgénes et oestrogenes. Actual. Scient. et Indus., nos. 925 and 926, Hermann, édit., Paris. 1950 Recherches expérimentales sur le développement de Vappareil génital et le fonc- tionnement des glandes endocrines des foetus de souris et de mulot. Arch. d’Anat. micr. et de Morph. exp., 39:518-569. , and Frilley, M. 1947 Destruction des glandes génitales de ’embryon de souris par une irradiation au moyen des rayons x, a l’age de 13 jours. Ann. d’Endocrinol., 8:400-419. Rudnick, D. 1944 Early history and mechanics of the chick blastoderm. Quart. Rev. Biol., 79: 187-212. Salzgeber, B. 1950 Sterilisation et intersexualité obtenues chez l’embryon de poulet par irradia- tion aux rayons x. Bull. Biol. France et Belgique, 84:295-933. Seevers, C.H. 1932 Potencies of the end bud and other caudal levels of the early chick embryo, with special reference to the origin of the metane- phros. Anat. Rec., 54:217-246, 490 Shimasaki, Y. 1930a Uber die Resektion des Nephrostomalkanalchens der Vorniere bei Bufo- larven. Jap. J. Med. Sci., I: Anat., 2:277-289. 1930b Entwicklungsmechanische Unter- suchungen iiber die Urniere des Bufo. Jap. J. Med. Sci., I: Anat., 2:291-319. Spofford, W. R. 1945 Observations on the poste- rior part of the neural plate in Amblystoma. I. The prospective significance of posterior neural plate mesoderm. J. Exp. Zool., 99:35-52. 1948 Observations on the posterior part of the neural plate in Amblystoma. II. The in- ductive effect of the intact part of the chorda- mesodermal axis on competent prospective ecto- derm. J. Exp. Zool., 107:123-159. Stoll, R. 1948 Actions de quelques hormones sexuelles sur le développement des canaux de Miiller de l’embryon de poulet. Arch. d’Anat. micr. et de Morph. exp., 37:118-135. Swift, C. H. 1914 The origin and early history of the primordial germ cells in the chick. Am. J. Anat., 15:483-516. Torrey, T. 1950 Intraocular grafts of embryonic gonads of the rat. J. Exp. Zool., 115:37-58. Tung, T.-C. 1935 On the time of determination of the dorso-ventral axis of the pronephros in Dis- coglossus. Peking Nat. Hist. Bull., 70:115. ,and Ku, S.-H. 1944 Experimental studies on the development of the pronephric duct in anuran embryos. J. Anat., 78:52-57. Van Geertruyden, J. 1942 Quelques précisions sur le développement du pronéphros et de l’ure- tere primaire chez les Amphibiens Anoures. Ann. Soc. Roy. Belgique, 73:180-195. 1946 Recherches expérimentales sur la formation du mésonéphros chez les Amphibiens Anoures. Arch. Biol., 57:145-181. Vannini, E. 1946 Sex differentiation in Am- phibia. Nature, 757:812. von Winiwarter, H. 1910 La constitution et Vinvolution du corps de Wolff, et le développe- ment du canal de Miller dans l’espéce humaine. Arch. Biol., 25:169-268. Waddington, C. H. 1938 The morphogenetic function of a vestigial organ in the chick. J. Exp. Biol., 15:371-376. Wells, L. J. 1950 Hormones and sexual differ- entiation in placental mammals. Arch. d’Anat. micr. et de Morph. exp., 39:499-514. , and Fralick, R. 1951 Production of an- drogen by the testes of fetal rats. Am. J. Anat., 8§9:63-107. , and van Wagenen, G. 1954 Androgen- induced female pseudohermaphroditism in the monkey (Macaca mulatta): anatomy of the re- productive organs. Contr. Embryology, Carnegie Inst. Washington, 35:93-106. Wharton, L. R. 1949 Double ureters and associ- ated renal anomalies in early human embryos. Contr. Embryology, Carnegie Inst. Washington, 33:103-112. Wiesner, B. P. 1934 The postnatal development of the genital organs in the albino rat, with a dis- cussion of a new theory of sexual differentiation. J. Obst. Gyn. British Empire, 417:867-922. 1935 The postnatal development of the SPECIAL VERTEBRATE ORGANOGENESIS genital organs in the albino rat, with a discussion of a new theory of sexual differentiation. Jour. Obst. Gyn. British Empire, 42:8-78. Willier, B. H. 1921 Structures and homologies of free-martin gonads. J. Exp. Zool., 33:63-127. 1933 Potencies of the gonad-forming area in the chick as tested in chorio-allantoic grafts. Roux’ Arch. Entw.-mech., 730:616-649. 1937 Experimentally produced sterile gonads and the problem of the origin of germ cells in the chick embryo. Anat. Rec., 70:89-112. 1939 The embryonic development of sex; in Sex and Internal Secretions. 2d ed. Wil- liams & Wilkins Co., Baltimore. 1950 Sterile gonads and the problem of the origin of germ cells in the chick embryo. Arch. d’Anat. micr. et de Morph. exp., 39:267- 270. 1952 Development of sex-hormone ac- tivity of the avian gonad. Ann. New York Acad. Sci., 55:159-171. , Gallagher, T. F., and Koch, F. C. 1935 Sex-modification in the chick embryo resulting from injections of male and female hormones. Proc. Nat. Acad. Sci., 27:625-631. , Gallagher, T. F., and Koch, F. C. 1937 The modification of sex development in the chick embryo by male and female sex hormones. Physiol. Zool., 70:101-122. , and Rawles, M. E. 1935 Organ-forming areas in the early chick blastoderm. Proc. Soc. Exp. Biol. & Med., 32:1293-1296. Witschi, E. 1929 Studies on sex differentiation and sex determination in amphibians. II. Sex reversal in female tadpoles of Rana sylvatica fol- lowing the application of high temperature. J. Exp. Zool., 52:267-291. 1934 Genes and inductors of sex differ- entiation in amphibians. Biol. Rev., 9:460-488. 1939 Modification of the development of sex in lower vertebrates and in mammals: in Sex and Internal Secretions. 2d ed. Williams & Wil- kins Co., Baltimore. 1948 Migration of the germ cells of human embryos from the yolk sac to the prim- itive gonadal folds. Contr. Embryology, Carnegie Inst. Washington, 32:67-80. 1950 Génétique et physiologie de la dif- férenciation du sexe. Arch. d’Anat. micr. et de Morph. exp., 39:215-240. 1951a Embryogenesis of the adrenal and reproductive glands. Recent Progress in Hormone Research, 6:1-23. (Proceedings, Laurentian Hor- mone Conference. ) 1951b Adrenal hyperplasia in larval frogs treated with natural estrogens. Anat. Rec., 111:35-36. Wolff, E. 1938 L’action des hormones sexuelles sur les voies génitales femelles des embryons de poulet. Trav. Stat. Zool. Wimereux, 73:825- 840. 1946 Recherches sur l’intersexualité ex- périmentale produite par la méthode des greffes de gonades a l’embryon de poulet. Arch. d’Anat. micr. et de Morph. exp., 36:69-91. 1947 Essai d’interprétation des résultats URINOGENITAL SYSTEM 491 obtenus récemment chez les Vertébrés sur l’inter- de la différenciation du pénis chez le Canard. sexualité hormonale. Experientia, 3:272-276, Arch. d’Anat. micr. et de Morph. exp., 37:155- 301-304. 167. Wolff, E. 1950 Le réle des hormones embryon- , and Wolff, E. 1951 The effects of castra- naires dans la différenciation sexuelle des oiseaux. tion on bird embryos. J. Exp. Zool., 716:59-97. Arch. d’Anat. micr. et de Morph. exp., 39:426- , Strudel, G., and Wolff, E. 1948 L’action 444, des hormones androgeénes sur la différenciation , and Ginglinger, A. 1935 Sur la trans- sexuelle des embryons de poulets. Arch. d’Anat., formation des poulets males en intersexués par d’Hist. et d’Embry., 37:237-310. injection d’hormone femelle (folliculine) aux Yamada, T. 1937 Der Determinationszustand des embryons. Arch. d’Anat., d’Hist. et d’Embry., 20: Rumpfmesoderms im Molchkeim nach der Gas- 219-278. trulation. Roux’ Arch. Entw.-mech., 737:151- , and Haffen, K. 1952 Sur le développe- 270. ment et la différenciation sexuelle des gonades 1950 Dorsalization of the ventral mar- embryonnaires d’oiseau en culture im vitro. J. ginal zone of the Triturus gastrula. I. Ammonia- Exp. Zool., 779:381-403. treatment of the medio-ventral marginal zone. , and Wolff, E. 1948 Sur le déterminisme Biol. Bull., 98:98-121. Section VII CHAPTER 7 Teeth ISAAC SCHOUR INTRODUCTION THE ontogenetic history of the tooth is an interesting and fruitful chapter in develop- mental histophysiology. The tooth does not belong to the osseous system. It is a highly specialized appendage and passes through many developmental stages. These permit an analysis of different types of growth and are governed by definite principles the integra- tion of which makes for orderliness in the elaboration of the form, size and function of the tooth. The continuously growing incisors are es- pecially useful for the quantitative and qualitative analysis of developmental proc- esses. Thus the rodent incisor may be re- garded as nature’s gift to research in tooth development (Schour and Massler, ’49). The component dental structures when growing are highly sensitive to physiological and metabolic processes which become per- manently recorded in the completed and calcified enamel and dentin. The tooth is an organ of mastication, but it can also be uti- lized as a permanent biological kymograph of the life history of the growing individual (Massler, Schour and Poncher, ’41). The tooth proper consists of the calcified enamel, dentin and cementum and the soft internal pulp. However, function calls for a close organic integration with the surround- 2) [Mesénchy me B Proliferation. Histodifferentiation Appositio Initiation (Bud stage) (Cap stage (Bell stage) and GROWTH nm CALCIFICATION nee (Into oral cavity) ing gingiva, periodontal membrane and al- veolar bone, and topographical correlation with the adjacent and opposing teeth. This section will be confined largely to a consid- eration of the growth of enamel and dentin. The progressive development of the teeth consists of the following stages (Noyes, Schour and Noyes, °48) (Fig. 191): I. Growth . Initiation (chemodifferentiation of Huxley ) . Proliferation . Histodifferentiation . Morphodifferentiation . Apposition II. Calcification III. Eruption IV. Attrition. Table 15 outlines the morphologic se- quence of events and the corresponding phy- siological processes. These overlap consider- ably and many of them may occur at the same time in different parts of the tooth. Table 15 also summarizes the developmental physiology of the tooth. — Or & OW bo GROWTH INITIATION Specific basal cells at definite sites on the dental lamina of the oral epithelium become ERUPTION ATTRITION Fig. 191. Diagrammatic representation of life cycle of tooth (from Schour and Massler, ’40). 492 493 TEETH MOT}I1}}e [vUIOUqe Ul Zux4]NsSeJ ‘(sIsorony) Jemvus JO uolvOYIO[vO RUsIDYeD :d1Ule3ySAG *U01} -oUN} IMosNU VAIsseodxe UMOJO JO BULIBVAM [euiouge ‘U01711438 uoydnia Aq paqyes 0J onp «vam jeuUIOUqe dAIsseoxy *(UOISN[DIO -uedWoy “IwaM [Buoy UMOJO JO sysluosejue jo [RAOWsl -eidns) uolyesuoya -ounj WOdj Burq[Nsed qysiey Ul 0} onp IvaM YUSIOYep :[BI0T U01}11978 JUSIOYIG 43009 jo Bulyieyy sssooid YyMOIB9q UOoljoNperYy uot. Vy uonMdnia Jo a4¥l UOISNPIO[V YL pesveloul JO; BUIMOTLP “uUOIzISOd|eyy *(UOIS SjsIuOSseIUR JO aN] S10} 0B] olua4ysAs Aq -njoo0-vidns) uo0les uorjouny [Bsnjo00 1O JeAowsd :uodnis pojJooye ATIpBal sso] -uoje ‘uolydnie aAtIs ul 43900} JUsov[pe pueB Bul JO 9301 pesvaioap Zuisneo !SUOTJIPUOD [BIO] Aq -Soox@ “UOISIEUIqnNS Sulsoddo 4aoul 03 44009 -uoljIsod WOIsNN00 oIWBWINeI} :[ROOT poajooye AjIsva AloA ‘uoldnise yusIDyeqd 4300} Jo BulIyIe yyy peul1oj JO UOTZBIBIPL [esnp209 uondnig S1ap10q (‘aye avuomentmanGh ane ‘SouRjUL ‘BuluvEM Pios3so opIm AIM ‘yusujsn{pe [eyeu auoq pue uwInjueUL -Oau) [BNplAIpul -99 ‘uljUep-eid opIM jeweus jo josnjeqsjeuoynyys ‘(uNUuep aelseIj pues Ayloedo a31yM-Ay[eYo sis -U09 S}OaYeI : UISTTO Iv[NqGO]s1984UT) SsaqI uoljoRIYIp ABI-x uoiydios -O10Nn[ YT “UljUep Ie[NqGo]s -qejow Apoq ul -Jaydsoo]vo Jo uOIsn} {seqyLiaydsoo[eo Jo -a1 o1s0j0IsAYd 03 yool -19}UL : AULOJDOPIOIAYY SUO}BIIvA [RWIOU -UOU SUIMOYS UIUEd uoIsNnj} JO 9va1Zap -qns jou ‘au0q exIyUuN -Bieg “uljuep IeB[Nqo]s udaABd 0} DATZISUGS ‘aiseiy pue Ayleyo Suoljovel BUT ‘uljuep pue joueun -19}UI pue UljUep-sid ‘queudojaAep ul *JUBYSISOI-ploB -UIe4S d1B0[04SIY *XLI]VUl UlezO1d UL S4]eVS ssoupiey apIM iq] sisoulwmezIAOdATT aseyd aaTjIsues 4ysopPY PUREIN|VUIWIIJeUIVU ‘JoJaWIOISUep ABI-X [RJeUlW Jo U0lWeIIdloeIg JO 90130q UOlBOYO[VD uloqjed oeuesoydioul ® uo pasodiedns saaino otuoW0Us petty JO sqsisuod yorum uleqyed [ejUEeWaIOUL UB 24;NSeYY ‘(ueds ej] [BuOoTjoUNy) uoyeinp pue ‘(Asojou uljuep jo uoIsnyo -o1yo) owl} +‘ (u1eqyed -UI Iv[nosvA pue viseid (uin1}uU0 14s YjMO1IZ puv S1e}U99 -odAy jeurevue :yjMOIs JO (UOIZBUIOJ io ‘opliony YjMO1S) SeqIs ‘squad uo queIpeis peqinjsip A[pe xljeul JO yunowe wUINIpos ‘g pel ule -Ipeids pue soqyel UdAl3 einjoniy4s -BUIIOJ JOO uoljeiqua -yievw fy sisoururezlAodAyy ys juatloyep) visejdodAyY -ziye) Surureqys yeqiA 38 xlujeul jo uolyIsodeq jo yunowy puv uMOID -loylpourxny uorljIsoddy ‘¢ srolount |e} ueWe0-oulQUep pure jewuevue-ouljuep uoly s]]e0 jo Zululyjng “peunsse azis so3v4s -B1yudle YIP wnIps py aZzIs puv WIO] IepnselIT jo JuewesuRiIy sl uloqed o1so0joydioyy pue wi0og [Jeq pue dep -oydloy, “Ff dn uaals st AyloedeBo dalqesojpl[OIg *paysl[qeyse oie APIATOR jo oje1 pue ueds ofl] jeuoljouny Ieyy “Uy -uep pur jawieus UIIOJ Aousloyep 0y «yjeQueyod yyMoIs O ulureziA “dind jo wioj almboe pue ‘ieuuNnyjoo ainjond4s uolze1}Ua UOT} BIVUBIOT pue UOljeULIOy UljUep [vo MOT ainjoniys [Rod AV Vy SIsA[BVuR 1Z0]OJAD = [BJ aUIODEq sT[J90 oYloodg = Jo AyTTeNH -JaBIpoqsi py -JIpoqsIFy “E -IdA}e@ :Y SISOUIUIeIAOdAP UBBIO DAIJVUIOT (djnd ojutr jo UOIRIOGRIY *€ uUOISBPAUL pueB uintjeyyide WO} UL asueyO a Jeulvus jo uolneiseyljoid azIs UL BsvelOUT *T peqiqiyulun) YW sisourut soin3y :UI BULQ[NSel UTZ so3e4S -eqaodAy *(4}M018 9A} o1707IW JO JaquUINNy -voldiyjnw [j90 pidey devo pue png UOlZRIOJIJOIg *Z -Blojijoid JUsToyep) BUT -uosiod pee] pu sulolyojop, YIMOIs eully aqluyed °Z jo soseyd yuenbesqns aZIS poulWlJoJopeld *T y4004 uo eIyua MOT [1B JO saouequnysiq 238 YjMOI3 JO UIZUQ, «jo JequUINN' Bulue,{eJUeq -loyIpowmeYy UoljzeIqIUy *T YJMOI SHONVAUNLSIG SAONGYGAUALNI =NI LTASaY AdOLS 40 SGOHLAW SLNGAG GNV :NO O1dINGdS DNIMOHS TIVINGWIYadXa INGNdOTAATA AAILVLITVND AGNV SOILSIUALOVUVHO GONAN TANI SADVLS ADOTONINUGL SADVLS SNOILIGNOO TIVLNAWINGdxXa OL ALIAILISNAS NI SNOILVUUEV AAILVLILNVAD OIDOTOISAHdOLSIH uOrvNN OIDOTOHAUOW S$, AGTXNH TVINGANdOTAAGG (Of, ‘42]SSD]J\ puv snoyog woif pajdvpy) uljuagq pup jaumuy ‘“yiI0oT, ay fo AsojoisAyg Joyuauidojaaeq “GT ATaVI, 494 odontogenic. Experiments on the induction of tooth development have been conducted in amphibians. Holtfreter (’35) transplanted indifferent ectoderm of the tree-frog, Hyla, to the ventral head region of the newt, Triton. Horny teeth developed which were typical of the donor species, but did not occur in the host species. Sellman (746), on the basis of extirpation and transplantation experiments in the uro- deles which are equipped with dentin teeth even as larvae, concluded that tooth forma- tion requires three components: (1) neural crest, (2) presumptive mouth ectoderm and (3) oral entoderm (the oral plate of the fore- cut). Andres (46) transplanted undetermined ventral ectoderm of a gastrula of a toad to the lateral head region of a salamander neu- rula. Though this ectoderm is induced by the host to form a variety of tissues, such as cartilage, teeth did not develop. PROLIFERATION The odontogenic cells undergo rapid mi- totic multiplication leading to changes in the form and size of the enamel organ (the bud, cap, and bell stages, Fig. 191). Adjacent connective tissue cells also proliferate to form the dental papilla (the pulpal organ) and the dental follicle (the periodontal organ). The enamel organ, the dental papilla, and the dental follicle together form the tooth germ. The enamel organ does much more than its name implies. In addition to supplying the inner enamel epithelium which gives rise to the ameloblasts it exerts an organizing influence on the adjacent mesenchymal cells and outlines the future dentino-enamel and dentino-cemental junctions. It would, there- fore, be more appropriate to call it the odon- togenic organ. It appears that even at this early stage the tooth germ contains the entire growth po- tential of the future tooth. Explants of this stage of development continue to develop in tissue culture through the subsequent stages of histodifferentiation and appositional growth (Glasstone, ’36). Proliferation of a given cell normally ceases with the assumption of the next stage of histodifferentiation. The inverse relation between proliferation and differentiation and specific activity is well illustrated in the ameloblasts. If they do not attain differentia- tion they continue to proliferate excessively with resulting cyst or tumor formation SPECIAL VERTEBRATE ORGANOGENESIS (Thoma (ameloblastoma ) ’46). and Goldman, HISTODIFFERENTIATION The formative cells undergo structural as well as chemical changes. They give up their proliferative activity. The cells of the inner enamel epithelium differentiate into tall col- umnar ameloblasts and exert an organizing influence upon the subjacent mesenchymal cells of the dental papilla, which then dif- ferentiate into odontoblasts. Chemical Interaction and Interdependence of Epithelium and Odontoblasts; Transplanta- tion Experiments. Von Brunn (1887) and others have shown that the presence of the inner epithelium is essential to the differen- tiation of the odontoblasts and the initiation of dentin formation. However, once differen- tiation has reached a certain stage, the odon- toblasts can proceed with dentin formation without the further presence of the epithe- lium. Transplantation experiments with tooth germs of higher vertebrates beginning with those of Legros and Magitct (1874) have yielded further evidence on the interdepend- encies in tooth development. Huggins et al. (34) observed dentin formation in 14 days following transplantation of odontoblasts to the abdominal wall. Enamel, on the other hand, will not form in the absence of odonto- blasts or dentin (Huggins et al., 34; Glass- tone, °36). Transplanted ameloblasts lose their cylindrical character, change to strati- fied squamous epithelium and fail to form enamel unless the odontoblasts accompany the transplant (Hahn, ’41). Dentin forma- tion, therefore, precedes and is essential to enamel formation, although the presence of the epithelial cells and their chemical inter- action precede and are essential to the differ- entiation of the dentin-forming cells and the initiation of dentin formation. The influence of dentin on enamel forma- tion may be indirect by reversing the stream of tissue fluid (Wassermann, ’44). The chem- ical influence of epithelial cells upon ad- jacent mesenchymal cells is not only evi- denced in the tooth but also has been demonstrated in bone growth. Transplanta- tion of epithelium of the gallbladder or of the urinary tract distal to the kidney causes the differentiation of the adjacent mesenchy- mal cells into osteoblasts and initiates the formation of bone (Huggins, 731). Concomitant with their morphological dif- TEETH ferentiation, the cells acquire their func- tional assignment and their appositional growth potential to form enamel or dentin. This potentiality may be defined in terms of the amount of work capacity of the cell and is expressed in an orderly sequence of events and processes which occur during the stage of apposition. The growth energy released at the initia- tion of development is unorganized and must be distributed according to a definite growth pattern. The process of differentiation is like the direction of a stage play. It occurs in that period in development at which “roles are assigned, cues fixed, appearances timed, and the stage set” (Weiss, 39). The orderliness of the actual performance (apposition) which follows is entirely dependent on the proper differentiation of the cells and their proper environmental condition. MORPHODIFFERENTIATION This stage marks the assumption of the morphological pattern of the tooth. The cells of the inner layer of enamel organ arrange themselves to outline the dentino-enamel junction, which serves as a blueprint pattern of the future form and size of the tooth. This junction must be established before any enamel or dentin is deposited, since these structures become calcified soon after they are formed and cannot change thereafter. Both histo- and morphodifferentiation first occur at the tip of the tooth and then proceed toward the apex. Hertwig’s Epithelial Sheath. At the margins of the bell-shaped enamel organ, the inner and outer layers of the enamel epithelium proliferate and give rise to Hertwig’s epithe- lial root sheath. This epithelial sheath out- lines the dentino-cemental junction and acts as the blueprint pattern for the shape, size and length of the future root or roots, just as the inner enamel epithelium outlines the shape and size of the crown. In addition, the epithelial sheath initiates the differentia- tion of the radical odontoblasts just as the ameloblasts initiate the differentiation of the coronal odontoblasts. The cementoblasts probably owe their differentiation to the chemical stimulus of calcified dentin. As soon as the formation of the dentin and cementum of the root is begun, the sheath disintegrates and vestiges can be found later as epithelial rests in the periodontium. A disturbance in morphodifferentiation and also in proliferation results in a dis- 495 turbance in the form and size of the dentino- enamel junction. This leads to abnormal forms and sizes such as the peg tooth or Hutchinson’s incisor (screwdriver-shaped in- cisor). The field concept of development has been applied by Butler (39) to phylogenetic prob- lems of tooth morphology. He pointed out that certain tooth characters are manifested in groups of teeth to a maximum degree in particular key teeth in each of the incisor, canine, and molar groups. Dahlberg (745) suggested that tooth anomalies are related specifically to the tooth districts rather than to the dentitions as a whole. Thus, certain points in the human dentition are more sus- ceptible to change than are others. APPOSITION In contrast to the rapid, multiplicative, mitotic and cellular type of proliferative growth, appositional growth is slow, additive, incremental and extracellular (Huxley, ’32). Apposition of enamel constitutes the fulfill- ment and full expression of the growth po- tential acquired by the ameloblasts during histodifferentiation. Appositional activity of the ameloblasts begins at specific sites, the growth centers, and proceeds at a definite time and chro- nology, at definite rates and gradients and for a definite number of days, the functional life span of the formative cells. The end re- sult is the incremental pattern (Schour and Massler, *40). Physiological Characteristics of Enamel and Dentin. An analysis of the incremental pat- tern is facilitated by three unique physiolog- ical characteristics of the enamel and dentin: 1. The Rhythmic Manner of Appositional Growth and Calcification. The hard struc- tures of the tooth, like the trunks of trees, grow by the regular and rhythmic formation of concentric layers or rings. 2. The High Sensitivity of the Growing and Calcifying Tissues to Fluctuations of Metabolic Processes (Particularly Mineral Metabolism). The growing tooth depends for its raw materials upon the substances elab- orated by the body. Fluctuations in the metabolism of the individual are, therefore, reflected in the layers or rings of the tooth forming at that time. The normal incre- mental rings are accurate records of the ontogenetic development of the tooth and the normal physiological fluctuations in the metabolism of the individual. 496 Accentuations of these rings may occur at any given time as a result of severe met- abolic fluctuations or disturbances. They are seen as the striae of Retzius in the enamel and Owen’s lines of contour in the dentin and may be of physiological or pathological origin. The neonatal ring, for example, re- flects the physiological readjustments coin- cident to birth and is the result of the brief neonatal arrest in growth (Schour, ’36). An analysis of various experimental endocrine and vitamin disturbances shows that each particular dysfunction produces character- istically disturbed incremental rings which are superposed on the basic formative pat- tern. The annual rings of the tree similarly reflect the variations and vicissitudes in cli- mate that the tree had experienced during its growth period. 3. The Permanence of the Tissues. Enamel once completely formed and calcified can be destroyed by oral environmental factors but not by systemic alterations. Although com- pleted, the dentin of deciduous teeth under- goes physiological resorption. Enamel, in fact, has lost its formative organ and has not even the power of repair. All available evi- dence shows that the dentin, as well as the enamel, is not subject to calcium withdrawal. These tissues serve, therefore, as permanent records of physiological or pathological dis- turbances of metabolism that may occur within the individual during their formative and calcifying stages. This permanence of structure is not found in bone although it also grows in an appositional manner and registers within its structure the effects of disturbances in body metabolism. The rec- ords in bone are erased by constant resorp- tions and reconstructions. Bone thus possesses two of these physiological characteristics of the tooth but lacks the third—its perma- nence. The Growth Centers. The dentino-enamel junction is characterized by definite high points which correspond to the number of cusps (in the posterior teeth) or lobes (in the anterior teeth). Amelogenesis and den- tinogenesis, just as proliferation and _histo- differentiation, begin at these individual points and proceed at a specific rate and gradient of growth. Each summit on the dentino-enamel junction thus acts as an in- dividual growth center from which the growth begins and radiates outward in a definite growth plan. The Incremental Cones. Beginning at each growth center, successively adjacent amelo- blasts begin their formation at successively SPECIAL VERTEBRATE ORGANOGENESIS later intervals, possibly a day apart. The cellular activity spreads peripherally along the dentino-enamel junction like a ripple will spread from a pebble dropped into calm water. Each ameloblast proceeds outwardly away from the dentino-enamel junction at its own characteristic rate and gradient of growth until the required length of the enamel rod is reached. Similarly each odon- toblast recedes inwardly away from the dentino-enamel junction. Any given in- cremental layer of enamel or dentin assumes in three dimensions a conical form. Its apex is directed occlusally and its base rests upon the dentino-enamel junction. These incre- mental layers are apposed at each growth center, one over the other in the enamel and one within the other in the dentin. The resultant incremental growth pattern con- sists of a series of gnomonic curves whose form is determined by the dentino-enamel junction. When the incremental layers of adjacent growth centers meet, as in the molar teeth, the subsequent incremental layers are de- posited as fusions of individual cones. Rates and Gradients of Appositional Growth. Vital staining with sodium fluoride (Schour and Poncher, °37) and alizarine red S (Schour et al., 41) offer a ready method for measuring the rate of apposition. Each injec- tion produces a distinct ring in the enamel or dentin formed at the time. The distance between two successive experimental rings represents the amount of deposition during the time interval that elapsed between the in- jections. In human teeth, the average rate of apposi- tion of enamel and dentin is approximately 4 micra per day. This rate, however, decreases as one proceeds from the incisal or cuspal tip toward the gingival level (locus gradi- ent); from the anterior incisors to the pos- terior molar teeth (anteroposterior gradient) ; and within the same cell, from the begin- ning of functional activity toward its ter- mination (age gradient). As Thompson (717) has pointed out, the method of appositional growth combined with growth gradients may result in a spiral form which is especially evident in the rodent incisor. Appositional Growth Potential and Forma- tive Life Span. In the tooth it is possible to measure the growth potential of the amelo- blast and to assess its functional life span (Massler and Schour, *46). Since only one ameloblast is responsible for a given enamel rod, its length may be taken as a measure of the growth potential and the growth work TEETH done by the cell. The formative life span in days may then be obtained by dividing the length of the enamel rod by the daily rate of formation. Thus, L (length of enamel rod) = G. P. (growth potential) = (Daily) rate X life span, = R X T = Time of cellular activ- ity (functional life span). In the dentin the length of the dentinal tubule represents the growth potential or growth work of the odontoblasts. Such quan- titative studies permit a detailed analysis of the growth pattern and size and form of dif- ferent classes and types of teeth. It is of interest to note that these rates and gradients of growth are not readily al- tered by environmental factors. Characteris- tic gradients in different regions of the same tooth and in different classes and types of teeth appear to correspond with their par- ticular form and contour. This conforms with the statement of D’Arcy Thompson (’17): “A very large part of the specific morphology of the organism depends upon the fact that there is not only an average rate of growth common to the whole, but also a variation of rate in different parts of the organism. The smallest change in the relative magni- tudes of these partial or localized velocities of growth will soon be manifested in more and more striking differences of form.” The functional life span of the amelo- blasts shows wide gradients, with a maximal life span for the cells over the growth center and a minimal one for those near the ce- mento-enamel junction. While the odonto- blasts, in contrast to the ameloblasts, persist and function throughout life, that period of their activity which is responsible for the formation of the primary dentin tends to be constant (in man, about 350 days for deciduous and about 700 days for the permanent teeth). Thus the presence and morphology of the pulp can be ex- plained on the basis of the limited and decelerating rate of appositional activity of the odontoblasts. CALCIFICATION Calcification consists of the deposition of mineral salts and their crystallization. It is a process which does not involve a change in size, but a reorientation of molecular structure and content of the deposited matrix leading to increased polymerization of the ground substance and the precipitation of minerals. The enamel and dentin in most species show a common basic incremental calcifica- 497 tion rhythm which recurs at intervals of ap- proximately 16 micra. The common basis of this 16 micron calcification unit is possibly associated with the physicochemical factors concerned in calcospherite formation and the Liesegang ring phenomenon. Striae of Retzius in the enamel and Owen’s lines of contour in the dentin constitute physiologi- cal or pathological accentuations of the nor- mal incremental rings (Schour and Hoff- man, °39). In the dentin, calcification follows apposi- tion in close succession (one day interval in the rat). In the enamel, the matrix first con- sists of organic substance (and water) and mineral salts in the ratio of two to one. During the secondary or final calcification the organic substance becomes increasingly impregnated by mineral salts until the ma- ture enamel consists of 96 to 98 per cent inorganic material. ERUPTION Eruption is the process by which the tooth migrates from its intraosseous position within the jaw into the oral cavity in order to reach and maintain articulation. The piercing of the tooth through the oral mucosa is only a momentary and transitory event. Eruption continues throughout the life of the tooth. In the rat, the rate of eruption is about 2 mm. per week in the upper incisors, and about 3 mm. in the lower. Studies in the rabbit show that most of the eruption occurs when the animal is at rest and the incisors are out of occlusion (Rink, ’29). Experi- mental removal of the opposing incisor re- leases the eruption potential and results in the doubling of the rate of eruption. ATTRITION Attrition may be defined as the normal wearing of the teeth due to functional ac- tivity. Attrition is a degrowth process and provides an exception to the rule that tooth development proceeds without influence of function (note the development of teeth in dermoid cysts). The charting of the rate of attrition results in a straight line, sug- gesting that it is a mechanical process which is independent of growth (Hoffman and Schour, *40). The continuous process of at- trition is compensated by eruption and serves to regulate articulation. In some species, as in the herbivora and rodentia, the teeth are made functional by attrition. In the rodent incisor, the rate of eruption 498 and of attrition is readily measured by notch- ing the exposed enamel surface at the gin- gival margin. The decrease in distance be- tween the notch and the incisal edge is a measure of the rate of attrition. The increase in distance between the notch and the gin- gival margin is a measure of the rate of eruption. REFERENCES Andres, G. 1946 Uber Induktion und Entwick- lung von Kopforganen aus Unkenektoderm im Molch (Epidermis, Plakoden und Derivate der Neuralleiste). Revue suisse Zool., 53:502-510. Brunn, A. v. 1887 Ueber die Ausdehnung des Schmelzorgans und seine Bedeutung fiir die Zahnbildung. Arch. f. mikr. Anat., 29:367-383. Butler, P.M. 1939 Studies of mammalian denti- tion. Differentiation of postcanine dentition. Proc. Zool. Soc., London, 109:1—36. Dahlberg, A. A. 1945 The changing dentition of man. J. Am. Dent. Assoc., 32:676-690. Glasstone, S. 1936 Development of tooth germs in vitro. J. Anat., 70:260-266. Hahn, William E. 1941 The capacity of develop- ing tooth germ elements for self-differentiation when transplanted. J. Dent. Res., 20:5-19. Hoffman, M. M., and Schour, I. 1940 Quantita- tive studies in the development of the rat molar. II. Alveolar bone, cementum and eruption (from birth to 500 days). Am. J. Orthodont., 26:854— 874. Holtfreter, J. 1935 Experimentell erzeugte Chi- maren aus den Organanlagen von Frosch- und Molchkeimen. Sitzgber. Ges. Morphol. Physiol. Miinchen, 44:24-32. Huggins, C. B. 1931 The formation of bone un- der the influence of eptihelium of the urinary tract. Arch. Surg., 22:377-408. , McCarroll, H. R., and Dahlberg, A. A. 1934 Transplantation of tooth germ elements and the experimental heterotopic formation of dentin and enamel. J. Exp. Med., 60:199-210. Huxley, Julian S. 1932 Problems of Relative Growth. Lincoln MacVeagh, The Dial Press, New York. Legros, Ch., and Magitot, E. 1874 Physiologie Experimentale—Greffes de follicules dentaires et de leurs organes constitutifs isolement. Compt. rend. Séances de l’Academie, 78:357—360. SPECIAL VERTEBRATE ORGANOGENESIS Massler, M., and Schour, I. 1946 The apposi- tional life span of the enamel and dentin-forming cells. J. Dent. Res., 25:145-150. , Schour, I., and Poncher, H.G. 1941 De- velopment pattern of the child as reflected in the calcification pattern of the teeth. J. Dis. Child., 62:33-67. Noyes, F. B., Schour, I., and Noyes.H. 1948 Oral Histology and Embryology, 6th ed. Lea & Febi- ger, Philadelphia. Rink, Karl 1929 Rules of growth in rabbits. Viertelj.-schr. f. Zahnh., 45:543-561. Schour, I. 1936 The neonatal line in the enamel and dentin of the human deciduous teeth and first permanent molar. J. Am. Dent. Assoc., 23:1946- 1955. , and Hoffman. M. M. 1939 Studies in tooth development. I. The 16 microns calcifica- tion rhythm in the enamel and dentin from fish to man. J. Dent. Res., 75:91-102. . Hoffman. M. M., Sarnat, B. G.. and Engel, M. B. 1941 Vital staining of growing bones and teeth with Alizarine Red S. J. Dent. Res., 20:411-418. . and Massler, M. 1940 Studies in tooth development: The growth pattern of human teeth. J. Am. Dent. Assoc., 27:1778-1793, 1918- 1931. , and Massler, M. 1949 The teeth; in The Rat in Laboratory Investigation, by Griffith and Farris, 2nd ed., Chapter 6. J. B. Lippincott Co., Philadelphia. , and Poncher, H.G. 1937 Rate of apposi- tion of enamel and dentin as measured by the ef- fect of acute fluorosis. Am. J. Dis. Child., 54:757- 776. Sellman, Sven 1946 Some experiments on the determination of the larval teeth in Ambystoma mexicanum. Odontologisk Tidskrift, 54:1-128. Thoma, Kurt K., and Goldman, Henry M. 1946 Odontogenic tumors: A classification based on ob- servations of the epithelial, mesenchymal, and mixed varieties. Am. J. Path., 22:433-471. Thompson, D’Arcy W. 1917 Growth and Form. Cambridge University Press, Cambridge, Eng- land. Wassermann, F. 1944 Analysis of enamel forma- tion in the continuously growing teeth of normal and vitamin C deficient guinea pigs. J. Dent. Res., 23:463-509. Weiss, Paul 1939 Principles of Development. Henry Holt & Co., New York. Section VII CHAPTER 8 Skin and Its Derivatives* MARY E. RAWLES INTRODUCTION ALTHOUGH much information is available on the detailed histological and physiological characteristics of the adult skin and _ its numerous derivatives in the vertebrates, com- paratively little attention has been given to the embryological origin of these character- istics. The purpose of this account is, there- fore, to analyze the interrelationship and interaction of the ectoderm and mesoderm, the two primordial tissue components which unite in early development to form the com- posite structure, the skin and its diverse types of derivatives. Special attention will be given to (1) the causal relations and interactions involved in the differentiation of the dermal and epidermal components, (2) regional specialization of the skin with particular reference to the role of the meso- derm in epidermal specialization, and (3) integumentary patterns, structural and pig- mentary. ORIGIN OF THE SKIN Early Beginnings of an Integrated System. Embryologically the skin of all vertebrates is a composite structure formed by the join- ing of two main contributions from separate sources, the ectoderm and the mesoderm. As a result of the formative movements of cells in gastrulation, prospective mesoderm is brought in contact with the inner sur- face of the epidermal portion of the prospec- tive skin ectoderm. During this process the mesoderm appears to acquire a positive “af- finity” for ectoderm, for prior to gastrula- tion prospective mesoderm resists fusion with ectoderm (Holtfreter, ’39). This union be- tween the two is probably the resultant of progressive changes in both ectoderm and mesoderm and is not to be regarded simply *The writer is indebted to her husband, Mr. John S. Spurbeck, for the preparation of the illustra- tions. 499 as an operation of chance. The embryonic layers are presumably held together at first by complementary forces residing directly in the naked contact surfaces of their con- stituent cells. Sooner or later these primary affinities are superseded by more permanent unions in the form of “cementing tissues” or basement membranes (see Weiss, 747). Thus the embryonic components enter into an intimate union at the surface of the body and behave afterwards as an_ integrated system. Source Material of the Epidermis and Der- mis. The entire surface ectoderm between the union of the neural folds in the median dorsal line and the median ventral line of the embryo may be regarded roughly as prospective skin epidermis. By means _ of vital staining and other techniques its ori- gin has been traced to a surface area in the early gastrula in the amphibians and in equivalent stages in the chick (Vogt, ’25, ’29; Pasteels, °37). Through growth and move- ments of the skin ectoderm and the accom- panying gastrulation processes, the entire embryo is covered by ectoderm at the late gastrula stage (amphibian) or later stages (chick). Strictly speaking, the prospective fate of the skin ectoderm is not entirely epi- dermal, for in certain loci its development is modified by induction in specific direc- tions, as for instance into lens and con- junctiva over the primordium of the eye- cup, oral epithelium and parts of teeth in the mouth, inner ear in the region of the myelencephalon, etc. Nevertheless, the ecto- derm in such loci is capable of forming epi- dermis when grafted to atypical positions, e.g., mouth ectoderm grafted to trunk forms epidermis (Stréer, 33). In a balanced physio- logical salt solution explanted ectodermal cells merely undergo proliferation and form only epidermis (Holtfreter, ’31); and the ectoderm of exo-gastrulae remains a wrin- kled mass of epidermal cells (Holtfreter, 33a, b). Furthermore, it is well known that 500 prospective conjunctival ectoderm in the ab- sence of contact with the optic vesicle re- mains pigmented and opaque like the sur- rounding true epidermis. The ectoderm thus appears to have a primordial capacity to form epidermis. Inasmuch as the mesoderm is a relatively thick layer, the upper surface of which is in contact with skin ectoderm while its lower surface is in close union with gut entoderm and is the source of a variety of organs and tissues, the question arises as to what portion of the mesoderm contributes to the formation of the dermis. Does the SPECIAL VERTEBRATE ORGANOGENESIS (the unsegmented mesoderm lateral to the somites, within which the body cavity arises) plus the closely united ectoderm and ento- derm will also produce normal skin and skin derivatives (feathers or hair, as the case may be) when grafted to the embryonic chick coelom as seen in Figure 192 (Rawles, 47; Straus and Rawles, ’53) or to the chorio- allantoic membrane (Murray, ’28). In view of the fact that it is assumed fre- quently on the basis of morphological studies (Engert, ’00, and others) that the dermis, not only of the dorsal and dorsolateral re- gions but of the other body regions as well. Fig. 192. Section through skin and down feathers developing in intracoelomic graft of lateral plate isolated from right side of 28-somite chick embryo at level of somites 22-25. Total age 10% days. Compare with skin and feather germs of normal, Fig. 193 C. Iron hematoxylin, 8u 75. dermis arise directly from mesodermal cells which are everywhere in contact with and subjacent to the skin ectoderm, or is its origin restricted to the so-called ‘“derma- tome,” one of the three arbitrary divisions of the somite? An unequivocal answer to this old and controversial problem has been fur- nished for the chick and the mouse by the use of experimental techniques. Isolates of early limb buds, taken prior to the entrance of cells from the neural crest, and consist- ing of somatic mesoderm covered by skin ectoderm, will produce structurally normal skin and skin derivatives, feathers and hair. respectively, when grown in the coelom of a chick embryo host (Hamburger, 7°39; Rawles, 47). In the case of the chick, the skin from such intracoelomic wing grafts has even been transplanted to the back of newly hatched host chicks, in which case it has continued to grow normally and produce typical wing plumage in normal succession throughout the life span of the bird (Rawles. 44). Moreover, isolates of the lateral plate is derived from the dermatome, the above experimental results are of considerable in- terest, for, under the conditions of the ex- periments, the dermis found in the grafts could not have been derived from derma- tome material. Inasmuch as the neural crest was also excluded from the grafted tissue, it too may be ruled out as a contributing fac- tor in the formation of the dermis of the grafted regions, contrary to the suggestion of Raven (731, °36) for the urodele amphibi- ans. It would appear from the experimental evidence, then, that the dermis of the limbs, flank, and ventral surface of the body is a derivative of the mesoderm of the somato- pleure. There is no doubt that the dermatome of the somite contributes to the formation of the dermis of the dorsal and dorsolateral body regions. Grafts of somites, including ectoderm and entoderm, without the adja- cent lateral plate bear this out (Straus and Rawles, *53). Furthermore, results obtained from marking, with finely powdered blood carbon, the mesoderm of various portions of SKIN AND ITs DERIVATIVES the somites between the limb regions of 2% day chick embryos, have shown clearly that in normal development the somite material does not migrate into the ventrolateral and ventral regions of the body wall. These cor- roborative results strengthen the conclusion that the dermis arises from mesoderm in contact with and subjacent to the skin ecto- derm. It is highly probable that the develop- ment of the dermis is a result of contact relationships between ectodermal and meso- dermal cells. The origin of the dermis in the amphibian has not been so clearly established. Again, the problem centers on whether the mesen- chymal cells subjacent to the skin ectoderm have a localized source, such as the derma- tome or neural crest, or whether their source is more generalized. That the dermatome is not the source of the dermis of lateral and ventral regions is indicated by the results of Detwiler (37a). When four successive brachial somites were excised on one side of Amblystoma embryos, and replaced by unsegmented, lateral mesoderm and _ over- lying ectoderm, the dermis of the skin on the operated side appeared to be as com- pletely developed as that on the normal side. The dermatome was thus eliminated as a source of the dermis of the lateral and ven- tral regions at least. Among the amphibians, two possible interpretations are evident. The dermis arises either directly from the ex- ternal layer of the somatic mesoderm of the lateral somatopleure (or ventral myotomic growth) or from the neural crest cells as Raven (31, °36) first suggested. According to Raven the greater portion if not all of the dermis of the trunk arises from neural crest cells. His interpretation was based on the interchange of prospective neural crest of the trunk region between embryos of dif- ferent species of urodeles (Amblystoma mex- icanum and Triturus taeniatus or T. al- pestris). Owing to species differences in nu- clear size and quantity of yolk granules he was able to follow the migration of neural crest cells into the dermis, i.e., into a posi- tion between the myotomes and skin ecto- derm. Detwiler (°37b) followed neural crest cells which had been stained with Nile blue sulfate to similar positions suggesting their contribution to the dermis or corium. Holt- freter (35) and Stearner (’46) have pre- sented further evidence in support of this hypothesis. Although these and other ex- periments leave little doubt that some of the neural crest cells migrate laterally be- tween the somite or myotome and the skin 501 ectoderm, it is not altogether clear whether these are precursor pigment cells or mesen- chymal cells of the dermis or a mixture of both. The problem needs reinvestigation in view of the demonstration for the chick and mouse that trunk and limb dermis arises in- dependently of any contribution from the neural crest. At present the most plausible view appears to be that the mesenchyme just beneath the skin ectoderm of the trunk is not restricted to any one site of origin. The mesenchyme of the head is largely if not entirely of neural crest origin; there- fore, it is quite probable that cells derived from the latter give rise to the dermis of the head skin. It is now known that the neural crest exhibits marked regional dif- ferences in its potentialities (Hoérstadius and Sellman, ’46; Niu, °47). Time of Localization or Specialization of the Ectoderm. It has been clearly shown by the methods of experimental embryology that the parts of a developing system possess greater developmental potentialities at an early stage than at any later period. The ectodermal cells of an amphibian gastrula, for instance, are potentially capable of dif- ferentiating into almost any cell type. At this stage their prospective fate has not be- come irrevocably established and the course which they follow in differentiation depends upon their reactions to inductive stimuli received from the region in which they hap- pen to be placed. During the process of gastrulation, however, the skin ectoderm ac- quires a certain amount of autonomy or capacity to self-differentiate which enables it to differentiate in a particular direction independent of stimuli from the surround- ing tissues. In other words, specialized prop- erties have now been acquired which govern its further histological differentiation (see Section VI, Chap. 1 for the evidence on change in potency during gastrulation). By the end of the gastrulation stage the ectoderm has become more or less a mosaic of areas or fields with differences in intrinsic organiza- tion. Although regional specialization has begun it is not yet final—changes are still possible. The time when the prospective fate of the areas becomes definitely established varies among the different organ-specific areas, and there are also time differences among species. Polar Organizaton of the Skin Ectoderm. The existence of a polarity in the skin ecto- derm of early amphibian embryos is evi- denced by the orderly beat of the cilia in a predominantly anteroposterior direction. 502 Experimentation has shown that the time in development when this polarity becomes established varies somewhat among the dif- ferent species. In many of the anurans and in the axolotl, polarization appears to be acquired during gastrulation, shortly after the round yolk-plug stage and thus consid- erably in advance of the actual appearance of the cilia (Woerdeman, ’25; Tung and Tung, °40). If a rectangular piece of skin ectoderm is excised from a blastula or young gastrula and reimplanted after a ro- tation of 180 degrees, the cilia arising from this area beat in the normal anteroposterior direction. But if the same rotation operation is performed later in gastrulation, i.e., at the circular blastopore stage and afterwards, the ciliary beat of the particular area is reversed, indicating that polarization had occurred prior to rotation. In Amblystoma punctatum similar experiments by Twitty (28) have shown that polarity is established in the ectoderm somewhat later in develop- ment, during the closure of the neural folds, hence about the time that the cilia would normally appear. Cilia developing on areas rotated after the closure of the neural folds were found to beat in opposition to those of the surrounding epidermis. About the time that the direction of the effective stroke of the cilia is established in Amblystoma other changes involving polarity are also taking place, such, for example, as the outgrowth of the placodes of the lateral-line system (Stone, °33) and the anteroposterior polari- zation of the ear rudiment (Harrison, ’36). At this time the ectoderm can no longer be turned inside out and develop into normal skin (Luther, ’34). All of these changes are indicative of some fundamental transfor- mation within the cells of the ectoderm. That polarization is imposed upon the skin ectoderm by the underlying entomesodermal tissue is evidenced by the fact that whenever ectodermic vesicles are formed, e.g., as a re- sult of faulty healing-in of the transplant, the ciliary beat of such areas is uncoordi- nated (Luther, °34). The partial exo-gastru- lae obtained by Holtfreter (’33a,b) afford an even more beautiful demonstration of this point. He found that the ciliary beat was regular and polarized in the portions of the ectoderm underlaid by entomesoderm, but was irregular and chaotic in regions not underlaid by this “organizing” tissue. Once ciliary polarity has been induced in the skin ectoderm, it in turn becomes capable of inducing polarity in younger, unpolar- ized ectoderm if brought into direct contact SPECIAL VERTEBRATE ORGANOGENESIS (Tung, Tung and Chang, °48). These in- vestigators have also presented evidence that the induction of polarity in the ectoderm involves a chemical interaction. Surface-interior differences in cells of the skin ectoderm have been demonstrated in the chick embryo by implanting small iso- lates of head skin ectoderm into the meso- derm of a wing bud (Willier and Rawles, 40). If the isolate becomes completely em- bedded in the surrounding limb mesoderm, the epithelial character of the ectoderm dis- appears and its constituent cells intermingle with and become indistinguishable from mesodermal cells of the wing bud. If, as sometimes happens, the isolate rolls up to form a vesicle with its original outer sur- face facing the cavity and the inner surface contiguous with wing mesoderm, the ecto- derm will maintain its epithelial character and differentiate into epidermis. In order to develop normally the original outer surface of the skin ectoderm must be free of cellular contact. The free or outer surface of the ectodermal cells appears to be incompatible with and to resist fusion with mesodermal cells. Source of the Pigment Cells of the Skin. Pigment cells variously designated as mel- anophores, chromatophores, pigmentophores, dendritic cells, et cetera, are common and distinctive components of the skin of all ver- tebrates, including man. They are found in both epidermal and dermal layers and, also, particularly in the lower vertebrates, in perineural and perivascular layers and in the peritoneal lining of the coelomic wall. Controversy over the site of origin of these highly specialized, branched cells has existed for nearly a century. Although many hy- potheses have been advanced, the most gen- erally accepted view up until 1934 held that the pigment cells were modified connective tissue cells, or at least originated from the mesoderm. The first suggestion that the neural crest might be the source of these cells is traceable to the observations of Bor- cea (09) in teleosts and Weidenreich (712) in amphibians. Within this same period Har- rison (710) in his studies on nerve regenera- tion in vitro found pigment cells in cultures of frog spinal cord and clearly predicted their origin from the neural crest. Many years elapsed before this problem received serious attention and was systematically in- vestigated by experimental methods (Dn- Shane, °35). The proof now firmly estab- lished by numerous workers for many species of amphibians (see reviews by DuShane, SKIN AND ITs DERIVATIVES 43; Rawles, ’48; Horstadius, 50) is based principally upon the following evidence: (1) extirpation of the neural folds (includ- ing the primordia of the crest) of neurulae results in a total absence of pigment cells from the operated trunk region; (2) isolated neural folds produce numerous pigment cells when cultured in vitro or (3) when trans- planted to the flank of another embryo of the same or of a different species. In the latter case, the pigment cells which develop in the foreign host are always of the donor type. While the entire neural crest is po- tentially capable of producing all of the various types of pigment cells—melano- phores, xanthophores, guanophores—found in the body of an amphibian, it nevertheless does exhibit regional differences as regards the number of these cells produced. The great majority appear to come from the crest of the trunk region (Niu, ’47; Twitty, ’49). In birds, as in amphibians, the neural crest origin of the pigment cells (melano- phores) has been unequivocally established by a variety of experimental results. The evidence briefly summarized is as follows: (1) explants including the neural crest pro- duce typical melanophores when cultured in vitro (Dorris, 38); (2) melanophores dif- ferentiate from isolates containing the crest, or any of its migratory cells, when grafted to the embryonic coelom (Eastlick, 39; Ris, 41) or to the early limb bud (Dorris, °39; Willier and Rawles, *40). Similar isolates without the crest never produce pigment; (3) neural crest cells grafted between em- bryos of different species of fowl, or even between wild birds and fowl, invariably pro- duce melanophores of the donor type whose activity in the feathers of the host is re- corded in the form of a typical donor colora- tion or pattern. Among mammals, the mouse embryo has so far been the only representative in which the origin of the pigment-forming cells has been tested experimentally. The results are clear cut, however, in showing that only those tissues containing prospective neural crest, histologically recognizable neural crest or cells migrating from the neural crest, can produce pigment in grafts (Rawles, °47). Thus the evidence has clearly established the origin of pigment cells of amphibians, birds, and mammals from the neural crest. A similar origin is indicated also for the lamprey (Newth, 51) and for the bony fishes (Borcea, ’09; Lopashov, °44; Orton, *53). It is highly probable that the neural crest or its equivalent, the dorsal neural border, is 503 the source of pigment cells in all of the vertebrates. Further proof would be welcome for some of the lower forms. Migration of Pigment-forming Cells (Mel- anoblasts) into the Skin. From their locus of origin in the neural crest, prospective pig- ment cells migrate gradually to all regions of the body of the embryo. During this period of dispersal they cannot be distin- guished with certainty either morphologi- cally or histologically from the other em- bryonic cells with which they are associated. Yet, by means of appropriate transplanta- tion experiments, it has been demonstrated clearly that they have reached all body re- gions of a chick embryo by the fourth day of incubation and all body regions of a mouse embryo by the twelfth day of gestation— long before there is any visible sign of pig- mentation. As development proceeds and bio- chemical conditions become suitable for the synthesis of melanin, these “colorless” cells begin their characteristic differentiation and henceforth are readily distinguishable from surrounding cells. Observed differences in the distribution patterns of these cells after the differentiation of melanin pigment gran- ules are not, however, necessarily a _ reflec- tion of corresponding differences in their original dispersal from the neural crest. Recent evidence indicates that the character- istic longitudinal stripes of certain larval amphibians are formed by a secondary re- arrangement of pigment cells originally more widely scattered over the lateral sur- faces of the somites (Rosin, ’43; Twitty, °45). Many prospective pigment cells remain un- differentiated (colorless) for long periods— until after metamorphosis in amphibians— and some undoubtedly become located in positions unfavorable for the synthesis of pig- ment. Both the reactivity of the pigment cells and the biochemical properties of the skin vary not only with the species, but even regionally within one individual. The mechanism by which precursor pig- ment cells (melanoblasts) reach the skin and other locations is by no means com- pletely understood. While it is evident from in vitro and other studies that these cells are capable of independent movements, much of their migratory activity is unques- tionably dependent upon contact relationship with certain other strains of cells and tissues. The fact that pigment cells are found pri- marily along surface membranes (basal layer of the epidermis, parietal membranes, etc.) is significant. The possibility also that their final distribution may be affected some- 504 what by morphogenetic movements and by growth movements of their tissue substrate should not be overlooked. The direction and paths of movement of the melanoblasts concerned with skin pig- mentation do not appear to be at random. Experimental evidence indicates that they follow more or less predetermined routes, proceeding dorsoventrally in the mesen- chymal tissues subjacent to the skin ecto- derm. This would seem to imply that their movement is somehow directed by proper- ties intrinsic to the skin and also by the interface between the ectoderm and _ the prospective dermis. The association of cells and tissues of distinctly different types pre- supposes some kind of surface compatibility or “affinity.” The migratory movement of the precursor pigment cells, therefore, could be dependent upon their specific interaction or contact relationship with cells of the der- mis. On this supposition, movement would continue until the contact relationship be- tween the combining elements reaches some sort of equilibrium (Holtfreter, 39; Tyler, "47; Weiss, ’47). Interesting experimental evidence in favor of selective association of cells and tissues of specific types has been obtained recently by Weiss and Andres (’52). They injected disso- ciated embryonic cells of the chick, includ- ing melanoblasts which served as markers, into the blood stream of host chick embryos. Since these injected cells became scattered at random throughout the body of the host, it was possible to determine whether or not any “selectivity” governed the definitive lo- cations of the pigment cells. The results demonstrated conclusively that donor mel- anoblasts proliferate profusely and synthe- size melanin granules of the color and shape characteristic of their genotype, only in lo- cations in the host identical to those in which they would normally have developed pig- ment in the donor individual. Never were they found in unusual cell and tissue associa- tions. Certain results obtained from grafting skin in fowl indicate that the invasion of melanoblasts is controlled by the skin and feather germs. When, for example, an area of skin, experimentally deprived of its nor- mal source of pigment cells, is grafted at hatching to a chick host of similar age, melanoblasts from the surrounding regions of the host skin migrate freely into the graft and establish themselves permanently (Rawles, 44). Such an invasion of melano- blasts does not take place when an area SPECIAL VERTEBRATE ORGANOGENESIS of normal skin containing its full comple- ment of melanoblasts is grafted simiiarly (Danforth and Foster, °29). It would appear, then, that invasion does not take place if a state of equilibrium has already been at- tained between the tissues of the skin and the melanoblasts. This phenomenon has been interpreted by Willier (48) to mean that a constant ratio has been established between the number of melanoblasts and the cells of the skin. The number of melanoblasts, ac- cording to this view, is limited not by a self-limitation of their capacity for multi- plication, but rather by the cell community (skin). Such a constant ratio may be tem- porarily thrown off balance by an active regenerating feather papilla in which special conditions are set up favoring the invasion of some of the melanoblasts from the dermis or its specialized unit, the dermal papilla, into the epidermal region (collar) which gives rise to the feather parts. As this in- vasion of melanoblasts into the regenerat- ing feather parts takes place, other melano- blasts of the dermal regions multiply to restore again the constant ratio. Thus a mechanism is provided for maintaining this constant relationship between the pigment cells and the feather cells throughout the life span of a bird (Willier, ’52). REGIONAL SPECIALIZATION OF THE SKIN In the higher vertebrates the epidermis enters into a closer and more intimate rela- tionship with the dermis or corium than in the lower vertebrates. The early smooth con- tour between the two, such as exists perma- nently in the lower forms, is lost owine principally to the formation of folds or papillae which project into the epidermis and alternate with similar downward pro- jections from the epidermis. The extent of development of the papillae varies greatly in the different body regions. On the palms and soles they reach their greatest height. Structural Differences. Studies on the de- velopment of the skin in the various types of vertebrates show that it is the ectodermal portion which undergoes characteristic struc- tural modifications to fit it for the carrying out of special functions. Increase in epider- mal surface is brought about by both evagi- nations and invaginations which develop in many different ways. Arising as external processes are the so-called epidermal ap- pendages, scales, feathers, hairs, nails, and teeth; arising as invaginations are a va- SKIN AND ITS DERIVATIVES riety of glands—mucous, cement, and poison glands of the lower vertebrates; sebaceous, sweat, and milk glands of the higher ver- tebrates. A cursory examination of the hair coat of a mammal or the feathers of a bird re- veals in various regions of the body—head, breast, back, tail, et cetera—a striking dif- ference in the size, shape, and structure of the epidermal outgrowths. Even one single region, the head of a bird for example, may display various adaptive modifications of the contour feathers to form ornamental plumes, ear coverts, facial bristles, and eye- lashes. Among feathers the variety of form is almost limitless, yet each is a modification of the same fundamental structure. Of in- terest also is the fact that one and the same hair or feather papilla of a particular re- gion may produce a succession of hairs or feathers which display morphological and color differences pertaining to definite stages of the life history. Not only are there marked regional dif- ferences in the epidermal outgrowths, but also variations in the skin proper. Quantita- tive histological studies of mammalian skin have demonstrated regional differences in erowth of the epidermis (Loeb and Haven, 29). Straus’s study (50) of the microscopic anatomy of the skin from ten selected re- gions of a female gorilla reveals significant differences in structure at all of the various regions examined. The skin is thickest over the back and thinnest over the chest. The epidermis reaches its greatest thickness on the palms and soles, the corium its greatest thick- ness on the back. In general the _ total thickness of the skin is a reflection of the thickness of the corium. The above findings hold for mammals in general, including man. Spectrophotometric analysis of living hu- man skin (Edwards and Duntley, ’39) has shown significant differences in the distri- bution of the pigments (melanin, hemo- globin, carotene) responsible for differences in the color of the skin in the various body regions. Oxyhemoglobin is especially abun- dant in regions of the skin where the ar- terial blood supply is rich and where for the most part the dermal papillae are high. Carotene is more abundant in regions where the stratum corneum is thickest. Female skin contains less hemoglobin and melanin but more carotene than male skin. Minor sex differences in distribution of the pigments also occur. Functional Differences. Differences in struc- 505 ture in the various body regions may be correlated often with integumental differ- ences in function, but there are also in- trinsic physiological differences among cells which are revealed only by their method of response to certain stimuli such as hormones, vitamins, light, and temperature. In man, regional differences in photosensitivity of the skin are believed to be dependent to a great extent on variations in thickness of the horny cell layer and upon the development of the skin capillaries. The horny layer func- tions as a superficial filter absorbing some of the light waves before they reach the liv- ing layers of the epidermis. But it is known that the quantity of pigment, the age and sex of the individual, and the season of the year are important factors, also, in deter- mining the regional and individual differ- ences in skin photosensitivity (Ellinger, ’41). A striking demonstration of the existence of intrinsic, qualitative differences among cells of the skin of different body regions is afforded by amphibians at metamorphosis. Under the influence of thyroid hormone, the skin of the tail of a frog tadpole, for in- stance, responds by undergoing degenerative changes while the response of the adjoin- ing skin of the trunk is one of proliferation. Histologically the two areas of skin response are sharply defined and show no transition. The specific response of these tissues is not altered by heterotopic transplantation (Lindeman, ’29). The dependence of certain sexual plumage types upon gonadal _ hor- mones in fowl is equally striking. Extensive studies of morphological and color changes in the feathers of the Brown Leghorn have shown that regional differences in the threshold of response to female sex hormone and to thyroxin are dependent on local dif- ferences in the growth rate of cells of the individual feather germ (Lillie and Juhn, 32). In many animals melanin pigment is produced by melanophores in certain local- ized regions of the integument in response to male hormone stimulation. Among cases described are the bill of the sparrow (Keck, 34); the lores, roof and floor of the mouth of the night heron (Noble and Wurm, ’40): the dorsolumbar spots of the hamster (Kup- perman, 744); the scrotum of the 13-lined eround squirrel (Wells, °45); the sternal spot of the Australian opossum (Bollinger and Hardy, °45). In the human female excess melanin is deposited by melanophores in the skin of certain localized areas (eyelids, nipples, areolae, linea nigra) in response to increased amounts of estrogenic substances 506 present in the body during pregnancy (Davis, Boynton, Ferguson, and Rothman, "45). There is much evidence that male hor- mone affects the vascularization of certain localized areas of skin. The highly vascular comb of fowl has long been used as a trust- worthy and measurable indicator of the presence of male hormone. The reddening of the sex skin in certain primates and the skin of the legs of the night heron are simi- lar examples of the response of local cu- taneous blood vessels to sex hormone stimu- lation. Hamilton (’48) emphasizes the im- portant role of the sex hormones in regulat- ing the vascularization of the human skin (complexion) ; and spectrophotometric analy- sis has shown that the pallor of the white eunuch is mainly the result of lack of hemo- globin in the cutaneous blood vessels. Regional Differences of the Skin Established Early in Ontogeny. Very little is known about the time of origin of regional specificity in the skin ectoderm, and this point has not been systematically tested. In a series of studies on the development of the dorsal fin in Amblystoma embryos Bodenstein (752) found some evidence of regional differentials in the response of the ectoderm of young tail-bud stages. If flank ectoderm, for exam- ple, is transplanted to the dorsal median region, no fin is produced; but if dorsal median ectoderm is transplanted to the flank, over the somites, a fin is produced. In the absence of more pertinent data, however, it can only be assumed that at some time dur- ing the latter part of embryogenesis regional specificity of the skin, which is clearly de- monstrable in the early larvae, becomes established. Experiments with a wide va- riety of larval amphibians have demon- strated that skin from various regions of the body—tail, back, flank, limb buds—trans- planted to different locations on the body of the same individual or of another individual of the same or of a different species, retains its individual characteristics. The transplant, moreover, metamorphoses in the manner typical of the region of the animal from which it came, even with regard to indi- vidual details of spotting pattern (Uhlen- huth, *17; Weigl, 13; Cole, ’22; Reis, °30; and others). Recent experiments have shown that the specific central reflex relations of sensory nerve fibers entering heterotopically located grafts of cornea (Weiss, *42) and skin (Miner, 51) depend upon the sites of origin of the innervated tissues and not upon the origins of the sensory nerves them- SPECIAL VERTEBRATE ORGANOGENESIS selves. Thus, in larval newts, tactile stimula- tion of cornea transplanted to the site of the ear or nasal organ will elicit the typical lid-closure reflex that is obtained by touch- ing the cornea of a normal eye. The in- trinsic capacity of the skin tissue for the differentiation of specific sensory nerve end- ings has been strikingly illustrated by ex- changing areas of skin between the bill and leg of the duck (Dijkstra, °33). The skin of the duck’s bill normally contains highly spe- cialized sensory nerve endings, the cor- puscles of Grandry and Herbst, found no- where else on the body. When an area of bill skin is transplanted to the leg it be- comes invaded by leg nerve fibers and the characteristic corpuscles appear. If leg skin is grafted to the bill, however, no such sensory endings develop, despite its innerva- tion by normal bill nerves. That the capacity of the skin to respond to hormones is acquired early is evidenced by the fact that skin from young amphibian larvae transplanted to much older larvae attains adult characteristics synchronously with the host, i.e., metamorphoses earlier than it normally would have done if left undisturbed (Uhlenhuth, °17; Weigl, °13); and metamorphic changes in local areas of larval anuran skin can be initiated much earlier than they would normally occur by implanting pellets containing thyroxin under the skin (Kollros and Kaltenbach, 752). During metamorphosis Woronzowa_ (32) found that different regions of the skin of Amblystoma tadpoles clearly showed dif- ferences in their threshold of response to a given quantity of injected hypophyseal hormone. In fowl the marked regional differences in the morphology and pigmentation pattern of the adult plumage are directly related to the order of origin of the feather papillae early in embryonic life. Holmes (°35) found that the feather papillae composing the in- dividual plumage tracts arise in a definite time and space order. In a breast tract, for instance, the first longitudinal row of papil- lae arises parallel to the anteroposterior axis of the body at a position off-center in rela- tion to the prospective tract as a whole, becoming the sixth row in mediolateral or- der when the tract is completed. After the formation of the papillae composing the pri- mary row, other rows arise in sequence to the right and left of the first, and always definitely oriented with reference to it, until nine parallel rows are formed. The relation of this definite time and SKIN AND ITS DERIVATIVES space order to asymmetry in the adult plum- age of the Brown Leghorn has been clearly shown by Juhn and Fraps (34a, b). These investigators found that the symmetry re- lations along transverse rows of feathers of the breast tracts follow an orderly distribu- tion with respect to a secondary axis lying approximately at the sixth row, counting eed eae Ss 507 when the characteristics of the humeral tract are becoming established has come from 180 degree rotation of small areas of ectoderm and underlying mesoderm of the dorsal sur- face of the wing buds of 3- to 4-day chick embryos (Saunders, 50). While the resulting wing plumage was normal in some cases, showing complete regulation of the area, Fig. 193. Sections showing stages in development of down feather. Iron hematoxylin, 8u. A, Early stage in feather formation, breast tract, 9-day embryo. Note aggregation of mesodermal cells (primordium of the dermal papilla) beneath two-layered ectoderm (epidermis). * 157. B, Later developmental stage from same feather tract. Note increased proliferation of mesodermal cells causing ectoderm to protrude beyond the skin level, forming the characteristic feather buds. & 157. C, Feather buds of breast tract of 11-day embryo. Note increase in length of feather bud and the thickening of ectoderm to form barb-ridges. < 75. laterally from the mid-line of the bird. The increasing degree of asymmetry with refer- ence to this secondary axis is in accord with observations that pigmentation patterns, nor- mal or hormone-induced, limited to one vane-half of feathers of the breast tract show the relation of mirror images within each tract. Furthermore, the degree of asymmetry in such feather patterns increases as the lat- eral margins of the tract are approached. Thus the degree of asymmetry of the adult plumage is found to increase with distance from row number 6 (the first row to arise) and to correspond with the time of origin of the various rows of feather papillae com- posing the tract. Some indication of the time in ontogeny there were a sufficient number of cases in which abnormalities occurred in the distri- bution and orientation of the feathers to indicate that the tract characteristics of the rotated area were partially established. Role of the Mesoderm in Epidermal Spe- cialization. It would be of considerable in- terest to know whether the highly specialized epidermal outgrowths, feathers and_ hairs, arise in situ through some localizing factor in the ectoderm itself or whether they are produced by reaction with the underlying mesodermal tissue mass. In other words, are these specialized skin derivatives the prod- ucts of embryonic induction? Experimental embryology has demonstrated that, in addi- tion to the central nervous system, inner ear 508 and lens, many other ectodermal organs, such as gills, balancers, fins, teeth, hypoph- ysis, etc., are dependent upon influences exerted by other tissues for their differen- tiation. Inductive influences, as Weiss (735) has pointed out, were originally thought of ae SPECIAL VERTEBRATE ORGANOGENESIS strating embryonic induction are extirpation and transplantation. To a certain extent both have been applied towards testing the in- ductive faculties of the tissues comprising the feather germs. Before considering the experimental evidence, a few facts regarding Fig. 194. Sections through the skin of saddle region of a normal Lakenvelder pullet, 3 weeks after hatching, showing the structure of the fully differentiated skin and definitive feathers developing in situ. Delafield’s hematoxylin, 10u. A, * 75; B, x 100. b, Cells of the barb system; ce, collar (thick ring of embryonic feather- forming cells); d, dermis or corium; dp, dermal papilla of feather; e, epidermis; em, erector muscles of feather; f, follicle cavity; 7m, striated muscles; p, pulp of feather; r, rachis. as being much more specific than they are considered to be at the present time. The term induction has now been extended to include a variety of “organizing” influences from unspecific activation to the very specific organization of typical patterns in space and time. Inductive phenomena are by no means physiologically uniform. The Feather. The two classic methods that have been employed in general for demon- the origin of feather germs should be re- called. The first indication of the site of a prospective feather is seen in the meso- dermal portion of the dermis or corium, at approximately the fifth day of incubation, in the form of an aggregation or condensa- tion of cells immediately beneath the thin, two-layered epidermis. This condensation is the primordium of the dermal papilla and precedes any visible epidermal response, as SKIN AND ITS DERIVATIVES indicated by the lack of alteration in the epithelial cells. Rapid growth and prolifera- tion of the dermal cells soon cause the over- lying epidermis to protrude, forming the characteristic protuberances or feather germs comprising the various feather tracts (see Fig. 193A, B,C). Before hatching, each der- mal papilla sinks beneath the surface of the skin in a tube-like follicle lined with epi- dermis. Later proliferation of the epidermal cells overlying the dermal papilla gives rise to the embryonic region from which the parts of the definitive feather arise. For in- formation concerning the structure of the definitive or permanent types of feather and its relationship to the fully differentiated skin, see Figure 194A and B. The effect of removing the ectoderm, from the upper surface of a 3-day wing bud, on the origin of feather germs of the chick has been tested by Saunders and Weiss (750). In the absence of ectoderm the prospective dermis is found to be incapable of organiz- ing either a typical corium or dermal papil- lae. A causal interaction of the inductor type between the ectoderm and subjacent meso- derm in feather germ formation is suggested by these experiments. Evidence of a more crucial nature has come from some recent transplantation experiments of Cairns (751), in which he was able to induce feather germs in a region of the wing skin ectoderm which normally produces no feather germs by implanting mesoderm from the leg bud of 4-day embryos. The most conclusive evidence that induc- tive principles are operative in feather de- velopment has come from the beautiful and systematically executed experiments of Lil- lie and Wang (41, 44) and Wang (’43) on the papillae of regenerating feathers of the adult fowl. When a feather is shed naturally through the process of molting, or when it is plucked, the dermal papilla, covered by a thin layer of epidermal (‘regeneration’) cells, is left behind in the base of the tubu- lar follicle. From this epidermal component is formed a thick ring of embryonic cells, the “collar” (Lillie and Juhn, *32), which gives rise to all of the epidermal parts of the regenerating feather. If the dermal papilla is removed, a feather is never formed from the epidermal cells of the follicle wall which grow over the site of extirpation. The dermal papilla is, therefore, essential for a feather-forming epidermal response. By means of a variety of skillful transplan- tations, Lillie and Wang succeeded in dem- onstrating clearly that the mesodermal or 509 dermal portion of the feather papilla func- tions as a feather “inductor” and determines the symmetry and orientation of the result- ing feather. The specificity of epidermal re- sponse, i.e., the specific type of feather in- duced, whether breast or saddle, was found to be dependent upon the tract origin of the overlying epidermis. For example, a saddle feather papilla from which the epidermal cells have been entirely removed, trans- planted into an “empty” breast feather fol- licle, induces a feather of the breast type from the epidermal cells of the breast fol- licle wall which grow over it, and vice versa. Breast or saddle papillae from which the epidermal cells have not been removed retain their specificity when transplanted into empty follicles of either tract. The Hair. Hairs, like feathers, are highly keratinized epidermal outgrowths which arise also in a definite time-space sequence. The first series of primordia are uniformly spaced. As noted previously for feather pri- mordia, so here, a second series arises to each side of the first and definitely oriented in relation to it, and so on in transverse rows until the number characteristic of the species is laid down. All of the hair papillae are formed during embryonic life or shortly after birth; hence the number and arrangement are the same in the adult as in the embryo. Growth of the connective tissue of the dermis later on, however, does tend to disrupt the earlier more orderly arrangement and _ to make the linear order somewhat more diffi- cult to discern. The first primordia to appear are those of the sensory hairs or vibrissae on the face around the nose and mouth. These arise very early in gestation, long before there is any indication of hair primordia elsewhere on the body surface. In their early develop- mental stages the sensory hairs differ in certain respects from the general body hairs. The future site of a sensory hair is fore- shadowed by a sub-epidermal condensation of the mesodermal cells of the prospective corium, the primordium of the dermal papilla. This precedes any appreciable change in the overlying epithelium. Rapid proliferation of the mesodermal cells raises the epidermis, forming rows of papilla-like elevations easily observable in surface view (Fig. 195A). These elevations (Hockerchen) have been described by many of the early investigators for various mammals—cat, sheep, pig, rabbit. In sections they are strik- ingly similar to down feathers (cf. Figs. 193A and 195A). This initial stage is of 510 short duration. Soon tongue-like thickenings of the basal cells of the epidermis grow downward into the underlying corium, form- ing the follicle with the dermal papilla situated in its base. Other hairs of the face, head, and trunk which arise later do not ete . os pore eet ae es my ek ee - ee re A » pen ee saa ae 7 Sess ee : PE ee PP a be = a ~3 4 é ‘= eg I = - * +. ee é © + S Oran ee eae ek Shree FAS RSE chi one : oN Shee eld ee ; ot «ws ; . . Ser: % “5 , - . Ce al wt aa Se yee Nate aa? os Shay ae eR a Os oP ates oe sé. ** oe Pg a Fee, Ce} fe 5g Sts 8 Oe e* ws a, ee » + ey » wei Ps TAG * ¢ 3 nme a oe ~ a tees, Cae J me asd . _ + s a ee » e =~ * e - 6 & 3 * th © “ % 4 * — * ™, * a, ey : > * ~ mn ~ yy = - a - Kor + * ae: “es 4 = ve = . ‘ i ae 2 - i x < t— o@~ ~~ _e ss Fig. 195. Sections illustrating early stages in hair formation. Iron hematoxylin, 8u. « 210. A, Begin- ning of a sensory hair (vibrissa), 22 mm. pig em- bryo, lower jaw. Note aggregation of mesodermal cells (primordium of dermal papilla) and elevation of two-layered, overlying ectoderm. B, Beginning of a body hair, 17-day mouse embryo. Note increase in ectodermal cells, to form the hair primordium, and aggregation of underlying mesodermal cells to form primordium of dermal papilla. C, Further down growth of ectodermal cells in the formation of the hair primordium. Same embryo as B. show such pronounced surface elevations of the corium and overlying epidermis. (A striking exception is the European hedgehog, whose spines develop quite like the sensory hairs of other mammals; Davies, ’89.) Each body hair begins as a minute epidermal nod- ule which forms by local proliferation of cells of the germinal or basal layer (Fig. 195A). Simultaneous with, or immediately follow- SPECIAL VERTEBRATE ORGANOGENESIS ing, the initial changes in the epidermis, a condensation of cells in the underlying mes- enchyme takes place to form the primordium of the dermal papilla. By continued growth downward of the epidermal tongue of cells (Fig. 195C), the follicle is established with the dermal papilla located in its base (Fig. 196A and B). The hair proper develops from epidermal cells covering the dermal papilla. Although the ectodermal and mesodermal components of the hair papilla have not been submitted to experimental analysis com- parable to that of the feather, the great similarities in the formation of hairs and feathers suggest strongly that the inductive mechanisms involved are similar in these two groups of skin derivatives. In fact, evi- dence in support of this view is forthcoming from Hardy’s (49) work on culturing mouse skin in vitro. She noted that no epidermal thickenings or “plugs” formed in the ab- sence of mesoderm, and, furthermore, no dermal papillae could be found in areas from which the epidermis had been removed. Normal hair differentiation was obtained from cultures in which the epidermis and dermis were not separated. Such results sug- gest strongly that an interaction of ectoderm and subjacent mesoderm is necessary for in- itiating hair formation. It would seem, then, highly probable that hairs, like feathers, are products of embryonic induction. Epidermal Ridges. Since the initiation of development of both feather and hair pri- mordia appears to be dependent upon an interaction between the ectoderm and the subjacent mesoderm, one is prompted to consider the possibility of the existence of similar inductive relationships in the forma- tion of other specializations of the epidermis, such, for example, as the epidermal ridge systems found on the under surface of the hands and feet. In all of the primates the skin of these regions is characteristically marked with fine, parallel ridges presenting a corrugated appearance. Hairs and _ seba- ceous glands are absent, but sweat glands are abundant and large. The minute details of the ridges and the very definite patterns— loops, whorls, arches—formed by them on the tips of the digits and in consistent sites on the palms and soles show regional as well as individual variation. Even the skin of a small area will show ridge details not found elsewhere on the same or any other indi- vidual. This, together with the fact that the ridge pattern in all of its detail remains un- changed throughout life, is the basis for the use of epidermal ridge patterns (der- SKIN AND ITs DERIVATIVES matoglyphics) in personal identification. The development of the epidermal ridges is intimately associated with the develop- ment of the touch balls or volar pads. These embryonic structures are definitely localized swellings or bulges found on the terminal Sila but in the deep germinal portion in contact with the mesoderm or corium, and is very likely a response to inductive stimuli from the mesoderm. The lower germinal layer begins to increase and form folds which grow downwards into the corium, Simul- Fig. 196. A, Section through head skin of newborn black mouse (C57) showing various stages in the formation of hairs. Delafield’s hematoxylin, 10”. * 157. B, Section through dorsal skin of black mouse (C57), 7 days after birth. Note fully differentiated, emerged, pigmented hairs. Lightly stained with Dela- field’s hematoxylin, 10u. « 75. segments of the digits and on the palms and soles. In the human, the volar pads of the hand are evident early in fetal life (sixth week) when the hand is still paddle-like. They become quite prominent about the twelfth to thirteenth week. Soon afterwards they regress and become relatively incon- spicuous. Ridge formation begins when the volar pads are at their peak. The first indication is seen, not on the surface of the epidermis, taneously, the surface of the corium in con- tact with the epidermis develops elevations or folds (papillae) which project upwards into the epidermis and alternate with the similar downward projections from that layer. Later the outer surface of the epi- dermis, which has up until now remained smooth, becomes raised into ridges, one cor- responding to each of those formed earlier on the lower surface. Like feathers and hairs, the epidermal SZ ridges form in a definite time-space order. On the fingertips, the first regions to show ridge differentiation, folding of the epidermis begins in the central portion of the apical pad. Subsequently other foci arise in the distal, lateral, and proximal regions. From these foci, ridge differentiation progresses in orderly sequence until the systems meet and the final pattern configuration is estab- lished. Less frequently differentiation of the ridges into the definitive pattern is com- pleted by extension from a single focal cen- ter on the apical pad (Bonnevie, ’27, ’29; Cummins and Midlo, ’43). That there is some relationship between pattern type and the de- gree of elevation and the contour of the volar pad is generally agreed upon. By the nineteenth week the pattern is permanently established in the human fetus in all of its minute detail. Ridges broaden and lengthen to keep pace with the growth of the hands and feet, but no new formations occur (Hale, °49). In the absence of experimental evi- dence there is no actual proof that the epi- dermal ridge patterns are the products of embryonic induction, but the fact that the mesodermal substratum is necessary for the regeneration of normal ridge patterns would appear to strengthen this point of view. Mesodermal Substrate Essential for Regen- eration of Epidermis and Cornea. The im- portance of the mesodermal substrate in the regeneration of normal skin epithelium in the adult human has been nicely demon- strated by Bishop (45). By removing skin from the forearm to various depths it was found that a portion of the papillary layer must remain to insure the regeneration of normal epidermis. Scar formation resulted when removal was sufficiently deep to in- clude the reticular layer and the bases of the hair follicles. It has also been pointed out by Cummins and Midlo (43) that wounds, burns, etc., produce no permanent effect on the epi- dermal ridge patterns of the volar surfaces unless the injury is deep enough to destroy the dermal papillae, in which case scar tis- sue is then formed. Since the epidermis of the palms and palmar surface of the fingers reaches a thickness of about 0.8 mm., it fol- lows that tissue damage to a depth of about 1 mm. would be necessary to prevent nor- mal regeneration of the characteristic ridges of the fingers and palm. It is generally known that skin grafted from one region to another of the same individual retains its original characteristics. In this respect the ridged skin of the volar surfaces offers no SPECIAL VERTEBRATE ORGANOGENESIS exception. When it is considered that a skin graft normally includes the papillary layer of the dermis, the retention of specificity of the graft is readily understandable. Further evidence of a necessary interaction between dermal and epithelial factors in regeneration is shown by the studies of Maumenee and Scholz (48) on the mam- malian cornea. Epithelial cells from the conjunctiva, which migrate over areas de- nuded of cornea, do not become transformed into typical corneal epithelium unless the underlying stroma is normal. INTEGUMENTARY PATTERNS Morphological Patterns. The origin of spe- cific skin patterns, as exemplified by hair and feather direction ana by the arrange- ment of epidermal ridges, presents many interesting developmental problems which are by no means fully understood. As pointed out by Wright (49a), the formulation of principles of gene action in relation to morphological pattern is of the greatest im- portance in relating genetics to the physi- ology of development. While certain broad generalizations have long been apparent, there is a definite need for a systematic study of the action and interaction of genes in the formation of specific patterns. It is well known that skin patterns are established in their permanent form early in ontogeny. The direction or slope of the hair, for instance, is recognizable imme- diately after the first indication of the hair primordia. Causal factors, therefore, must be looked for early in embryonic life. In order to identify such factors and to analyze their role in the origin, growth, and differentia- tion of the definitive pattern, direct experi- mental evidence is essential. Although nu- merous experiments have been done to ana- lyze the factors concerned with hair di- rection in a variety of mammals, the ma- jority have been carried out after the hair primordia were established and have given negative or inconclusive results. Some recent experiments of Kiil (’49) on newborn rats, at a time when the hair primordia are in the processes of development, appear to be the most decisive. By observing tattoo marks that penetrated the upper and lower layers of the skin of the tail and by excising pieces of skin from various regions including the ventral side of the neck where two natural whorls occur, data were obtained which in- dicate that the pattern of organization of the skin is primarily a phenomenon of dif- SKIN AND ITS DERIVATIVES ferential growth within the layers of the skin itself, and that divergent whorls and rosettes are controlled from their centers by excess growth in the outer layer of skin. In lieu of experimental data, the study of the development of epidermal ridge patterns of anomalous hands and feet of human em- bryos has led Cummins (’26) to a similar ex- planation. In many respects hair arrange- ment is similar to dermatoglyphic configura- tions. Regions in which hairs slant uniformly in one direction are comparable to open fields of ridge pattern, while irregularities of hair direction localized at the points of juncture of several different hair slants cor- respond to triradii. The genetic investigations of Wright (49a, b; °50) on the guinea pig have contributed importantly towards an interpretation of the relation of genes to the development of pat- tern. By means of extensive breeding experi- ments he has shown that three genes, R, M, and St, exert major effects on hair direction in the guinea pig. A schema of the general physiological processes leading to local al- terations in skin growth which are due to these genes and their interactions has been presented (Wright, °50, p. 59). Wright’s studies of the genetics of normal and ab- normal growth patterns of the guinea pig have led to the view that the specificity of gene action is always a chemical specificity —the production of enzymes which guide metabolic processes along particular chan- nels. The development of any morphological pattern is, according to this view, a chain of reactions in which each gene reacts only in the presence of certain conditions, in part environmentally relative to the cell lineage in question, in part the result of the action of genes previously called into action (Wright, ’34a, b). Pigmentation Patterns. Experimental stud- les in recent years have contributed much towards an understanding of the numerous factors involved in the development of color patterns, especially those in which the im- portant and widely distributed melanins or granular pigments are involved. As the in- vestigations have broadened it has become more and more evident that the principles underlying pigmentary pattern formation are remarkably similar among the verte- brates in general. In analyzing the develop- ment of specific pigmentation patterns, at- tempt has been made to determine to what extent the migration, differentiation, and orientation of the melanin pigment-forming cells (melanophores) into distinctive pat- a3 terns are dependent on their intrinsic prop- erties imparted to them by their genetic constitution, and to what extent upon ex- trinsic, environmental factors, 1.e., the tis- sue substrates with which they become asso- ciated ultimately. The Pigment Cell and Pattern Formation. Owing to the well-established fact that pros- pective pigment cells (melanoblasts) arise from a transitory embryonic structure, the neural crest, and reach the tissues with which they become associated later on through their migratory activities, it has been possible to employ a wide variety of appropriate transplantation, explantation, and deficiency experiments towards clarify- ing their role in pattern formation. In fowl and other birds melanoblasts from individuals of varieties exhibiting a specific type of color and pattern have been intro- duced into feather primordia of individuals exhibiting an entirely different color and pattern. Such combinations have been ac- complished by means of various grafting methods (see Rawles, ’48). The large body of results obtained has been consistent in showing that melanophores retain their spe- cific characteristics (shape and color of the pigment granules) and react with the for- eign feather germs into which they have been introduced, to produce a typical donor color and pattern. In other words, the genetic constitution of the pigment cell governs the type of response which it manifests in a developing feather. Further proof that the color and pattern manifested in feathers is controlled by the particular assemblage of genes with which the pigment cell is en- dowed has come from grafting melanoblasts from male and female embryos of varieties of fowl showing sex-linked differences in plumage pattern, such as the Barred Plym- outh Rock and F, hybrids from crosses between Rhode Island Red males and Barred Plymouth Rock females. Without exception, melanoblasts from prospective males reacted with foreign host feather germs to produce a male plumage pattern; those from prospec- tive females, a female plumage pattern, re- gardless of the sex of the host (Willier and Rawles, 44a, b). With reference to the lack of any influence from the sex hormones of the host, it should be mentioned that among fowl, pigment cells show a differential sen- sitivity in their response to sex hormones. Melanoblasts of the Barred Plymouth Rock and the cross referred to above are represen- tative of a type which may be classified as ‘insensitive,’ their phenotypic manifestation 514 being independent of sex hormones. This is in contrast to melanoblasts of breeds like the Brown Leghorn which are “sensitive” to sex hormones and dependent upon them for ex- pressing their phenotype (see Willier, ’50). In amphibians it has been possible, also, to introduce precursor pigment cells of spe- cies exhibiting one characteristic type of pigmentation pattern into individuals ex- hibiting a different type, by exchanging seg- ments of neural folds (including the neural crest). Results agree in showing that the specific patterns of pigment cell distribution and orientation are dependent primarily upon intrinsic genetic differences in the pig- ment cells themselves (Twitty, *49). Numerous histological studies of melanin granules deposited by melanophores in the epidermal cells of feathers and hairs have revealed that within any one genotype the size, shape, and color of the granules exhibit a remarkable specificity. In the numerous cases in which melanoblasts have been trans- planted to foreign feather germs, they have always deposited in the feather cells gran- ules characteristic of their own particular genotype. Tissue Environment and Pattern Forma- tion. While it has been clearly demonstrated in birds and amphibians that the genotypic constitution of the pigment cells is a con- trolling factor in phenotypic expression of color and pattern, it is equally clear that the surrounding tissues exert a definite influ- ence on the realization of their potentiali- ties. A number of workers with amphibians have shown the influence of the epidermis, somites, and neural tube upon the melanin- forming capacity and the orientation of these specialized cells into patterns (DuShane, ’43; Twitty, 49; Dalton, *50). Evidence of the influence of the tissue substrate on melanophore pattern formation is strikingly brought out by grafting neural crest between widely unrelated individuals such as anurans and urodeles (Baltzer, ’41, ’43; Leuenberger, *42). Under such condi- tions the orientation and the distribution of the grafted melanophores are quite definitely altered. The morphology of the individual melanophores, however, is not changed. They retain their specific characteristic size, color, type of branching, etc., and can be easily distinguished from those of the host. In gen- eral the results of neural crest exchanges among amphibians have shown that altera- tions in the normal pigmentation pattern produced by melanophores of a given geno- type are progressively more pronounced as SPECIAL VERTEBRATE ORGANOGENESIS the donor and host become farther apart phylogenetically. This would appear to in- dicate that the arrangement of melanophores into a distinctive pattern is largely a particu- lar kind of response drawn forth by a par- ticular set of physiological conditions ex- isting in their tissue substrates. Changes in one bring about changes in the other; in other words, there is interplay or interaction between intrinsic and extrinsic factors. In fowl the position of the feather germ on the body—its tract location—determines the specific type of pigmentary response given by melanophores potentially capable of a range of responses, such as black and red or black and white barring, etc. This has been demonstrated convincingly by grafting such melanoblasts to feather germs of vari- ous regions of the body. The final pattern obtained is always typical of the region— wing, breast, saddle, etc.—showing that the locus of differentiation, i.e., the particular feather germ or portion of the feather germ in which the melanophore differentiates, de- termines which of its potencies is realized. It should be emphasized that there is no correlation between the locus of origin of the melanophores of any one genotype and their differential response to feather germs of various body regions. In fact, melanoblasts from the parietal lining of the coelom which would ordinarily never enter feathers will, upon being introduced into the epidermis of developing feathers of various regions, pro- duce all of the intricacies of pattern char- acteristic of homologous feathers of their genotype (Rawles, °45). Experimental evi- dence indicates that, in their undifferen- tiated or melanoblast stages, pigment cells of any one genotype are all alike. This ap- pears to be true, also, for some if not all of the amphibians (Stearner, ’46). The importance of the tissue environment, the feather papilla, in governing the color pattern response of melanophores of the Brown Leghorn to sex hormones has been fully demonstrated experimentally. It has been known for a long time that a variety of female colored, reddish bands, can be produced in the normally black breast feath- ers of a caponized Brown Leghorn by inject- ing a known quantity of female sex hormone at definite time intervals after plucking (Lillie and Juhn, ’32). In the light of later information regarding the developmental po- tentialities of the pigment-forming cells, it would appear that in the breast papillae these cells respond to female sex hormone by depositing red melanin granules into the SKIN AND ITS DERIVATIVES feather-forming epidermal cells. This par- ticular response to estrogen does not occur in vitro, and does not occur in vivo until the epidermis of the feather germ has at- tained a certain developmental stage (Trin- Kaus, °48). Willier (50) has interpreted the response of the estrogen-sensitive melano- blasts, in the zone of differentiation of the feather papilla, in terms of the physiologi- cal reaction gradients established by Lillie and Juhn (732). Since feather papillae of the various tracts and even within the same tract show distinct differences in reaction gradients, the melanophore response as re- corded in the finished feather pattern varies in a conformable manner. According to Wil- lier, the feather papilla is to be regarded as an endocrine receptor endowed with special prop- erties for responding to sex hormones. Dif- ferences in the responsiveness of the feather papilla appear to be genetically controlled. Danforth (43) finds that a simple altera- tion in genotype, e.g., a mutational change of a single H gene in fowl, suffices to change the response of the feather germ from one of sensitivity to estrone and indifference to testosterone, to one in which the tissues re- spond equally to both, 1.e., appear to be insensitive to the difference between these two hormones. Reactions Between Pigment Cells in Pat- tern Formation. In the formation of melanin pigmentation patterns there is not only a constant reaction between melanophores and their tissue environments, but also reactions between the individual melanophores them- selves. The importance of such reactions in the formation of rhythmic, barred patterns has been emphasized by Nickerson (’44) in his study of the Barred Plymouth Rock and the Silver Campine, two varieties of fowl with distinctly different types of black and white barring patterns. Having established that periodicity is intrinsic to the melano- phore and, further, that melanophores of the white bands are able to produce pigment under suitable conditions, Nickerson con- cluded that the barring rhythm is controlled primarily through the medium of diffusible substances produced by the active melano- phores within the black band which inhibit pigment formation by precursor melano- phores in their immediate neighborhood (subjacent white band). As growth proceeds and the black band becomes removed from the zone of differentiation of the feather germ, this region will lie beyond the in- hibiting influence and a new black band may now be formed. Certain properties of the SS feather germ, such as growth rate, size of the barb ridges, etc., are necessarily in- volved. It should be mentioned that, although Nickerson for good reasons favored the dif- fusion hypothesis, he did not lose sight of the theoretical possibility that, in the syn- thesis of melanin by the active melano- phores, some substance essential for melanin production might be removed from the epi- dermal substratum of the developing barbs of the white region. In either case the bar- ring rhythm would be associated with melanin production by certain groups of ac- tive melanophores. To what extent the pigmentary patterns in the hair coat of mammals are influenced by interactions between pigment cells (mel- anophores) awaits investigation. For more than fifty years it has been known that the skin epithelium of a white area of a spotted guinea pig gradually becomes black when in contact with an area of black skin trans- planted from another region of the same individual. Recently in an attempt to ex- plain this well-known phenomenon Billing- ham and Medawar (’48, °50) have assumed the passage of a hypothetical, cytoplasmic “ingredient” from the contiguous black pig- ment cells into the supposedly “white” pig- ment cells of the skin epithelium of the white area. A “white” pigment cell so trans- formed in turn transforms other contiguous “white” cells. According to this conception the process of pigmentation proceeds in a manner formally equivalent to a virus in- fection. In the light of well-established facts regarding the migration of precursor pig- ment-forming cells in amphibians, birds and other mammals, further and more crucial evidence is imperative to substantiate this view of “infective” transformation of pig- ment cells. In larval salamanders, the process of pig- mentation appears to be profoundly affected by interactions between developing pigment cells. Certain experimental studies have shown that pigment cells which have an advantage in age or in rate of development are able to inhibit or suppress the differ- entiation of younger or less rapidly differen- tiating precursor pigment cells (Twitty, 49; Lehman, *50). In fact, Twitty, on the basis of his extensive experimentation with Tri- turus, is of the opinion that influences ex- erted mutually by the pigment cells are of primary importance in their migrations and their arrangement into specific pigmentary patterns. 516 CONCLUSION From the foregoing it is evident that the skin of vertebrates is a complex organ com- posed of a variety of tissue elements and, like other organs and organ systems, pro- eressively increases in diversity during ontogeny. Tissue interrelationships appear to play a vital role in its organization and differentiation. Within recent years infor- mation has been obtained concerning the temporal and spatial restriction of develop- mental potencies of the skin ectoderm and cellular reactions between it and the under- lying mesodermal layer in bringing about regional specialization in the skin and its various derivatives. At the present time lit- tle is known about intrinsic, qualitative dif- ferences among integumentary cells that are revealed only by their particular methods of response to various types of stimuli, both internal and external. It is highly probable, however, that through continued and con- certed experimental attack and the use of new and sensitive biophysical and biochemi- cal methods it may yet be possible to obtain definite knowledge of the nature of these intrinsic qualities and thus increase our understanding of how they fit into the pat- tern of developmental organization of the skin and its diverse derivatives. REFERENCES Baltzer, F. 1941 Untersuchungen an Chimiaren von Urodelen und Hyla. Rev. suisse Zool., 48: 413-482. 1943 Weitere Beobachtungen an Pig- mentchimaren von Amphibien. Archiv. Julius Klaus-Stiftung, 78:664-670. Billingham, R. E., and Medawar, P.B. 1948 Pig- ment spread and cell heredity in guinea pigs’ skin. Heredity, 2:29-47. , and Medawar, P. B. 1950 Pigment spread in mammalian skin: serial propagation and immunity reactions. Heredity, 4:141-164. Bishop, G. H. 1945 Regeneration after experi- mental removal of skin in man. Amer. J. Anat., 76:153-181. Bodenstein, D. 1952 Studies on the development of the dorsal fin in amphibians. J. Exp. Zool., 120: 213-245. Bollinger, A., and Hardy, M. H. 1945 The sternal integument of Trichosurus vulpecula. J. & Proc. Roy. Soc. N. S. Wales, 78:122-133. Bonnevie, K. 1927 Die ersten Entwicklungssta- dien der Papillarmuster der menschlichen Fin- gerballen. Nyt. Mag. f. Naturv., 65:19-56. 1929 Zur Mechanik der Papillarmuster- bildung. I. Die Epidermis als formativer Faktor in der Entwicklung der Fingerbeeren und der Papillarmusterbildung. Roux’ Arch. Entw.- mech., 717:384420. SPECIAL VERTEBRATE ORGANOGENESIS Borcea, M. I. 1909 Sur Vorigine du coeur, des cellules vasculaires migratrices et des cellules pigmentaires chez les Téléostéens. Compt. Rend. Acad. Sci. Paris, 149:688-689. Cairns, J. M. 1951 Induction of regional speci- ficity in feather structure. Anat. Rec., 117:36-37. Cole, W. H. 1922 The transplantation of skin in frog tadpoles, with special reference to the ad- justment of grafts over eyes, and to the local specificity of integument. J. Exp. Zool., 35:353- 419. Cummins, H. 1926 Epidermal-ridge configura- tions in developmental defects, with particular reference to the ontogenetic factors which condi- tion ridge direction. Amer. J. Anat., 38:89- 151. ,and Midlo,C. 1943 Finger Prints, Palms and Soles. The Blakiston Co., Philadelphia. Dalton, H. C. 1950 Inhibition of chromatoblast migration as a factor in the development of ge- netic differences in pigmentation in white and black axolotls. J. Exp. Zool., 775:151-174. Danforth,C.H. 1943 Gene H and testosterone in the fowl; in Essays in Biology, pp. 159-165. Uni- versity of California Press, Berkeley, California. , and Foster, F. 1929 Skin transplantation as a means of studying genetic and endocrine fac- tors in the fowl. J. Exp. Zool., 52:443-470. Davies, H. R. 1889 Die Entwicklung der Feder und ihre Beziehungen zu anderen Integument- gebilden. Morph. Jahrb., 75:560-645. Davis, M., Boynton, M., Ferguson, J., and Rothman, S. 1945 Studies on pigmentation of endocrine origin. J. Clin. Endocrinol., 5:138-146. Detwiler, S. R. 1937a Substitution of lateral for axial mesoderm in relation to the development and segmentation of spinal ganglia. J. Exp. Zool., 76:35—-45. 1937b Observations upon the migration of neural crest cells, and upon the development of the spinal ganglia and vertebral arches in Amblystoma. Am. J. Anat., 67:63-94. Dijkstra, C. 1933 Die De-und Regeneration der sensiblen Endkérperchen des Entenschnabels (Grandry- und Herbst-Kérperchen) nach Durch- schneidung des Nerven, nach Fortnahme der ganzen Haut und nach Transplantation des Hautstiickchens. Zeit. f. mikro.-anat. Forsch., 34:75-158. Dorris, F. 1938 The production of pigment in vitro by chick neural crest. Roux’ Arch. Entw.- mech., 138:323-334. 1939 The production of pigment by chick neural crest in grafts to the 3-day limb bud. J. Exp. Zool., 80:315-345. DuShane, G. P. 1935 An experimental study of the origin of pigment cells in Amphibia. J. Exp. Zool., 72:1-31. 1943 The embryology of vertebrate pig- ment cells. Part I. Amphibia. Quart. Rev. Biol., 18:109-127. Eastlick, H. L. 1939 The point of origin of the melanophores in chick embryos as shown by means of limb bud transplants. J. Exp. Zool., 82: 131-157. Edwards, E. A., and Duntley, S. Q. 1939 Pig- SKIN AND ITS DERIVATIVES ments and color of living human skin. Amer. J. Anat., 65:1-33. Ellinger, F. 1941 The Biologic Fundamentals of Radiation Therapy. Elsevier Publishing Co., Inc., New York. Engert, H. 1900 Die Entwicklung der ventralen Rumpfmuskulatur bei Végeln. Morph. Jahrb., 29:169-186. Hale, A. R. 1949 Breadth of epidermal ridges in the human fetus and its relation to the growth of the hand and foot. Anat. Rec., 105:763-776. Hamburger, V. 1939 The development and in- nervation of transplanted limb primordia of chick embryos. J. Exp. Zool., 80:347-389. Hamilton, J. B. 1948 Influence of the endocrine status upon pigmentation in man and in mam- mals. Spec. Pub. N. Y. Acad. Sci., 4:341-357. Hardy, M. H. 1949 The development of mouse hair in vitro with some observations on pig- mentation. J. Anat., 83:364-384. Harrison, R. G. 1910 The outgrowth of the nerve fiber as a mode of protoplasmic movement. J. Exp. Zool., 9:787-848. 1936 Relations of symmetry in the de- veloping ear of Amblystoma punctatum. Proc. Nat. Acad. Sci., 22:238-247. Horstadius, S. 1950 The Neural Crest. Oxford University Press, Oxford, England. , and Sellman, S. 1946 Experimentelle Untersuchungen iiber die Determination des knorpeligen Kopfskelettes bei Urodelen. Nova Acta Reg. Soc. Sci. Upsaliensis, Series 4, No. 8, 13:1-170. Holmes, A. 1935 The pattern and symmetry of adult plumage units in relation to the order and locus of origin of the embryonic feather papillae. Amer. J. Anat., 56:513-537. Holtfreter, J. 1931 Uber die Aufzucht isolierter Teile des Amphibienkeimes. II. Ziichtung von Keimen und Keimteilen in Salzlésung. Roux’ Arch. Entw.-mech., 124:404466. 1933a Die totale Exogastrulation, eine Selbstablosung des Ektoderms vom Entomeso- derm. Entwicklung und funktionelles Verhalten nervenloser Organe. Roux’ Arch. Entw.-mech., 129:669-693. 1933b Organisierungsstufen nach _ re- gionaler Kombination von Entomesoderm mit Ektoderm. Biol. Zentralbl., 53:404431. 1935 Morphologische Beeinflussung von Urodelenektoderm bei xenoplastischer Trans- plantation. Roux’ Arch. Entw.-mech., 133:367- 419. 1939 Gewebeaffinitat, em Mittel der em- bryonalen Formbildung. Arch. exp. Zellforsch., 23:169-209. Juhn, M., and Fraps, R. M. 1934a Pattern anal- ysis in plumage. I. Curve of barb growth. Proc. Soc. Exp. Biol. & Med., 37:1181-1183. , and Fraps., R. M. 1934b Pattern analy- sis in plumage. III. Action of thyroxin in high concentrations. Proc. Soc. Exp. Biol. & Med., 37: 1185-1187. Keck, W. N. 1934 The control of the secondary sex characters in the English sparrow, Passer domesticus (Linnaeus). J. Exp. Zool., 67:315-347. Sli7 Kil, V. 1949 Experiments on the hair slope and hair pattern in rats. J. Exp. Zool., 110:397-439. Kollros, J., and Kaltenbach, J. C. 1952 Local metamorphosis of larval skin in Rana pipiens. Physiol. Zool., 25:163-170. Kupperman, H. S. 1944 Hormone control of a dimorphic pigmentation area in the golden ham- ster (Cricetus auratus). Anat. Rec., 88:26. Lehman, H. E. 1950 The suppression of melano- phore differentiation in salamander larvae fol- lowing orthotopic exchanges of neural folds be- tween species of Amblystoma and Triturus. J. Exp. Zool., 114:435-464. Leuenberger, T. 1942 Das Verhalten der Farb- zellen von Triton in Larven der Unke (Bomb- inator pachypus) bis zur Metamorphose. Rev. suisse Zool., 49:236-241. Lille, F. R., and Juhn, Mary 1932 The physi- ology of development of feathers. I. Growth-rate and pattern in the individual feather. Physiol. Zool., 5:124-184. , and Wang, H. 1941 Physiology of de- velopment of the feather. V. Experimental morphogenesis. Physiol. Zool., 74:103-133. , and Wang, H. 1944 Physiology of de- velopment of the feather. VII. An experimental study of induction. Physiol. Zool., 77:1-31. Lindeman, V. F. 1929 Integumentary pigmenta- tion in the frog, Rana pipiens, during metamor- phosis, with especial reference to tail-skin his- tolysis. Physiol. Zool., 2:255-268. Loeb, L., and Haven, F. L. 1929 Quantitative studies on the growth of the epidermis. Anat. Rec., 42:217-241. Lopashov, G. V. 1944 Origin of pigment cells and visceral cartilage in teleosts. Compt. Rend. Acad. Sci. U.R.S.S., 44:169-172. Luther, W. 1934 Untersuchungen iiber die Um- kehrbarkeit der Polaritat zwischen Aussen- und Innenseite des Ektoderms von Amphibienkeimen. Roux’ Arch. Entw.-mech., 737:532-539. Maumenee, A. E., and Scholz, R. O. 1948 III. The histopathology of the ocular lesions produced by the sulfur and nitrogen mustards. Johns Hop- kins Hosp. Bull., 82:121-147. Miner, N. 1951 Cutaneous localization follow- ing 180° rotation of skin grafts. Anat. Rec., 109: 66-67. Murray, P.D. F. 1928 Chorio-allantoic grafts of fragments of the two-day chick, with special ref- erence to the development of the limbs, intestine and skin. Australian J. Exp. Biol. & Med. Sci., 5: 237-256. Newth, D. R. 1951 Experiments on the neural crest of the lamprey embryo. J. Exp. Biol., 28: 247-260. Nickerson, M. 1944 An experimental analysis of barred pattern formation in feathers. J. Exp. Zool., 95:361-397. Niu, M. C. 1947 The axial organization of the neural crest, studied with particular reference to its pigmentary component. J. Exp. Zool., 105:79- VAS: Noble, G., and Wurm, M. 1940 The effect of testosterone propionate on the _ black-crowned night heron. Endocrinology, 26:837-850. 518 Orton, G. L. 1953 Development and migration of pigment cells in some teleost fishes. J. Morph., 93:69-100. Pasteels, J. 1937 Etudes sur la gastrulation des vertébrés méroblastiques. III. Oiseaux. Arch. de Biol., 48:381-488. Raven, C. P. 1931 Zur Entwicklung der Gang- lienleiste. I. Die Kinematik der Ganglienleisten- entwicklung bei den Urodelen. Roux’ Arch. Entw.-mech., 725:210-292. 1936 Zur Entwicklung der Ganglien- leiste. V. Uber die Differenzierung des Rumpf- ganglienleistenmaterials. Roux’ Arch. Entw.- mech., 734:122-146. Rawles, Mary E. 1944 The migration of melano- blasts after hatching into pigment-free skin grafts of the common fowl. Physiol. Zool., 17:167-183. 1945 Behavior of melanoblasts derived from the coelomic lining in interbreed grafts of wing skin. Physiol. Zool. 78:1-16. 1947 Origin of pigment cells from the neural crest in the mouse embryo. Physiol. Zool., 20:248-266. 1948 Origin of melanophores and their role in development of color patterns in verte- brates. Physiol. Rev., 28:383-408. Reis, K. 1930 Untersuchungen iiber das Ver- halten der Transplantate larvaler Amphibien- haut auf Larven und auf erwachsenen Amphib- ien, mit besonderer Beriicksichtigung der Meta- morphose. Roux’ Arch. Entw.-mech., 722:494— 545. Ris, H. 1941 An experimental study on the ori- gin of melanophores in birds. Physiol. Zool., 14: 48-66. Rosin, S. 1943 Experimente zur Entwicklungs- physiologie der Pigmentierung bei Amphibien. Rev. suisse Zool., 50:485-578. Saunders, J. W. 1950 An analysis of the spatial distribution, tract specificity and orientation of feather germs in the humeral tract of the chick wing. Anat. Rec., 108:32-33. , and Weiss, P. 1950 Effects of removal on the origin and distribution of feather germs in the wing of the chick embryo. Anat. Rec., 708: 93. Stearner, S. P. 1946 Pigmentation studies in sal- amanders, with especial reference to the changes at metamorphosis. Physiol. Zool., 19:375-404. Stone, L. S. 1933 The development of lateral- line sense organs in amphibians observed in liv- ing and vital-stained preparations. J. Comp. Neur., 57:507-540. Straus, W.L., Jr. 1950 The microscopic anatomy of the skin of the gorilla; in The Anatomy of the Gorilla, edited by W. K. Gregory, pp. 213-226. Columbia University Press, New York. , and Rawles, Mary E. 1953 An experi- mental study of the origin of the trunk muscula- ture and ribs in the chick. Am. J. Anat., 92:471- 509. Stréer, W. F. H. 1933 Experimentelle Unter- suchungen iiber die Mundentwicklung bei den Urodelen. Roux’ Arch. Entw.-mech., 730:131- 186. Trinkaus, J. P. 1948 Factors concerned in the SPECIAL VERTEBRATE ORGANOGENESIS response of melanoblasts to estrogen in the Brown Leghorn fowl. J. Exp. Zool., 109:135-170. Tung, T. C., and Tung, Y. F. Y. 1940 Experi- mental studies on the determination of polarity of ciliary action of anuran embryos. Arch. de Biol., 51:203-218. , Tung, Y. F. Y., and Chang, C. Y. 1948 Studies on the induction of ciliary polarity in Amphibia. Proc. Zool. Soc. London, 118:1134— 1179. Twitty, V. C. 1928 Experimental studies on the ciliary action of amphibian embryos. J. Exp. Zool., 50:319-344. 1945 The developmental analysis of specific pigment patterns. J. Exp. Zool., 100:141- 178. 1949 Developmental analysis of amphib- ian pigmentation. Growth Symp., 9:133-161. Tyler, A. 1947 An auto-antibody concept of cell structure, growth and differentiation. Growth Symp., 6:7-19. Uhlenhuth, E. 1917 A further contribution to the metamorphosis of amphibian organs. J. Exp. Zool., 24:237-301. Vogt, W. 1925 Gestaltungsanalyse am Amphib- ienkeim mit 6rtlicher Vitalfarbung. I. Methodik. Roux’ Arch. Entw.-mech., 106:542-610. 1929 Gestaltungsanalyse am Amphibien- keim mit 6rtlicher Vitalfarbung. II. Gastrulation und Mesodermbildung bei Urodelen und Anuren. Roux’ Arch. Entw.-mech., 120:384-706. Wang, H. 1943 The morphogenetic functions of the epidermal and dermal components of the papilla in feather regeneration. Physiol. Zool.. 16:325-350. Weidenreich, F. 1912 Die Lokalization des Pig- mentes und ihre Bedeutung in Ontogenie und Phylogenie der Wirbeltiere. Zeit. £. Morphol. u. Anthropol., Sonderhft., 2:59-140. Weigl, R. 1913 Uber hom@doplastische und heter- oplastische Hauttransplantationen bei Amphib- ien, mit besonderer Beriicksichtigung der Meta- morphose. Roux’ Arch. Entw.-mech., 36:595— 625. Weiss, P. 1935 The so-called organizer and the problem of organization in amphibian develop- ment. Physiol. Rev., 15:639-674. 1942 Lid-closure reflex from eyes trans- planted to atypical locations in Triturus torosus. J. Comp. Neur., 77:131-169. 1947. The problem of specificity in growth and development. Yale J. Biol. Med., 79: 235-278. ,and Andres, G. 1952 Experiments on the fate of embryonic cells (chick) disseminated by the vascular route. J. Exp. Zool., 1721:449-487. Wells, L.J. 1945 Pigmentation of the scrotum as a sensitive indicator for androgen. Anat. Rec., 91: 43-44, Willer, B. H. 1948 Hormonal regulation of feather pigmentation in the fowl. Spec. Pub. N. Y. Acad. Sci., 4:321-340. 1950 Specializations in the response of pigment cells to sex hormones as exemplified in the fowl. Arch. Anat. micros. Morph. expér., 39: 451-466. SKIN AND ITS DERIVATIVES Willier, B. H. 1952 Cells, feathers, and colors. Bios, 23:109-125. , and Rawles, Mary E. 1940 The control of feather color pattern by melanophores grafted from one embryo to another of a different breed of fowl. Physiol. Zool., 13:177-199. , and Rawles, Mary E. 1944a Genotypic control of feather color pattern as demonstrated by the effects of a sex-linked gene upon the mel- anophores. Genetics, 29:309-330. , and Rawles, Mary E. 1944b Melano- phore control of the sexual dimorphism of feather pigmentation pattern in the Barred Plymouth Rock fowl. Yale J. Biol. Med., 17:319-340. Woerdeman, M. W. 1925 Entwicklungsmech- anische Untersuchungen iiber die Wimperbewe- gung des Ektoderms von Amphibienlarven. Roux’ Arch. Entw.-mech., 106:41-61. Woronzowa, M. A. 1932 Analyse der weissen S19 Fleckung bei Amblystomen. Biol. Zentralbl., 52: 676-684. Wright, S. 1934a Physiological and evolution- ary theories of dominance. Amer. Nat., 68:24-53. 1934b Genetics of abnormal growth in the guinea pig. Symp. Quant. Biol. Cold Spring Harbor, 2:137-147. 1949a On the genetics of hair direction in the guinea pig. I. Variations in the patterns found in combinations of the R and M loci. J. Exp. Zool., 112:303-324. 1949b On the genetics of hair direction in the guinea pig. II. Evidence for a new domi- nant gene, Star, and tests for linkage with eleven other loci. J. Exp. Zool., 772:325-340. 1950 On the genetics of hair direction in the guinea pig. III. Interactions between the processes due to the loci R and St. J. Exp. Zool., 113:33-63. Section VIII ENERGY EXCHANGE AND ENZYME DEVELOPMENT DURING EMBRYOGENESIS* E..43;, BOEEL INTRODUCTION: ENERGY REQUIREMENTS OF THE EMBRYO ENERGY is required by the developing or- ganism to sustain four major groups of proc- esses: maintenance, growth, differentiation, and specific functional activities. Mainte- nance processes continue throughout life, but, to a certain extent, the others occur in sequence during development. Normally, they are not completely independent of each other, and, although they can to some ex- tent be separated conceptually,; they are not easily dissociable even by experimental means (Needham, *42). Maintenance metabolism may be defined as the sum total of energy-yielding or en- ergy-requiring processes concerned with the preservation of the living system as an or- ganized entity in its environment. It is at the bottom of the steady state equilibria through which, at least in part, integrity of the organism is preserved; it operates in processes of so-called “active transport,” and it is essential for replacement of parts of the living machinery that have been sub- ject to metabolic wear and tear through what Pollister (54) has termed ‘“mainte- nance protein synthesis.” For the embryo, as for all living beings, maintenance metabo- lism represents the basic cost of living. Dur- ing development, maintenance metabolism * This paper was completed during the tenure of a Fulbright Award for research at the Carlsberg Laboratories in Copenhagen. The author wishes to express his gratitude to Dr. Heinz Holter, Chair- man of the Division of Cytochemistry, and to Dr. Soren Levtrup for their critical reading of parts of the manuscript. +One of the major limitations in conceptual separation of the fundamental processes in ontogeny is that the embryo’s response to a given set of ex- perimental conditions is usually the same no matter whether the investigator is thinking more about maintenance than growth and differentiation at the time, or vice versa. 520 would be expected to increase, not because maintenance becomes progressively more dif- ficult as the embryo increases in organiza- tional complexity, but simply because there is more embryo to be maintained. Growth can best be defined in terms of protein synthesis.{ It is obvious that such { Although the concept of embryonic growth has been under consideration for a long time, there is still no measure of its magnitude that is completely adequate or free from objection. If growth is defined as increase in mass, it is apparent that the process may be independent of synthesis of new materials and may simply involve increase in size by imbibi- tion or deposition of inorganic materials. To meas- ure growth in terms of increase in solids, that is, as dry weight, is satisfactory in the case of embryos whose cells do not contain appreciable quantities of the raw materials for development, but it is com- pletely inapplicable to highly lecithal eggs. In the eggs of amphibians, for example, dry weight de- creases during development as yolk is consumed. The use of protein nitrogen as a measure of growth has the same limited applicability. Increase in num- ber of cells or of nuclei may be used to assess growth in certain instances, but in a number of embryos cell number increases while the size of the organ- ism remains essentially unchanged. The use of DNA, suggested by Berenblum, Chain, and Heatley (39) and recently emphasized by Davidson and Leshe (750) may be an adequate measure of growth within certain limits. It is clear the DNA content of the individual cells of a given species is constant (Boivin, Vendrely, and Vendrely, ’48; Vendrely and Vendrely, 49; Mirsky and Ris, 49). Accord- ingly, increase in DNA should presumably indicate relative increase in cell number. But it is also equally clear that the karyoplasmic ratio is not con- stant—at least, not in early development. There- fore, DNA content, while in some cases giving an indication of the relative changes in numbers of nuclei, is not an effective measure of cytoplasmic mass. Furthermore, it may be noted that the use of DNA as a nuclear measure may be questioned. Hoff-Jorgensen and Zeuthen (52) have reported that the frog egg contains a store of DNA enough for several thousands of new nuclei, and Lindahl (53) has demonstrated recently that micromeres ENERGY EXCHANGE AND ENZYME DEVELOPMENT DURING EMBRYOGENESIS 521 an arbitrary definition has certain shortcom- ings, one of the most important being that a great deal of what is commonly regarded as differentiation is also embraced by it. However, this may be just as well, for, in normal development, growth and differentia- tion usually proceed apace. Little is known directly about the energetics of protein syn- thesis in the embryo. The recent studies on the incorporation of labeled amino acids into the microsomes of tissue homogenates and the association of this process with oxidative phosphorylation, through synergistic action between microsomes and mitochondria, are, therefore, of fundamental importance to em- bryologists (see Pollister, °54, for a review of the literature). Energy is required for peptide bond synthesis, and it may reason- ably be expected that embryonic growth, in- volving as it does the synthesis of protein, will require energy expenditure on the part of the embryo. Differentiation involves the progressive spe- cialization of cells both structurally and functionally. Its visible manifestations are the form changes associated with cyto- and histo- genesis, but, in addition to these important macroscopic or microscopic changes, there are events of equal significance at a molecu- lar or macromolecular level of organization (Porter, 54). The chemical changes in the ontogeny of mitochondria or the elaboration of specific chemical substances, for example Nissl substance, actomyosin, or phosphatase, represent differentiation just as surely as do changes in cell shape or the develop- ment of pigment or secretion granules. This May appear obvious, but it is mentioned to emphasize the fact that differentiation involves synthesis of specific materials, as does growth, and such syntheses require energy. Much confusion and controversy have at- tended consideration of the energetics of dif- ferentiation (Needham, ’31, ’42; Tyler, ’42; Brachet, 50). One of the reasons for this state of affairs stems from a number of overly enthusiastic attempts to assess the cost of differentiation in precise quantitative terms from the results of experiments which in the sea urchin egg are haploid. RNA, since it is so intimately linked with protein synthesis (Caspers- son, 47; Brachet, ’47) has also been suggested, par- ticularly as a measure of cytoplasmic growth. Here too, difficulties are encountered, for not all of the RNA is located in the cytoplasm. Moreover, Herr- man and Nicholas (749) and Flexner and Flexner (51) have reported lack of correlation between RNA content and cell volume, were not designed to yield such information. Another difficulty arises from confusing what may legitimately be considered as energy for differentiation, i.e., energy for the synthesis of specialized materials, with the so-called “Organizational Energy,” or OF, of Need- ham’s (731) terminology. OE refers to a hypothetical quota of energy which is in- timately tied up with the organization of the embryo and which should be released as a measurable quantity when the embryo is disorganized by cytolysis. If OE really exists, it follows that the energy expenditure dur- ing development, as measured directly by heat production, should be significantly less than that calculated from data on respiratory exchange and the foodstuffs consumed dur- ing development. Comparison of the results of direct and indirect calorimetry has failed to reveal such a quota of energy (Bohr and Hasselbalch, 03), as have attempts to deter- mine calorimetrically the evolution of heat during cytolysis (Needham, 31). A later study of the bee moth, Galleria mellonella (Crescitelli, ’35; Taylor and Crescitelli, ’37), indicates that energy expenditure during pu- pation is higher when measured indirectly from respiratory data than when determined calorimetrically. But definite conclusions from these observations cannot be drawn in the absence of chemical data on the sources of energy used during pupation. Smith’s (46, °52) investigations of energetics during the development of the rainbow trout show conclusively that loss in total fuel value of the embryo plus yolk is exactly equivalent to the heat produced. Failure to find a definite quantity of en- ergy that could be ascribed to OE has been interpreted by some as indicating that dif- ferentiation occurs without cost, but such a conclusion is as unwarranted as is the view that differentiation processes require a large proportion of the embryo’s total energy ex- penditure. Butler (46) has commented on this as follows: “If an organism can syn- thesize peptide bonds, it appears that it will have no great difficulty putting together pro- tein molecules of any degree of complica- tion. The free energy must come from the metabolic processes going on in the organ- ism. The complete oxidation of a glucose molecule to carbon dioxide and liquid water yields approximately 700,000 cal. of free energy per mol. This is of the order of mag- nitude sufficient for the building up into pro- teins of about a hundred amino-acid resi- dues. Thus there is no outstanding difficulty in accounting for the synthesis of living 522 ENERGY EXCHANGE AND ENZYME DEVELOPMENT DURING EMBRYOGENESIS structures with a fairly modest expenditure of food.” Activity metabolism refers to the energy required to sustain the specific activities of embryonic structures whose functional ca- pacities have been realized. Most organs do not become functional as soon as they appear in the embryo. Nevertheless, when function does commence, the energy expenditure of the organ increases. As an example, the heart may be mentioned. The respiratory rate (mel. oxygen consumed per vg. per hour) of the beating rat heart, at the earliest stage that it can be removed from the embryo, is 40; it is only 6 for the heart at rest (Boell and Nicholas, unpublished). The difference in metabolic rate in this instance clearly represents the energy used for muscular con- traction. It is of interest to note that total oxygen consumption of an embryo from which the heart has been removed is indis- tinguishable from that of an embryo with its heart in situ. This means simply that the heart is so small in relation to the total em- bryo that its absence produces no noticeable effect on total metabolism. Another example of activity metabolism is seen in the development of the grasshopper. The smooth course of respiratory increase, which characterizes postdiapause develop- ment, is punctuated, between the third and fifth days, by a significant decline in respira- tory rate. “At precisely this time the lateral walls of the embryo beat more slowly than they do immediately before or afterward; moreover, blastokinesis (revolution of the embryo around the yolk) has just been com- pleted on the third day. Since Slifer (’32) showed that blastokinesis ‘is accomplished by vigorous movements on the part of the em- bryo itself’ the decreased respiration may be attributed to lessened embryonic activity” (Boell, °35).* It may be concluded that the processes as- sociated with development—maintenance, growth, differentiation, and functional activ- ity—require energy. In the long run, these energy requirements are met through Ox1- dative processes. The embryo is not required to pay a premium for the large synthetic job *Tuft (53) apparently has misunderstood this description, for, in referring to the experimental ob- servation in support of his contention that “phases (of development) during which the cells of the embryo spread over the yolky parts of the egg are accompanied by a decrease in Oz uptake,” he states that “a similar phenomenon seems to occur at the same stage in the eggs of Melanoplus differentialis (Boell, ’35), but this author attributes it to a tem- porary decrease in the frequency of the heart beat.” it has to do other than that necessitated by its own inefficiency as an energy trans- former, but this is a characteristic shared by all living organisms. Neither is the embryo exempt from the construction cost of protein synthesis. “Developmental phenomena,” as Weiss ('53) has recently pointed out, “‘do not violate the laws of thermodynamics. . . the old problem of ‘energy of shape’ is still with us, presumably because of the fact that the energy requirements in growth and differentiation may be greatly overshadowed by the energy requirements for the continu- ous anabolic renewal of the protoplasmic SW SUCTE +) cpus Fors ENERGY RELEASE DURING PERIODS OF REDUCED OXYGEN SUPPLY OR ANAEROBIOSIS The eggs and embryos of most species will develop continuously and normally only in the presence of oxygen, but no very close correlation exists between the oxygen ten- sion of the environment and that required to sustain normal development. The sea urchin egg will respire and develop normally under oxygen tensions as low as 40 mm. Hg (Am- berson, ’28); normal development and _ res- piration in the grasshopper egg are possible in a gas mixture containing 10 per cent of oxygen (Bodine and Boell, ’34a), and even an egg as large as that of the frog can with- stand some variation in oxygen tension with- out developmental retardation (Parnas and Krasinka, ’21). On the other hand, numerous investigators have shown that respiration of the mammalian embryo in vitro requires an atmosphere of pure oxygen, and Philips (’41) has suggested that the normal oxygen ten- sion in air may not be sufficient to sustain an optimal level of respiration in the chick embryo during the period before circulation commences. Oxygen supply seems to be indispensable for continued development, but most em- bryos can safely withstand the effects of oxygen lack or reduced respiration for some time. But the embryos of different species vary considerably with respect to resistance to anoxia, and, by implication, with respect to their ability to derive energy from an- aerobic reactions or to accumulate an oxygen debt. In the sea urchin, Arbacia punctulata, the activation of the egg and the initial cor- tical changes associated with the process can occur in the complete absence of oxygen (Kitching and Moser, ’40), but cell division is immediately blocked as soon as oxygen is ENERGY EXCHANGE AND ENZYME DEVELOPMENT DURING EMBRYOGENESIS 523 withdrawn. Indeed, Amberson (’28) has shown that cleavage is retarded when the oxygen tension falls to 11 mm. Hg, and be- low 4 mm. mitotic activity is completely blocked. Hultin (’53) has concluded that oxygen is necessary from the beginning to assure development in the sea urchin. Ascaris eggs are likewise very dependent upon oxygen supply in order to cleave (Brachet, ’50, p. 165). It is of interest, there- fore, that mitosis continues in Ascaris eggs after centrifugation at 400,000 xg (Beams and King, ’36) although this treatment re- duces the rate of oxygen consumption to 25 per cent of normal (Huff and Boell, ’36). The eggs of Fundulus, on the other hand, are remarkably resistant to complete removal of oxygen. Loeb (1895) maintained Fundu- lus eggs under vacuum for four days and noted that cleavage could occur. In the pres- ence of cyanide, which exerts a strong de- pressing effect on respiration, development appears to proceed qualitatively normally but at a slower rate than in controls. By contrast, the mitotic process is blocked by cyanide in the eggs of the mackerel, cunner, and scup (Philips, *40). Frog eggs have been shown to develop through the cleavage stages in the absence of oxygen and in the presence of cyanide when respiration is only 10 per cent of normal (Brachet, ’34; Barnes, ’44; Spiegel- man and Moog, 45). Brachet believes that energy is supplied, during short periods of anaerobiosis, through the utilization of an oxidative reserve, typified by but not neces- sarily identical with glutathione, rather than through processes leading to the formation of lactic acid. During recovery from anaero- biosis, the respiratory quotient of frog eggs is not unity, as one would expect if accumu- lated lactic acid were being burned, but ranges between 0.27 and 0.52. During longer periods of anaerobiosis lactic acid is pro- duced, however (Lennerstrand, ’33; Brachet, 34; Barth, ’46). That avian and mammalian embryos can derive energy through anaerobic processes is a well established fact (Needham, ’31, ’42). However, most of the studies on these em- bryos have been concerned with the analysis of glycolytic mechanisms (see p. 541 ff.) rather than with the ability of the embryo to survive under conditions of reduced oxygen supply. A comparative study of the rate of anaerobic glycolysis and respiration in pre- somite and early somite rat embryos has re- vealed the interesting fact that the embryo can derive more energy from anaerobic breakdown of glucose than from oxidation (Boell and Nicholas, unpublished). This may have considerable significance for the embryo, for a condition of “Everest in utero” probably exists during the period before placentation as well as at birth. Examples could be multiplied, but those enumerated show that many embryos have the ability to derive energy, and in some cases to develop, under conditions where oxygen is lacking. Survival, during the exi- gencies of anaerobiosis, seems to be associated with the ability of the embryo to contract an oxygen debt. Sooner or later, however, this must be redeemed aerobically. For the vast majority of embryos, development de- pends upon an adequate oxygen supply and release of energy through respiratory proc- esses. RESPIRATION Oxygen Consumption During Development. With minor variations, the curve shown in O05 CONSUMPTION IN CC.//HOUR/60 GM. EGG 24 6 sO 2a 46 INCUBATION AGE IN DAYS Fig. 197. Oxygen consumption during development of the chick (from Romanoff, ’41). 18 20 22 Figure 197 describes the course of respira- tion during development of a wide variety of animals, both invertebrate and vertebrate. The chief characteristics of the curve are an initial period in which there is a con- stant percentage increase in oxygen con- sumption during equal intervals of time, then a period in which the rate of increase gradually lessens, and finally one, near the 524 end of development or of a particular phase of development, in which respiration in- creases only slightly, if at all. In the sea urchin, the course of respiratory increase during development can best be rep- resented, as shown in Figure 198, by two z 50 = [e) a 40 7Z (U) = => 7 z 236 2 2 oi 65 [op a a wre ” | : oe ! f O5 Zz x= Le) E < -: ze 10 |. DEVELOPMENTAL 2.DEVELOPMENTAL 3. DEVELOPMENTAL PERIOD PERIOD PERIOD ) 4 8 12 Seo Ole 4a es HOURS AFTER FERTILIZATION Fig. 198. Course of oxygen consumption during development of the sea urchin. Curves A and B were obtained from two different lots of eggs. (After Lindahl, 39.) connecting curves (Lindahl, ’39; Borei, °48). The sigmoid part of the curve represents respiration of the embryo from fertilization to the hatching of the blastula; the ascending limb of the second curve coincides with gas- trulation, and the terminal portion covers the period in which differentiation is the predominant developmental event. Metabolic Rate. It may be asked whether the curves shown in Figures 197 and 198 indicate that the rate of metabolism per unit of embryonic material is constantly in- creasing during development. This is not an easy question to answer owing to the dif- ficulties involved in determining accurately the actual amount of metabolically active embryonic material as distinct from the non- metabolizing components in the developing system (see footnote { on p. 520). However, in some cases, it is possible to separate the embryo from its yolk supply pretty com- pletely, and the study of respiratory rate in these embryos provides a negative answer to the question raised above. This is illus- trated in Figure 199, representing a com- posite graph of data from various sources on the metabolic rate (oxygen consumed per milligram dry weight per hour) of the chick embryo. The points for the first six days of development represent in vitro measurements on the isolated embryo; the terminal points ENERGY EXCHANGE AND ENZYME DEVELOPMENT DURING EMBRYOGENESIS were taken from Romanoff’s (’41) study of the respiration of the intact hen’s egg dur- ing incubation and have been corrected for the respiratory activity of the extraembry- onic membranes through the use of percent- age figures from Needham (’32a). The graph shows that metabolic rate is essentially constant, at an average of 10.1 (range 7.9 to 12.3) between 16 and 288 hours of development. It should not be concluded that Qo, remains at this level during the remainder of the period to hatching, for Romanoff has shown that it declines to 8 on the thirteenth day of development, and by the nineteenth day it is only slightly more than 5. The decline in respiratory rate undoubtedly reflects the fact that such meta- bolically sluggish components as feathers, cartilage, and connective tissue now make up an appreciable portion of the total em- bryonic mass. But from the twelfth day of development, the decline is also brought about in part by a decrease in the respira- tory activity of a number of tissues, particu- larly muscle and liver (Romanoff, ’43). Slow decline of basal metabolism during the life span of the individual seems to be the gen- eral rule (Needham, ’31, 42), but the proc- ess may not begin as early in embryonic life as was previously thought. The chief factor responsible for the in- crease in rate of oxygen utilization during the first 16 hours may be regarded as the opposite of that which later contributes to its decline. Philips (42), who has made the most complete study of the metabolism of {6} % 12\. ———————ry 6L ° 4 —————— a a ee -—=— (9) 20 40 60 80 100 120 140 ours 260 280 Fig. 199. Qog (ul. oxygen per mg. dry weight per hour) of chick embryo. Values for first 144 hours represent measurements of respiration of chick em- bryos in vitro by Philips (741, ’42), Romanoff (741, *43), Needham and Nowinski (’37), Dickens and Greville (’33b), Dickens and Simer (’30, ’31), and Warburg, Posener, and Negelein (’24). Final three points represent respiratory rate of chick embryos in vivo from Romanoff (41). the very young chick embryo, has suggested that the early increase in rate of respiration is correlated with the conversion of intra- cellular yolk materials into active cellular constituents. Essential constancy of respiratory rate, ENERGY EXCHANGE AND ENZYME DEVELOPMENT DuRING EMBRYOGENESIS 525 during a substantial part of embryonic life, may be seen in a number of other cases. This was clearly shown by Gray (726) for the trout embryo, and more recently Hayes, Wilmot, and Livingstone (751) reported that the respiratory rate of the Atlantic salmon embryo remains unchanged throughout the entire period during which weight data could be obtained. The actively developing grasshopper embryo, which at certain stages can be completely separated from yolk, also has a constant respiratory rate over a con- siderable period of development (Bodine and Boell, ’36a, ’37). Of course, during diapause respiratory rate falls, but this special situa- tion will be discussed later. In the rat em- bryo, the Qoy is approximately 30 during the cleavage stages, but it soon falls to a value of about 12 (Boell and Nicholas, ’48) which is maintained until around the fif- teenth day of development (Dickens and Greville, ’33a; Negelein, ’25). Respiratory Increase and Growth. It seems in those cases in which the yolk content of the embryo is relatively small, or where the yolky parts of the egg can be successfully separated from the embryonic materials, the respiratory rate of the embryo is uniform throughout a major part of the total period of development. In other words, as pointed out by Gray (27), the total respiratory ex- change at a given stage of development is proportional to the amount of metabolicaliy active embryonic material. It is, therefore, perhaps more than mere coincidence that respiratory data during development should follow a sigmoid curve (Gray, ’29a, b; Brody, ’45; Thompson, 42). During its early phases, growth appears to be an exponential process—that is, ap- proximately the same percentage increase occurs during successive equal intervals of time. The equation x = a.e*™! has been found empirically to fit growth data in a great many cases, and a plot of the logarithm of the magnitude of the growing entity against time thus yields a straight line.* When * It is not intended to attach any strict biological significance to the values of a or & in the equa- tion xz = a.e*'. The semilogarithmic plots of growth and respiratory data are intended as purely descrip- tive; the chief justification for their use lies in the fact that a convenient method is thereby provided for comparing and contrasting curves. In arithmetic plots, the similarities and differences between such curves are not always readily apparent. The reader is referred to Sholl’s (’54) paper in which is con- tained a thoughtful analysis of the utility and lm- itations of empirical curve fitting (see also Levy, 62). growth rate changes abruptly during devel- opment, a series of intersecting straight lines will result. If respiratory increase parallels growth of metabolically active embryonic mass, and the evidence reviewed above strongly suggests such a relationship, the data shown in Figure 197 should yield a linear curve when log respiration is plotted against time (Fig. 200). The points do not & S Log Oxygen Uptake Ss (6) 5) 70 15 Days Fig. 200. Semilogarithmic plot of oxygen con- sumption during development of the chick (data from Table 1 of Romanoff, 741). fall on a single straight line, however, for between the seventh and eighth days of de- velopment an inflection appears, and from then on respiratory increase proceeds at a lower rate than earlier. A similar inflection in the semilogarithmic plots of dry weight or wet weight also occurs on the eighth day. What is responsible for the inflection in the respiratory curve is not known. It is interesting to point out that it occurs at about the time that the embryo shifts from predominantly carbohydrate catabolism to protein (see p. 533), and that the rate of protein absorption for growth is minimal on the eighth day (Needham, ’31, Table 111, column 13). Novikoff and Potter (48) have obtained data on increase in PNA and DNA in the chick embryo, and similar results have been described by Reddy, Lombardo, and Cerecedo (752). A semilogarithmic plot of Novikoff and Potter’s PNA figures yields a pair of intersecting curves almost identical in slope with those for respiration, and with a break on the eighth day. Thus a number of lines of evidence suggest that the period around the eighth day of development is one of transition in the chemical growth of the chick embryo. 526 ENERGY EXCHANGE AND ENZYME DEVELOPMENT DURING EMBRYOGENESIS It will be noted that the points for the second, third, and fourth days deviate from the straight line drawn in Figure 200. These may represent biologically significant varia- tions since they are all in the same direc- tion, but it may be remarked that Romanoff has shown that the deviations coincide with periods of high variability among embryos CMM.05/EGG/HOUR STAGE 12 13 141516 10 20 30 40 50 60 HOURS Fig. 201. Oxygen consumption during develop- ment of Rana pipiens. Small points represent experi- mental determinations. Large circles represent cal- culated values from equations x = 0.095e9.977' and xX = 0.45e9-931', (From Moog, ’44a.) due to differences in morphological age or mortality during early development. Respiration of Amphibian Embryos. Typi- cal data for the respiratory exchange of the developing frog embryo are shown in Fig- ure 201 (Moog, *44a). Similar curves have been obtained for a large number of am- phibian species, both urodeles and anurans, and in all cases where careful attention has been paid to assure normal development and to avoid injury to the embryos during respir- atory Measurements, the curves are qualita- tively identical to that shown in Figure 201. It has been claimed that respiratory increase does not occur during cleavage in the am- phibian embryo, but this may be due to the use of insufficiently sensitive techniques for measuring the very low rates of gaseous exchange in early stages. As first indicated by Atlas (’38) and sub- sequently confirmed by Moog (’44a) and a number of others (Barnes, ’44; Spiegelman and Steinbach, °45; Boell, °45; Barth, *46; Ten Cate, 53), respiratory data for the am- phibian embryo can be closely approximated by the equation x=ae*t, providing the value of k is changed at an appropriate point in development. On a semilogarithmic plot a break in the curve thus appears similar to that in Figure 200. In addition to the pa- pers already cited, semilogarithmic plots of the results obtained by Bialascewicz and Bledovski (715), Wills (36), Fischer and Hartwig (738), and Hopkins and Handford (43) reveal that the break in the course of respiratory increase is a regularly occurring phenomenon in amphibian development. There are a few exceptions, but in most anurans the break is found approximately at gastrulation; in most urodeles it seems to occur during the tail-bud stages—in Am- blystoma punctatum at Harrison’s stage 32 to 34. The cause of the break is unknown and it is difficult to assess its developmental significance, especially since it does not ap- pear at the same developmental stage in the embryos of different species. There is some indication that it may be correlated with the development of “total respiratory sur- face” (Moog, *44a; Boell, ’45, °48). If this be the case one might expect that the break would occur somewhat later in embryos reared at low temperatures because of the increased solubility of oxygen and decreased respiration of the embryo. A comparison of the work of Atlas, Barth, and Moog, on Rana pipiens, shows that this may be the case.* However, Ten Cate (’53) has fol- lowed the respiration of a number of dif- ferent amphibians over a temperature range from 8 to 26°, and there seems to be no systematic variation in the position of the break with temperature. Unfortunately, some of Ten Cate’s experimental points show such wide scatter that it is difficult to dis- tinguish so-called breaks from aberrations due probably to technical errors or abnormal development. It seems reasonable to suggest that the chief reason for the increase in respiratory rate in the amphibian embryo is the same as in the fish, chick or mammal, and this appears to be that the oxygen consumed at * In experiments conducted at the Carlsberg Lab- oratory, I have found that a break in the respiratory curve occurs at stage 16 in Xenopus laevis. Before and after the break the respiratory curves are per- fectly linear when plotted semilogarithmically against time. These experiments were done at 15° on individual embryos, and the same embryos were used throughout the entire period of development. ENERGY ExcCHANGE AND ENZYME DEVELOPMENT DURING EMBRYOGENESIS 527 a particular stage of development is related to the amount of embryonic material.* Such a view is consonant with results of respira- tory measurements on gastrula explants (Barth, ’42; Boell, 42) in which it appears that oxygen uptake is inversely related to yolk content. It is well known that yolk, separated from the protoplasmic matrix in which it is held, has a negligible oxygen consumption. Synthesis of more embryonic material at the expense of the raw materials stored in the egg would thus be expected to result in increased oxygen consumption. Unfortunately, practically nothing is known about the transformation of yolk into active protoplasm or the rate at which the process occurs, because so far it has proven impossible to separate effectively “ac- tive” and “inactive” materials and to meas- ure them with any degree of precision. The term “yolk” has various meanings to differ- ent investigators. To some it refers to dis- tinct cytological entities—the yolk platelets or lipochondria. Yolk may also have a broader connotation and may be used to refer to any stored substances in the egg which are used by the embryo for energy and for growth. Yolk platelets clearly do not contain all of the reserve materials for de- velopment. This is particularly true of lip- oidal substances. When homogenates of am- phibian eggs are strongly centrifuged, the lipoidal material is driven to the centripetal pole and can be seen to consist of a whitish layer, probably lipochondria and a droplet of clear oil (Boell, ’42). In the intact egg the oil is most likely dispersed as an emul- sion in the cytoplasm. This interpretation agrees with Lgvtrup’s (’53a) calculation that 30 to 50 per cent of the fat in the amphibian egg is outside the yolk platelets. The yolk platelets also do not contain reducing car- bohydrate (Gregg and Loevtrup, 50). If evidence of visible change in the yolk platelets or lipochondria is taken as an index of yolk utilization, one would be forced to believe that the embryo does not call upon its yolk supply for growth or as a source of energy until late in development (Bragg, °39; Daniel and Yarwood, °39; Holtfreter, 46). But visible change in size or pattern of yolk platelets is not a very precise indi- cator of the beginning or course of yolk consumption. Kutsky (750) has shown, through the use of P#2, that turnover of phosphoprotein in yolk occurs from the very *It should be unnecessary to mention that this does not mean that rate of development is controlled by respiratory rate. beginning of development in R. pipiens and continues at a low rate throughout. Barth and Jaeger (’47b) found that the enzymes associated with phosphate transfer from yolk are active before gastrulation. Additional evidence for the involvement of yolk in early development has been provided by Friedberg and Eakin (749). These workers exposed eges of Hyla regilla to glycine, containing C14, and found that the amino acid is taken up by yolk in the early embryo. Although these studies show that yolk enters the ‘metabolic pool” much earlier than has generally been supposed, they give no indication of the rate of yolk utilization for growth. But, if it may be concluded that increase in rate of oxygen consumption is related to increase in magnitude of meta- bolically active components, it may be pos- sible to use respiratory activity, in the ab- sence of better and more orthodox measures, as an index of the transformation of yolk into protoplasm. It is clear from Figure 201 and from the work of others (Atlas, Barth, Boell, Fischer and Hartwig, Hopkins and Handford, Spie- gelman and Steinbach, and Wills) that abrupt changes or cyclic variations in respi- ration do not occur during amphibian de- velopment. Nor is there evidence of increased energy expenditure for such processes as gastrulation, differentiation of the primary germ layers, or their further elaboration into the definitive tissues and organs of the em- bryo. Any energy expenditure specifically required for these processes, since they occur gradually, is merely represented as part of the general rise in oxygen consumption that occurs during development. In an earlier study of amphibian respira- tion, Parnas and Krasinka (’21) claimed that sudden increases in respiratory rate occurred at distinct stages of development, with pla- teaus of constant oxygen uptake between. The first of these abrupt changes was pre- sumed to occur at the time of gastrulation; another, during closure of the neural folds; and a third, at the time the external gills made their appearance. From these observa- tions, Parnas and Krasinka were led to con- clude that major morphogenetic events were associated with increased energy expendi- ture. There can be no objection to such a conclusion, but it is apparently not validated by the experimental findings of the majority of more recent workers. Tuft (53) has reported, however, that the course of respiration in Xenopus laevis shows a period, interpreted as coinciding with gas- 528 ENERGY EXCHANGE AND ENZYME DEVELOPMENT DURING EMBRYOGENESIS trulation, during which respiratory increase does not occur, and he has cited the work of Barnes (44) as showing a similar phenome- non. Lévtrup (°53a,b) and Ten Cate (53) have also emphasized that periods of rela- tively constant oxygen consumption have been noted in their experiments. It should be mentioned that, in a number of instances, the period immediately following the break in the rate of respiratory increase seems to be one in which respiration remains con- stant, but such plateaus, if they may be so designated, are generally of short duration. Extremely long periods of essentially con- stant respiration, similar to those shown by Tuft or Barnes, are not the rule. It is likely that these plateaus have little fundamental significance, for they are not found con- sistently, even by the same investigator. Moreover, when they are found, they do not occur uniformly at gastrulation, as stated by Tuft. In this connection, it may be noted that the amphibian embryo is particularly sensitive to injury during gastrulation, and Smith (46) has found the same to be true for the trout. It seems possible, therefore, that long respiratory plateaus during gastru- lation may be an indication of damage to the embryo. The possibility of injury cannot be eliminated from Barnes’ experiments, for the eggs used were shaken at the rate of 120 cycles per minute over an amplitude of 8 cm. for 6 to 8 hours during a mano- metric run, and she has reported that a care- ful census of the eggs was made routinely after each experiment “to check on the num- ber living.” It is also possible that plateaus of respiration represent intervals of develop- mental retardation due to handling of the eggs or other unfavorable conditions. Under such circumstances chronological time and developmental time would not be equivalent. Time is an adequate base line for de- velopment only when it is certain that devel- opment is normal. That acceleration of de- velopmental rate occurs after embryos have been depressed by adverse conditions is clear from the work of Buchanan (738, 40). The heightened respiratory activity after a pla- teau may reflect a regulatory phenomenon of this kind. Energy for Maintenance vs. Energy for De- velopment. Respiratory Metabolism and Cell Division. Calculations of the amount of en- ergy necessary for cytoplasmic cleavage or for the production of “new surfaces” during mitosis have generally yielded figures so small as to suggest that the energy cost of cell division is negligible. The problem can- not be dismissed in this way, however, for the cell may not be completely efficient as an energy transformer during division, just as muscle is unable to channel all of the energy released during contraction into mechanical work. Apparently, respiratory increase does not necessarily follow the increase in number of cells. This was clearly shown in the work of Gray (’27), and more recently Tuft (’53) and Zeuthen (753) have indicated that in- crease in number of nuclei bears only a casual relationship to increase in respiration. A number of indirect approaches to the question of the energetics of cell division have been made through measurements of respiration when cell division is depressed. The difficulties of interpreting the results of such experiments are illustrated by the fact that hypertonic sea water and phenyl ure- thane inhibit cleavage without appreciably affecting oxygen consumption (Warburg, 08, 10), cyanide depresses respiration with- out affecting cleavage (see p. 523), and the substituted phenols block cleavage reversibly and at the same time bring about a tremend- ous increase in oxygen utilization (Clowes and Krahl, ’34, ’36; Tyler and Horowitz, 38a). Brachet (’38) studied the respiration of Chaetopterus eggs that had been artificially activated by potassium chloride. Such eggs will develop into ciliated objects resembling trochophores through a process of “differen- tiation without cleavage.” Brachet observed that the oxygen consumption of KCl-activated eggs increased less than in normal eggs. One might be tempted to consider such a differ- ence in respiration as representing the energy cost of cleavage, but Brachet went on to show that the rate of development of the KCl- activated eggs was much slower than normal. This was indicated by the fact that their DNA content was only 30 per cent of that in normally activated eggs. Similarly, Tyler and Horowitz (’38b) investigated respiration of cleaving and non-cleaving parthenoge- netic Urechis eggs and found that the res- piration of cleaving eggs increased at a faster rate than it did in non-cleaving ac- tivated eggs or in fertilized eggs whose cleavage had been blocked by phenyl ure- thane. Just as in Brachet’s experiments, the absence of cleavage was not the only differ- ence between the lots of eggs. To conclude from such experiments that the difference in respiration between normal and non- dividing eggs represents the energy cost of cleavage would, as Holter (’49) has pointed ENERGY EXCHANGE AND ENZYME DEVELOPMENT DuRING EMBRYOGENESIS 529 out, be completely unwarranted. Indeed, Andresen, Holter, and Zeuthen (’44) have shown that respiration in the absence of cleavage in the egg of the ascidian Ciona intestinalis is probably not significantly less than in eggs in which cleavage is occurring. These workers believe that the difference in respiration is altogether due to difference in developmental rate. A systematic study of the relation be- tween oxidative metabolism and cell di- vision, as revealed by the use of inhibitory confirmed this observation, but the period of enhanced carbon dioxide output during the mitotic cycle did not agree with that reported by Lyon. It should be noted that the carbon dioxide measured in these experiments may not have had a respiratory origin at all; it may have represented nothing more than carbon dioxide released from sea water by acids liberated from the egg (see p. 534). In contrast with the findings of Lyon and Vles, Meyerhof (711) reported that calori- metric determinations of energy exchange TasiE 16. The Effect of Various Chemical Agents on Cell Division and Res- piration in the Sea Urchin Egg (Arbacia punctulata)* AGENT RESIDUAL RESPIRATION AS PERCENTAGE OF NORMAL EFFECT ON CLEAVAGE Oxygen lack Complete block 20-30 KCN Complete block 20=99 Carbon monoxide Complete block 30 Sodium azide Complete block 50 Sodium sulfide 50% block 90 Sodium sulfide Complete block t 50 Diethyldithiocarbamate Complete block ¢ 100 Substituted phenols Complete block >100t Phenyl urethane Complete block 70 Iodoacetate 90% block 54 Barbiturates Complete block 20 * Data for this table abstracted from Krahl (750). + Cleavage block irreversible. t At a concentration slightly less than that required to produce complete reversible block to cell division, various substituted phenols increase respiration from 20 to 260% above normal. and stimulatory substances, has been con- ducted during the past two decades by Clowes and Krahl and their collaborators. Table 16 summarizes results for the sea urchin ege abstracted from the compre- hensive review of Krahl (750). From the data summarized in the table, and from what is known about the specific action of a number of the agents listed, it is possi- ble to conclude that energy release during cleavage involves the participation of the cytochrome oxidase system and a system of enzymes concerned with oxidative phos- phorylation (see p. 540). It is clearly impos- sible to state what the relative amount of respiration necessary to support cell division might be. A number of direct observations of res- piratory changes during mitosis have been made. As early as the turn of the century, interest was focused on this problem through the pioneer work of Lyon (’04) when he found that carbon dioxide production after cleavage of the sea urchin egg was con- siderably greater than before. Vles (722) during development of Strongylocentrotus lividus eggs failed to reveal any evidence of rhythmic variations that could be correlated with cell division. Gray (25) then investi- gated the problem by making respiratory measurements during cleavage in the eggs of Echinus miliaris, but he concluded that rhythmic changes specifically associated with cell division were not superimposed on the over-all course of oxygen consumption. Runn- strom (734) pointed out, however, that lack of synchrony in cleavage in the large num- bers of eggs used by Gray may have obscured an intrinsic respiratory rhythm. In a repetition of Gray’s experiments Runn- str6m found that oxygen consumption was higher during the early phases of the mi- totic cycle than later in the process. The above observations were all made on the eggs of sea urchins, but similar experi- ments have been performed on amphibian eggs. Trurnit (cited by Brachet, *50) meas- ured heat production of Triton eggs by actually inserting one of a delicate pair of thermocouples into the egg, and he observed 530 EnerRGY EXCHANGE AND ENZYME DEVELOPMENT DURING EMBRYOGENESIS a peak of heat production during the meta- phase of mitosis. Similar results were ob- tained in suspensions of sea urchin eggs. Stefanelli (’37) determined respiration of single frog eggs in a capillary respirometer, and, although oxygen consumption showed some erratic fluctuations with time, there was some indication of a peak in respiration just before the appearance of the cleavage furrow. A second minor peak is also evident in his curve, and this occurs a few minutes after the cells have separated. In Rana fusca eggs, respiratory peaks were noted at the time of cleavage furrow formation by Bra- chet (50), and secondary peaks were also seen between the major oscillations. The most systematic and thorough study of respiratory rhythms during cleavage has been made by Zeuthen (46, ’50a,b, °51, 53; see also the theoretical paper dealing with Zeuthen’s results by Linderstrom-Lang, *46). In the frog egg, only the difference between oxygen taken up and carbon dioxide liber- ated was measured; the results obtained indicated that divisions one to four were accompanied by rhythmic waves in respir- atory rate, the amplitude of the wave being 4 to 5 per cent of the basal rate. Similar phe- nomena were found in the eggs of Urechis and four species of sea urchin, but in these experiments the total oxygen consumption was measured, not the difference between ox- ygen and carbon dioxide. Zeuthen noted some variation in the time relations of respiratory rhythms and various phases of the mitotic cycle, but in most cases it appeared that the cytoplasm divided during the period of decreasing respiratory rate, and Zeuthen believes, therefore, that the respiratory rhythms are more intimately related to nuclear events than to cytoplasmic cleavage. It would appear from the foregoing ac- count that the association of rhythmic varia- tions in respiration with cleavage had been well established, but recently Scholander et al. (52) have expressed doubt on the occurrence of fluctuations in oxygen con- sumption during cell division. Working with single sea urchin eggs in a diver of ingenious design, Scholander et al. observed cycles of respiration in some cases, but they failed to find them in others although nuclear and cytoplasmic division were proceeding. When cycling was found, the period of enhanced oxygen utilization seemed to coincide with cytoplasmic cleavage rather than with nu- clear division as reported by Zeuthen, and they reported in addition that cycling tended to become strongly damped after the first cleavage and to disappear altogether after two or three divisions. Zeuthen (753) has calculated that the standard error of his measurements on the Urechis egg is only 0.66 per cent, much smaller than the error in Scholander’s work, and he has suggested that the spread of experimental points in the observations by Scholander et al. would make it impossible to detect respiratory rhythms consistently in single eggs. Metabolism During Developmental Block. A number of cases are known in which a period of development is followed by one of complete developmental block. The exam- ples to be discussed are of interest because they represent situations in which complete disengagement of growth and differentiation from maintenance is brought about without experimental interference. In amphibian hy- brids (R. pipiens eggs X R. sylvatica sperm) development proceeds normally through the beginning of gastrulation, but then, although the hybrid embryos remain healthy and vi- able, further development ceases (Moore, *41, 46). Barth (46) has been able to show that both hybrid and normal embryos consume oxygen at the same rate during cleavage and blastula stages. At gastrulation, when developmental block sets in, respiration fails to increase in the hybrids as it normally does. Lactic acid production is also decreased in hybrid embryos, but not quite so drastically as respiration. Apparently, however, metabo- lism is qualitatively identical in the two types of embryos, for the respiratory quo- tients were found to be the same. A somewhat analogous situation exists in Drosophila melanogaster eggs with chromo- somal abnormalities. Absence of the X- chromosome leads to developmental failure at about the time of gastrulation. Just as in the case of the amphibian hybrids, ‘“no-X” eggs respire normally so long as development con- tinues, but respiration ceases to increase when development fails (Boell and Poulson, ’39). The embryos are viable and continue to respire for many hours, but the rate of res- piration, in the absence of development, re- mains unaltered at a very low level. By con- trast, ‘‘attached-X” eggs (those with an extra chromosome) respire and develop normally. The diapause in insects represents a situa- tion where a reversible block to development is interposed between two periods of active growth and differentiation. During diapause in the grasshopper Melanoplus differentialis, cell division, growth, and differentiation oc- cur so slowly, if they occur at all, that one ENERGY EXCHANGE AND ENZYME DEVELOPMENT DURING EMBRYOGENESIS 531 may consider the diapause period as one in which development is completely suspended. * Respiration of the egg, or of the embryo iso- lated from it, is much lower during diapause than when development is going on even though blocked and developing embryos are of identical size and morphologically indis- tinguishable (Bodine, ’29; Bodine and Boell, ’34a,b, °36b, °37; Boell, ’35). Quantitative respiratory data for the two types of embryos are summarized in Table 17. During dia- pause, the embryo, as far as can be ascer- tained, is simply maintaining itself, and it is reasonable to suppose that the total energy expenditure, as indicated by the oxygen con- sumed, is channeled into maintenance proc- esses. The same can be said for the amphib- ian hybrid and the “no-X” Drosophila em- bryo. Are Different Respiratory Mechanisms In- volved in Maintenance and Development? Fisher and his co-workers have attempted to determine the relative proportions of the total respiratory activity concerned respec- tively with maintenance and with activity (Fisher, ’41; Fisher and Henry, °40, 44; Fisher, Henry and Low, °44; Fisher and Stern, 42; Henry and Henry, ’45). Working with yeast, luminous bacteria, and sea urchin egos, they have produced evidence that a fraction of respiration that is relatively sensi- tive to urethane or chloral hydrate is con- cerned with such cell activities as division or luminescence, and that another fraction, much more resistant to these narcotics, is concerned with maintenance. The mass law relation has been applied to their data through the equation U SS Wee Wf U is the fraction of respiration not inhibited by narcotic and is proportional to the free or active enzyme in the cell; J represents the fraction of respiration that can be inhibited by narcotic and is proportional to the amount of inactivated enzyme, and [N]¢, for prac- tical purposes, may be regarded as concentra- tion of narcotic. Such a relation assumes, and reasonably so in the case of most narcotics, a reversible combination of narcotic and en- U zyme. When Fisher plotted log 7 against the log of narcotic concentration, not one straight line resulted, but two, and these in- * Fitzgerald (’49) has shown, however, that al- kaline phosphatase increases late in the diapause period. tersected at a narcotic concentration about the same as that required to block cleavage completely. At this concentration respiration was inhibited about 70 per cent. Fisher and his co-workers have thus been led to con- clude that two respiratory mechanisms oper- ate in the cell: one, accounting for about 70 per cent of the total oxygen consumption, is concerned with cellular activities; the other, corresponding to approximately 30 per cent of the total respiration and more difficult to inhibit with narcotic, is regarded as in- volved in maintenance metabolism. Tasxe 17. Respiratory Rates of Diapause and Developing Eggs and Embryos EGGS O2 UPTAKE PER 100 EGGs PER HOUR Prediapause 40 Diapause 18 Postdiapause 43 EMBRYOS* Qo» (ul. O2 PER MG. DRY WT. PER HOUR) Prediapause 152, Diapause 0.59 Postdiapause 1.47 * The embryos were dissected free of yolk. A similar attack on the problem of energy for maintenance and energy for development has been made by Moog (’44a) through the inhibition of respiration and development of the frog egg with Chloretone. Moog suggests that “from the neurula on, normal develop- ment requires almost all of the normal oxy- gen consumption, and that development can- not proceed at all unless at least 50% of the normal oxygen consumption is operating.” In Chloretone solution capable of suppress- ing development completely without killing the embryo, she finds that respiration persists at a value between 40 and 50 per cent of normal, and she interprets these results as suggesting the presence of two systems of respiration concerned respectively with main- tenance and growth-differentiation. These conclusions are supported in part by applica- tion of the mass law in the same manner as done by Fisher. The action of lithium on respiration and development in the sea urchin egg may be appropriately mentioned at this time. Lith- ium, it will be recalled, exerts a vegetalizing influence on the sea urchin embryo. In an attempt to discover biochemical correlates 532 ENERGY EXCHANGE AND ENZYME DEVELOPMENT DURING EMBRYOGENESIS with this phenomenon, Lindahl (’39) was led to study the effects of lithium on respira- tion. In brief, he found that exposure of sea urchin eggs to lithium soon after fertilization largely prevents the rapid increase which occurs normally between fertilization and hatching of the blastula (see Fig. 198). He thus concluded that respiration in the sea urchin egg consisted of two fractions. One remained constant, or essentially so, through- out development; the second, the lithium- sensitive fraction, rose continuously from fertilization and was largely responsible for the exponential increase in respiration. Pre- sumably normal development depends upon the proper operation of both fractions; if the lithium-sensitive portion of respiration is de- pressed, vegetalization of the embryo results. Studies of the effects of carbon monoxide and cyanide on the respiration of yeast, sea urchin eggs, and grasshopper embryos pre- sent analogous situations. Here too, carbon monoxide- and cyanide-sensitive and -insen- sitive fractions have been demonstrated. But some question seems now to exist as to whether there is complete cyanide-insensitiv- ity in any case. Robbie, Boell, and Bodine (38) showed that approximately 20 per cent of the respiration of the diapause egg, formerly regarded as completely insensitive to cya- nide, could be depressed by 0.001 M potas- sium cyanide when care was taken to avoid loss of cyanide from the medium surround- ing the eggs. And Robbie (’46) went on to show that the unfertilized sea urchin egg, long regarded as the classic example of cy- anide insensitivity, also had reduced respira- tory activity in the presence of this reagent. Critique and Evaluation. From the ac- count given above, it is apparent that it is difficult, if not impossible, to assess the en- ergy requirements of particular events in the developmental process by a study of the over- all gaseous metabolism of the embryo. One can never be certain that minor ripples on the total curve of respiratory increase are significantly correlated with the develop- mental phenomena with which they appear to coincide temporally. They may represent nothing more than random and fortuitous fluctuations brought about by developmental conditions. Furthermore, it should be empha- sized that a period of increased oxygen utili- zation is not necessarily contemporaneous with an energy-requiring developmental event. The energy needed at the time the event is occurring might be supplied by breakdown of high-energy phosphate com- pounds such as adenosinetriphosphate (ATP). Thus a period of enhanced oxygen consump- tion would occur after the event and would be concerned only with regeneration of the energy stores. This would be identical to the situation in muscle, where oxidative recov- ery follows contraction. Nevertheless, when all the observations reviewed above are taken together, they strongly suggest that a certain fraction of the total respiratory exchange is used by the embryo for purposes of maintenance while another fraction is used to support develop- mental processes. The question may now be asked whether the energy requirements for these processes can be fixed quantitatively or even relative- ly. (For various points of view on this ques- tion, the reader may consult Needham, 742; Tyler, °42; Brachet, 50; Tuft, 53.) One may be tempted to conclude that the amount of energy needed to maintain a developing em- bryo is the same as the total energy released by a blocked embryo, so that the difference in metabolic level between embryos in the two states might be regarded as representing the cost of developmental work. Similarly, it might be thought that the energy require- ments for development could be derived by subtracting the total metabolism of the un- fertilized egg from that of the fertilized egg. Such considerations lead into difficulties, however, for in some cases fertilization is not associated with change in respiratory rate, and in others it actually decreases (Whitaker, ’33). Before firm conclusions could be drawn from data on blocked and developing em- bryos, or fertilized and unfertilized eggs, it would have to be demonstrated that a unit of oxygen consumed had the same calorific value for the two kinds of eggs or embryos. To obtain information on this would require very precise knowledge not only of the amount but also of the kind of foodstuffs be- ing oxidized as energy sources. Although it might be possible to demonstrate by chemi- cal analysis that a decrease in a potential energy source occurs during development, it may be pointed out that it does not follow that such disappearance is brought about ex- clusively by oxidative or other energy-yield- ing processes. The substance in question may be synthesized into some other compound, or it may be lost from the egg without par- ticipating in energy-yielding processes. There seems to be evidence that considerable quan- tities of metabolites may disappear from the egg by processes which do not involve the simultaneous utilization of oxygen (Lovtrup, 75/210) ENERGY EXCHANGE AND ENZYME DEVELOPMENT DURING EMBRYOGENESIS 533 There is an additional reason for ques- tioning the validity of considering the main- tenance requirements of a developing embryo as identical with those of a blocked embryo. It is entirely conceivable that the cost of maintenance for a developing system would be significantly greater than for a non-developing one. The rate of chemical turnover in the unblocked embryo is cer- tainly more extensive, the deterioration, through activity, of various enzymes is prob- ably greater, and “maintenance protein synthesis,’ as well as other synthetic proc- esses, 1s undoubtedly going on at a faster rate. It seems premature, therefore, to at- tempt to fix either the absolute or relative costs of maintenance and development, either as a total process or in terms of its sub- divisions, from the kind of data available at the present time. On the question of the existence of differ- ent respiratory mechanisms concerned with maintenance and development, it can be said that there appears to be little doubt that X-sensitive and X-insensitive fractions of respiration are found in the developing sys- tem. But the relationship of such systems to development on the one hand and to maintenance on the other is more obscure. One fundamental difficulty in using data derived from studies with inhibitors, in attempting to assess the relative amounts of energy exchange concerned with mainten- ance or with development, lies in the fact that the magnitude of the inhibitor-insensi- tive fraction of respiration, supposedly con- cerned with maintenance, varies widely and seems to depend upon the particular inhib- itor used. Thus, Fisher and Henry (744) found that sulfanilamides inhibit cell di- vision at concentrations that reduce respi- ration by no more than 45 per cent, and the concentration of penicillin required to block cleavage completely was found to have ab- solutely no effect on respiration (Henry and Henry, 45). The very multiplicity of the agents represented by X, and the fact that in certain cases X-sensitive and X-insensitive fractions can both be inhibited by the same substance, permit certain reservations as to the validity of considering the various frac- tions of respiration as proceeding through different mechanisms. ENERGY SOURCES DURING DEVELOPMENT The Ontogenetic Sequence: Carbohydrate, Protein, Fat. Needham (731, 42) has gathered together an impressive body of data to es- tablish the fact that the chick embryo uses the primary energy sources during develop- ment in the order—carbohydrate, protein, fat. Evidence for this has been derived from three sources: (1) chemical analysis of the ege contents and embryo at various stages of development, (2) respiratory quotient data for the egg and isolated embryo, and (3) estimations of the amount and nature of nitrogenous wastes produced by the embryo and extra-embryonic structures. The results of in vitro studies are in harmony with the concept that carbohydrate is the energy source of major importance for the early chick embryo (Dickens and Gre- ville, 33a,b; Philips, °41, ’42). The more recent work of Spratt (48, ’50) points in the same direction, for he found that differ- entiation of chick blastoderms, when culti- vated in synthetic media, would not occur unless glucose was present. Essentially similar results have been obtained by Taylor and Schechtman (749). Protein metabolism occurs throughout de- velopment, with most of the protein absorbed by the embryo being used for growth. Protein utilization for catabolic purposes begins to be significant on about the fifth or sixth day and is maximal on the eighth or ninth day. Conversely, anabolic protein util- ization is high during the first few days of development and minimal on the eighth or ninth day. Subsequently, the rate of absorp- tion of protein for growth rises gradually to a mew peak reached on the fifteenth day. Novikoff and Potter (48) have shown that the pentose nucleic acid content of the em- bryo fluctuates in phase with the processes outlined above.* Fat is the predominant energy source in avian development. About 60 per cent of the total fatty acid present in the egg at the beginning is combusted, and from this the embryo derives over 90 per cent of its caloric yield. Fat uptake, leading to combustion, begins at about the fourth or fifth day, and continues to increase in intensity, paralleling the increase in size of the embryo, through- out the remainder of the incubation period. In some few other embryos, the ontoge- netic sequence shown for the chick seems to * Novikoff and Potter compare their curve for PNA with data from Needham (’31) showing pro- tein content of the embryo in milligrams per cent dry weight. Had they used instead the intensity of protein absorption by the embryo (which is perhaps a better measure of protein metabolism) the paral- lelism of their results with those summarized by Needham would have been even more striking. 534 ENERGY EXCHANGE AND ENZYME DEVELOPMENT DurING EMBRYOGENESIS hold, but much of the evidence is derived from respiratory quotient data. Fiske and Boyden ((26) showed long ago that respir- atory quotients can be misleading as indica- tors of the kinds of metabolites burned, for the combustion of protein will yield res- piratory quotients of approximately 1, 0.8, or 0.7, depending upon whether the nitrog- enous end product is ammonia, urea, or uric acid. Moreover, with some methods of respiratory quotient determination, three different samples of living material are required, and, unless due consideration be given to assure uniformity of the eggs or embryos used, distorted values will result. In the grasshopper embryo, the sequence of energy source utilization seems to be the same as in the chick (Slifer, ’30; Boell, ’35; Hill, °45). The phase characterized by pre- dominant carbohydrate metabolism is very short, however, lasting not more than a day or two. It also appears that protein occupies a relatively insignificant role as an energy source, but that some protein is broken down is indicated by the accumulation of urates in the embryo (Bodine, ’46). Approximately 75 per cent of the total oxygen consumption can be accounted for as fat oxidation. How- ever, during the first two weeks of develop- ment, fat combustion accounts for no more than a third of the embryo’s respiration, while during the post-diapause period almost 90 per cent of the total oxygen consumption is at the expense of fat. Fat has also been shown to be the major energy source in a number of other insects (references in Boell, "35; Crescitelli, °35; Needham, °42; Ludwig, 60a, b), and in the silkworm the utilization of energy sources occurs in the same sequence as in the chick and grasshopper (Needham, "42). A progressive change in respiratory quo- tient, with high values initially and lower ones, characteristic of fat oxidation, predom- inating in later development, has been re- ported for Fundulus (Amberson and Arm- strong, 733), Carcinus (Needham, ’33), Urechis (Horowitz, °40), and Trichurus (Nolf, 32). Thus the ontogenetic sequence— carbohydrate, protein, fat—may operate in these embryos also, but, as was pointed out above, respiratory quotient data alone are not adequate proof. The Fish Embryo. Both in 1931 and 1942, Needham noted that the ontogenetic sequence —carbohydrate, protein, fat—apparently did not hold in certain cases. Echinoderms and amphibians represented outstanding excep- tions. Smith’s (’46, 52) thorough studies of the energetics of the rainbow trout, Salmo irideus, indicate that this form also repre- sents an exception. Smith has shown that the amount of carbohydrate present at any one time and consumed during particular periods of development is so small in pro- portion to the total protein and fat combus- tion that carbohydrate as an energy source can be ignored. Carbohydrate combustion is limited to three short periods of develop- ment: (1) immediately after gastrulation, (2) at the time of hatching, and (3) at the onset of starvation after exhaustion of the yolk supplies. In the Atlantic salmon also, carbohydrate metabolism seems to be rela- tively unimportant, for, according to Hayes and Hollett (40), glycogen is absent in early development and the amount of glucose pres- ent is exceedingly small. It may be remarked that the failure of Hayes and Hollett to find glycogen in the early embryo has been at- tributed to faulty technique. In Salmo salar, Daniel (’47) reports that glycogen is present at the beginning of development, but he agrees with Smith and Hayes and Hollett that the quantity is small. In the rainbow trout, protein and fat are used simultaneously and account for 99 per cent of the total caloric yield during develop- ment. There is some indication that the percentage of protein burned is higher dur- ing early development (39 to 42 per cent) than later on (27 to 30 per cent), and it seems clear that consumption of water-sol- uble phosphatide fat precedes that of glycer- ide fat. During early development, Smith’s analyses show that ammonia may account for as much as 95 per cent of the total nitrogen excreted. Possibly the initial high respiratory quotients reported for Fundulus (Amberson and Armstrong, ’33) may have been due to protein combustion with am- monia as the end product. The Sea Urchin Embryo. Although res- piratory quotients of unity were found for the sea urchin egg immediately after ferti- lization by some of the earlier workers, it appears now that these values were abnor- mally high and were biased by the inclusion of non-respiratory carbon dioxide. Runnstrém (33) showed that an acid is liberated from sea urchin eggs at fertilization of sufficient strength to release carbon dioxide from bi- carbonates in sea water. Accordingly, Runn- strom (734) obtained “respiratory quotients” of 2, or even higher, during the first 15 min- utes after fertilization. The origin of the acid is of interest al- though it is only peripherally related to ENERGY EXCHANGE AND ENZYME DEVELOPMENT DuRING EMBRYOGENESIS 535 the question of energy sources. It is probably derived from carbohydrate, for Lindberg (43) was able to account for it quantita- tively by the amount of glycogen lost at fertilization. Moreover, Rothschild (39) was able to inhibit the production of acid in cytolyzing eggs with phlorizin, but it should be noted that the production of acid under these circumstances may not be the same as in intact eggs. The acid is not lactic acid (Runnstrém, 733; Rothschild, ’39), and there is little likelihood that it is fatty acid, as suggested by Hayes (38). Lindberg believes that the acid is not a single substance, but a mixture of intermediates of carbohydrate breakdown. More recent studies by Ohman (40) in- dicate that in Paracentrotus lividus the respiratory quotient is 0.73 after fertiliza- tion; it rises during subsequent development to a value of 0.85. Laser and Rothschild (’39) also noted that the respiratory quotient was low immediately after fertilization. In Arbacia punctulata, the respiratory quotient rises after fertilization, but the initial value (0.85) is not as low as that reported by Ohman nor is the increase so marked (Hut- chens, Keltch, Krahl, and Clowes, °42). It may be significant that the rise in respiratory quotient is temporally correlated with the development of lithium sensitivity, for Lin- dahl (39) has suggested that this fraction of respiration involves carbohydrate catabo- lism. The low initial respiratory quotient is consonant with the finding by Hutchens, Keltch, Krahl, and Clowes (742) that, during the first few hours after fertilization, the small quantities of glucose lost from the egg of Arbacia cannot account for the oxygen uptake. It also agrees with Lindberg’s (’43) observation that glycogen loss, after an in- itial rapid breakdown associated with acid formation, proceeds at a low and fairly con- stant rate. It is also in harmony with the report by Hayes (738) that fat is lost from the egg throughout development (45 hours) but that the rate of disappearance is most rapid during the three or four hours im- mediately after fertilization. Gustafson and Hasselberg (751) have re- ported that Kjeldahl nitrogen is constant from fertilization to the pluteus stage in Paracentrotus lividus (see the discussion on p. 520). Nevertheless, some protein must be burned, for Hutchens, Keltch, Krahl, and Clowes (42) have shown that protein com- bustion leading to ammonia as the end prod- uct can account for a portion of the oxygen consumption after fertilization. These col- lective findings thus suggest that fat is prob- ably the major energy source for a short time after fertilization and that during the period of exponential increase in respira- tion metabolism is mixed but with carbo- hydrate becoming increasingly important as development proceeds. It is of interest in passing to note that almost 90 per cent of the respiration of sea urchin spermatozoa can be accounted for by oxidation of phos- pholipid (Rothschild and Cleland, ’52). The Amphibian Embryo. The sequence of energy sources in Amphibia was recognized by Needham (742) as being different from that in the chick. Carbohydrate loss during early stages has not been demonstrated, but that it begins at gastrulation has been shown Taste 18. Respiratory Quotients of Rana pipiens Embryos STAGE NUMBER OF DETERMINATIONS RO oF 4 0.92 10 2, 0.90 11 4 iO 12 10 0.98 13-14 6 0.88 tS 1 0.85 * The first method of Dickens and Simer was used in making these determinations (Dixon, 751). by histochemical tests (Woerdeman, ’33; Raven, ’35) and chemical analyses (Brachet and Needham, ’35; Heatley and Lindahl, 37; Jaeger, °45; Gregg, ’48). The claim of Gregg and Pomerat (742) that glycogen dis- appears during cleavage has been withdrawn by Gregg. Loss of glycogen determined by chemical analysis, during development of Amblystoma mexicanum has been reported by Lovtrup (53a), but his analyses were confined to the beginning and end of devel- opment and hence do not show when con- sumption begins. That disappearance of glycogen is due to combustion rather than storage in the form of some other carbohydrate has been sug- gested by Gregg, since he found that the total free carbohydrate remains constant at a very low level until late in development. This seems reasonable, but it does not elim- inate the possibility that glycogen may be synthesized into some non-carbohydrate product. Brachet’s (°34) study of the respiratory quotient during development in Rana fusca is in harmony with the chemical analyses in showing that carbohydrate utilization be- 536 gins at gastrulation. The values obtained were 0.66 for early cleavage stages, 0.70 for advanced blastulas, and 1.03 for gastrulas. During subsequent development, the respir- atory quotient fell slightly. A series of re- spiratory quotient values for R. pipiens em- bryos, during and after gastrulation, are shown in Table 18 (Boell, unpublished). The data in the table confirm the findings of Brachet for respiratory quotients of ap- proximately unity during gastrulation. Un- fortunately, determinations were not made during cleavage stages. However, additional evidence that low values are characteristic ENERGY EXCHANGE AND ENZYME DEVELOPMENT DURING EMBRYOGENESIS inally present in the egg. That protein is burned during development is indicated by the fact that ammonia and urea are excreted by the embryo (Brachet, ’39, see columns 1 and 2 of Table 19). Boell, Needham, and Rogers (739) reported that ammonia was produced anaerobically by explants of Rana temporaria gastrulas, but this result may be in error in view of Gregg and Ornstein’s (52) inability to find ammonia production by explanted tissue of R. pipiens. Column 7 of Table 19 suggests that a not inconsidera- ble portion of the total oxygen consumption can be accounted for as protein combustion, TaB1E 19. Utilization of Protein as Energy Source by Rana fusca* NITROGEN EXCRETED BY PROTEIN Oz NEEDED TO Os: UPTAKE % Ovo UPTAKE STAGE 100 EMBRYOS PER DAY EQUIV. TO. BURN PROTEIN OBSERVED{ AS PROTEIN NH; TN Tea IN (s{oe, 3) IN COL. 4 COMBUSTION 1 2, 3 4 5 6 7. Morula- blastula 55 24 38.4 240 228 292 78 Gastrula 32 25 38.1 238 225 520 43 Neurula 31 22 2589 224 Pall 72 800 2, * Figures in columns 1, 2, 3, and 4 represent ug., in columns 5 and 6, ul. } Data for oxygen consumption in column 6 from Brachet (’34); for columns 1 and 2, from Brachet (’39). of the period before gastrulation may be seen in the observation that explants from the blastula of A. mexicanum have a re- spiratory quotient of 0.75 (Boell, Koch, and Needham, ’39). Barth (46) has also obtained high values for the respiratory quotient dur- ing gastrulation and later development, al- though the average is somewhat lower than in the experiments already mentioned. This may possibly be due to the fact that oxygen consumption and carbon dioxide output were not determined on the same sample of eggs. Barth measured the respiratory quotients of cleavage stages but found that they were only slightly lower than during gastrulation. However, Barth apparently feels that the absolute value of the respiratory quotient for early cleavage must be regarded as un- known from his work, for his figures varied from 0.7 to 1.05, depending somewhat on the method used. The question of protein as an energy source is highly confused. Most of the data summarized by Needham (’31) showed com- bustion of protein before hatching and some- what enhanced utilization afterward. There was, however, little agreement as to the amount lost in relation to the amount orig- and that relatively more energy may be de- rived from protein combustion during early development than later. These calculations, taken with the observations on carbohydrate catabolism, suggest that protein utilization precedes that of carbohydrate, and it prob- ably continues throughout all of develop- ment. The question of energy source utilization in A. mexicanum has been investigated by Lovtrup (53a) in a study combining re- spiratory and reduced weight measurements. His calculations apparently show that car- bohydrate is used exclusively as an energy source during cleavage stages. But reduced weight data are of limited utility in estab- lishimg this convincingly, and Levtrup states that “the chemical analyses must of course be decisive on this point.” He has also cal- culated that protein may serve as an energy source and that the amount of protein so used corresponds to 16 per cent of that in- itially in the egg. Since his observations were continued into the feeding period, Lovtrup has concluded that some of the protein lost, perhaps a major portion, might have been spared had his animals been permitted to feed. It is not possible to determine in his ENERGY EXCHANGE AND ENZYME DEVELOPMENT DURING EMBRYOGENESIS 537 experiments how much protein was used before the animals were capable of feeding, or the extent to which the animals were forced by starvation to call upon their tissue reserves. It seems reasonable to conclude, however, that the degree of inanition was not very great at the time his observations were terminated, for the respiratory rate had not yet begun to decline as it does dur- ing starvation. The results of chemical analyses of the changes in total nitrogen during develop- ment apparently do not support the view that protein serves as an energy source. In contrast with the earlier findings which but the changes observed were very slight and did not exceed in magnitude the vari- ation between different samples of eggs or embryos. * The amphibian egg consists of a large amount of protein in relation to the total dry weight; Needham (731, p. 1105) gives a figure of 61 per cent. Most of this protein is stored as yolk and serves as a reserve on which the embryo draws for the elaboration of new materials during growth and differen- tiation. In view of the great quantity of protein in the egg at the beginning, any de- crease, as by combustion, is bound to be relatively small. Hence what appears to be TasiE 20. Total Nitrogen in Amphibian .Embryos+ SPECIES uc. N INITIAL uc. N FINAL % DIFFERENCE AMmoulystOmanpunciatlile sty she ares 430 0 AGLI SORAE. VOTDTOS 55 0 6006006004040 0K 280 S208 WE TLUNUSPLOTOSUS LMI ean ee ears 230 —12 IRGILET? TY WIA oc. Gwe cot DIGS ae canto Ao tee A eae 190 +13 { Data from Wills (36). In computing the average figures shown above, the first determinations at the beginning of development and the last determinations before the feeding stages were used. The number of determinations represented in the averages are as follows: for Amblystoma punctatum and Triturus torosus, the first five and the last five; for A. tigrinum, the first two and the last two; for Rana pipiens, the first nine and the last nine. showed some loss of protein by combustion, Wills (36) reported that total nitrogen was essentially constant throughout the _ pre- feeding period of development in four species of amphibia: Amblystoma punctatum, A. tigrinum, Triturus torosus, and Rana pipiens. A portion of Wills’ analyses have been ab- stracted and are shown in Table 20. Wills included the egg jelly in the material an- alyzed, but he claimed that its presence did not materially affect the quantity of nitrogen measured. This view is contested by Gregg and Ballentine (46), who found that jelly nitrogen averages 16 micrograms per embryo, that is about 10 per cent of the total nitrogen in egg plus jelly, and that the vitelline mem- brane contains about two micrograms. They did, however, confirm Wills, as did also Lgvtrup (753a), by stating that total nitrogen remains nearly constant throughout develop- ment. A minor decrease noted at stage 16+ (Shumway) was attributed to loss of the vitelline membrane and the fluid contained in the vitelline space. Some change in the distribution of nitrogen in various constitu- ents of the embryo was found, indicating that protein metabolism was going on even though it did not eventuate in combustion, constancy of protein may simply mean that changes in total nitrogen are obscured by variability in the samples of eggs and em- bryos used for analysis. That this factor may not have been adequately taken into account is illustrated both in the data of Gregg and Ballentine and in those of Wills. Gregg and Ballentine obtained an average figure of 162 ue. total nitrogen per egg at the be- ginning of development from a series of nine samples which ranged from 141 to 202 ug. The standard deviation for their array of figures is 17.2 ug., or 10.6 per cent and this corresponds to 107 ug. of protein. For the combustion of this amount 102 ul. of oxygen would be required. This could mean, using Atlas’ (38) figure of 250 wl. for the total oxygen consumption of one Rana _ pipiens embryo throughout development, that as * They found also that the quantity of ultracen- trifugable nitrogen, presumably representing gran- ular material, is constant throughout development. This is interesting in view of the fact that much of the RNA of the cell is located in such granules and that Bodenstein and Kondritzer (748), Kutsky (50), and Krugelis, Nicholas, and Vosgian (52) showed that RNA increases between ten- and twenty-fold. 538 ENERGY EXCHANGE AND ENZYME DEVELOPMENT DURING EMBRYOGENESIS much as 40 per cent of the total respiration of the embryo is absorbed by the standard deviation. Wills’ data for R. pipiens (Table 20) actually indicate more nitrogen present at the end of development than at the be- ginning, and it can be shown that the differ- ence is statistically significant! The yolk of amphibian eggs contains a relatively high proportion of phosphoprotein. The phosphate linkages are present in the form of high-energy bonds, and Harris (46) discovered a special phosphoproteinphospha- tase capable of releasing phosphate without prior action of protease. The mechanism of phosphate transfer from yolk is complex and involves ATP, other proteins, and several enzymes (Barth and Jaeger, °47a,b; °50a,b; Barth and Barth, °51). The high-energy bonds of phosphoprotein apparently represent a relatively insignificant store of energy, however, for it has been calculated that the net amount available is only 0.15 per cent of the total energy consumption during develop- ment (Levtrup, 53a). Relatively little work has been done on fat metabolism during amphibian embryo- genesis. That some fat is burned is evident from the analytical data of Bialascewicz and Bledovski, Barthelmy and Bonnét, and Parnas and Krasinka (cited from Needham, ’31). Furthermore, Atlas (’38) has concluded, from a comparison of the loss in dry weight and oxygen consumption of R. pipiens, that some fat is used as an energy source. Lovtrup’s findings also point in the same direction. The majority of these workers agree that fat utilization takes place predominantly during the latter part of development, although there may be some indication that it begins earlier than was formerly believed (Lovtrup, 53a). In conclusion, then, it may be stated that the exact order in which energy sources are used during amphibian ontogeny is unknown. There actually may be no sequence at all in the sense that it can be demonstrated in the chick, but there is some indication that pro- tein combustion precedes that of carbohy- drate in early development, and that after gastrulation oxidation of all three foodstuffs occurs. Factors Responsible for Changes in Energy Source Utilization. A problem of considerable interest concerns the mechanism by which the embryo shifts from one class of foodstuffs to another during development. It might ap- pear reasonable to assume that one energy source would be used preferentially until its concentration in the egg became so reduced that it could no longer be mobilized in ade- quate amounts; then a second energy source would be called upon. However, most of the information available at present from experi- mental studies suggests that the change to a second energy source is not brought about by the exhaustion of the one previously used. It has been shown that the chick embryo will enter upon the “protein phase” of me- tabolism even though carbohydrate is still present or when the carbohydrate supplies are artificially increased by injection of glu- cose into the egg. There is also evidence that failure to use an energy source is not the result of lack of necessary enzymatic machin- ery. Dickens and Greville (’33b) have found that chick embryos, in vitro, can burn protein leading to ammonia production during the period of development when the normal fuel is carbohydrate. Likewise, Needham (’32b) has shown that early chick embryos will burn protein when treated with fluoride. The grasshopper embryo, during the stage when 90 per cent of its energy yield is derived from fat oxidation, will burn glucose when it is supplied (Bodine and Boell, ’36b), and the embryo can be induced to burn protein by treatment with dinitrophenol (Bodine and Boell, ’38). It thus appears that the sequence of energy source utilization depends upon conditions in the embryo itself, but what these are is at present unknown. The question as to whether various devel- opmental processes may require special or unique sources of energy, different from those used for maintenance, cannot be definitely answered. But it may be suggested, on the following grounds, that they do not. It is be- coming increasingly clear that high-energy phosphate bonds in such compounds as ATP represent the immediate energetic source for development, and it seems likely that the ox- idative processes required to generate phos- phate bond energy would be the same no matter what the ultimate destiny of the en- ergy stored in them may be. The real prob- lem for the student of development is not how high-energy bonds are synthesized, but rather how the energy released from them is directed to bring about specific end results in the embryo. On this question ignorance is complete. RESPIRATORY MECHANISMS The Embden-Meyerhof, Citric Acid Cycle, Warburg-Keilin System. In the breakdown of carbohydrate to carbon dioxide and water, in the systems that have been most thoroughly investigated—yeast and skeletal muscle—the ENERGY EXCHANGE AND ENZYME DEVELOPMENT DuRING EMBRYOGENESIS 539 initial changes occur equally well in the absence or presence of oxygen. The so-called processes of anaerobic glycolysis are mediated by what is commonly called the Embden- Meyerhof scheme of phosphorylating gly- colysis, and the end result of the breakdown of glycogen or glucose is pyruvic acid. In the absence of oxygen, pyruvic acid is reduced to lactic acid by hydrogen transferred from reduced diphosphopyridine nucleotide (DPN). During the anaerobic reactions, some energy is released, and this is stored in high-energy phosphate bonds such as those in ATP. Pyruvic acid is broken down to carbon di- oxide and water by a series of oxidative steps involving the tricarboxylic or citric acid cycle. Pyruvate is first oxidatively decarbox- ylated, by processes involving the participa- tion of coenzyme A, and the two-carbon frag- ment remaining combines with oxaloacetate to enter the cycle as citrate. Electrons and hydrogen derived from the degradation of citrate back to oxaloacetate are transferred to DPN or to triphosphopyridine nucleotide (TPN) and thence to oxygen via the War- burg-Keilin system of the cytochromes and cytochrome oxidase. During the operation of the citric acid cycle, more high-energy phosphate bonds are generated. It has been shown during the past few years that the citric acid cycle also plays a role in the oxidation of fatty acids and amino acids. Fatty acids are first broken down by 8-oxidation to two-carbon fragments, and these, through coenzyme A, may enter the citric acid cycle. Many amino acids, after de- amination, also enter the cycle—in some cases through coenzyme A, in others more directly. These processes and the relationships of the Embden-Meyerhof scheme, the citric acid cycle, and the Warburg-Keilin system are shown diagrammatically in Figure 202. A conservative estimate shows that some 20 or 30 different enzymes are required to operate the complete scheme, and, in addi- tion, several coenzymes, “factors,” and other chemical agents are needed. It does not fol- low that the scheme outlined above operates in every living cell, but it has been shown to be widely applicable, not only in the case of various vertebrate tissues but also in yeasts and other microorganisms, and one is perhaps not far wrong to conclude that the basic theme, with minor variations, probably applies to most aerobic organisms. Our problem now is to determine to what extent this complex array of biochemical machinery is present and operative in the embryo. It will be apparent in what follows that many of the enzymic mechanisms which operate in cellular respiration in the adult organism are present, and functional, during early development, and the egg, far trom being a mass of protoplasm with simple enzyme equipment, is provided with or soon synthesizes an impressive battery of enzymes. Respiratory Mechanisms in Echinoderm Eggs. The sea urchin egg and embryo have Oxygen Cytochrome oxidase t Cytochrome a Cytochrome ¢ Cytochrome b ft Flavoproteins ir DPN and TPN Citric acid cycle t Coenzyme A 7. Pyruvic acid A Embden- Meyerhof scheme Z Carbohydrates Fatty acids Amino acids Fig. 202. Schematic representation of mechanisms concerned with oxidation of foodstuffs to carbon dioxide and water. been more thoroughly investigated, with re- spect to their oxidative enzyme equipment, than those of any other form. The systematic investigation of respiratory mechanisms dur- ing echinoderm development had its be- ginnings in the thorough studies of Runn- strom more than twenty years ago. Follow- ing Warburg (710), Runnstrém (’30) showed that cyanide strongly depressed respiration of the fertilized sea urchin egg but that the unfertilized egg was completely resistant to this inhibitor. In addition, he found that carbon monoxide, also effective as a depres- sant of respiration of fertilized eggs, stimu- lated oxygen consumption before fertilization. Next, Runnstrém (’30, °32) investigated the ability of eggs to oxidize dimethyl-p- phenylenediamine, the substrate used to test for Atmungsferment (cytochrome oxidase), and found that the potential activity of the 540 ENERGY EXCHANGE AND ENZYME DEVELOPMENT DURING EMBRYOGENESIS enzyme was the same before fertilization as afterward. Therefore, he concluded that the increased respiration and sensitivity to cya- nide after fertilization were not due to syn- thesis of new enzyme but were brought about by enhancement of the activity of enzyme previously present. In some way, fertilization had accomplished a change in the relationship of the enzyme with other links in the respiratory chain. Later work by Krahl et al. (’41) confirmed these observations. Cytochrome oxidase ac- tivity, measured in the presence of cyto- chrome c, was the same in unfertilized and fertilized eggs, and the enzyme seemed to have the properties usually associated with cytochrome oxidase from other sources. It is of interest to note that cytochrome oxidase is not localized in mitochondria but rather is associated with non-mitochondrial par- ticulates in the supernatant solution of centrifuged egg homogenates (Hutchens, Ko- pac, and Krahl, °42).* Gustafson (52) has assumed that cytochrome oxidase is grad- ually built into the mitochondria at the time when the mitochondrial population increases after the mesenchyme blastula stage (Gustafson and Hasselberg, 751; Deutsch and Gustafson, 52; Gustafson and Lenicque, °52). Cytochrome c¢ is abundant in sea urchin sperm (Ball and Meyerhof, ’40; Rothschild, ’48), but so far it has not been detected in the egg. However, both Ball and Meyerhof and Borei (50) observed absorption bands of iron porphyrins related to the cytochromes, and Borei believes that sub-detectable amounts of cytochrome may nevertheless be present and functionally significant. That sea urchin eggs contain dehydro- genases was first demonstrated by Runn- strom, and this was confirmed by Ballentine (38, 40), who showed in addition that the total dehydrogenase activity after fertiliza- tion was approximately three times greater than before. However, attempts to discover the various specific dehydrogenases con- cerned with the operation of the citric acid cy- cle were unsuccessful. The absence of succinic dehydrogenase has been frequently reported (Ball and Meyerhof, ’40; Ballentine, °40; Krahl et al., ’41), but recently Gustafson * Weber (personal communication) has obtained results which are at variance with those previously reported. He has found that approximately 90 per cent of the cytochrome oxidase of homogenates of unfertilized eggs of Paracentrotus lividus can be recovered in the particulate fraction sedimenting at 600 to 12,000 x g. and Hasselberg (’51) claimed to have found the enzyme in egg homogenates from two species of sea urchin. It should be noted, however, that Bodine, Lu, and West (’52) showed that reduction of triphenyltetrazol- ium chloride, used by Gustafson and Hassel- berg to measure succinic dehydrogenase activity, is not a specific test for the enzyme. DPN was found to be present in Arbacia eggs by Jandorf and Krahl (’42), and Barron and his co-workers were able to show that the eggs could oxidize pyruvic acid (Goldinger and Barron, ’46). This was also shown to be true for the eggs of Echinus esculentus by Cleland and Rothschild (’52b). It seems, therefore, that the dehydrogen- ases of the citric acid cycle must be present but that something in crude egg homogen- ates interferes with their activity. A possible explanation of the failure of others to ob- serve active specific dehydrogenases was offered by Keltch et al. (50) when they showed that echinochrome inactivates a num- ber of dehydrogenases by oxidizing their sulfhydryl groups. Indeed, they found that cell-free, non-mitochondrial particulate sys- tems of unfertilized eggs could oxidize a-ketoglutarate, oxaloacetate, and succinate, and they demonstrated that generation of high-energy phosphate bonds accompanied these oxidations. Thus, they confirmed the earlier observation of Lindberg and Ernster (48) of oxidative phosphorylation by homo- genates of Strongylocentrotus droebachiensis eggs. Clowes, Keltch, Strittmatter, and Wal- ters (50) went on to show that oxidative phosphorylation could be inhibited by nitro- or halo-phenols, and it was concluded that generation of high-energy phosphate bonds occurs, as it does in muscle or kidney, through the citric acid cycle. That the citric acid cycle operates in the egg of Echinus esculentus has been con- vincingly demonstrated by Cleland and Rothschild (’52b). Evidence for the presence of the enzymes in the cycle rests upon (1) demonstration that malonate, long known to act as an inhibitor of succinic dehydrogenase, blocks the endogenous respiration of egg homogenates, (2) finding that oxygen con- sumption of egg homogenates can be stimu- lated by all of the intermediates in the citric acid cycle (see Table 21), and (3) demon- stration of the complete oxidation of pyruvic acid. Evidence for the operation in sea urchin eggs of a typical Embden-Meyerhof glyco- lytic cycle has also been provided. Gustafson and Hasselberg (’51) demonstrated the pres- ENERGY EXCHANGE AND ENZYME DEVELOPMENT DURING EMBRYOGENESIS 541 ence of aldolase in the eggs of Paracentrotus lividus, and Cleland and Rothschild (’52a) showed that lactic and pyruvic acids could be produced from the following intermedi- ates in the Embden-Meyerhof scheme: gly- cogen, glucose, glucose-1-phosphate, glucose- 6-phosphate, fructose-6-phosphate, hexosedi- phosphate, and phosphoglycerate. Further- more, they found that glycolysis was greatly depressed by phenylmercuric nitrate, phlor- izin, fluoride, and iodoacetate. Respiratory Mechanisms in Embryos Other than Echinoderms. Respiratory mechanisms in the embryos of other species have not been so thoroughly investigated as in the sea urchin. However, the functional signifi- cance of cytochrome oxidase is indicated for most embryos by their sensitivity to cyanide. In a number of cases, cytochrome oxidase has been assayed histochemically (Moog,’43) or manometrically, and the re- sults of the quantitative measurements show that the enzyme is present in sufficient amounts to account for the total respiratory activity of the intact embryo. This is true for the grasshopper (Bodine and Boell, ’36a; Allen, °40), various amphibians (Brachet, 34; Spiegelman and Steinbach, *45; Boell, 45), the chick (Albaum and Worley, °42; Albaum, Novikoff, and Ogur, ’46; Levy and Young, 48), and the mammalian cerebral cortex (Flexner, Flexner, and Straus, *41). Attempts to demonstrate the presence of cytochrome c in early embryos have been no more successful than in the sea urchin already mentioned. In Amphibia, the pres- ence of pigment, of course, makes direct spectroscopic examination impossible. Yaoi (28-29) was unable to find cytochrome c in the chick embryo, and he stated that if it was present at all between the fourth and eleventh day its concentration was so low as to be negligible. Likewise, Potter and Dubois (42) were unable to find cytochrome c in the chick embryo until the tenth day. In the cerebral cortex of the pig fetus, the bands of reduced cytochrome c were only faintly visible after the middle of gestation (Flexner, Flexner, and Straus, ’41). Stotz (39) stated that cytochrome c is present in the early rat embryo (age not specified) in a concentration of 0.3 mg. per gram of tissue (dry weight); in the late rat embryo, he found 0.18 mg. per gram. It has been claimed that succinic dehydro- genase and other dehydrogenases in the citric acid cycle are absent from the early chick embryo (Banga, ’37; Elliott and Greig, 38; Greig, Munro, and Elliott, 739). But Booth (35) showed that succinic, lactic, hexosediphosphate, a-glycerophosphate, and glucose dehydrogenases were present on the eighth day. More recent work has shown the presence of an active succinoxidase system as early as the twenty-fifth hour of incuba- tion (Albaum, Novikoff, and Ogur, °46), and Spratt (752) has found succinic dehydro- genase in the early blastoderm. Potter, Schneider, and Liebl (45) made measure- ments of the succinoxidase system in rat liver and brain from the seventeenth day of gestation through the thirtieth day after birth and showed that the activity increased TaBLe 21. Effect of Citric Acid Cycle Inter- mediates on Oxygen Consumption of Sea Urchin Egg Homogenates* SUBSTRATE OXYGEN CONSUMPTION AS 10M. PERCENTAGE OF CONTROL Citrate 1135 a-Ketoglutarate 145 Succinate 142 Fumarate 115 Malate 121 Glutamate 110 Pyruvate 120 * Data from Cleland and Rothschild (’52b). during this period. Flexner and Flexner (46) believe that the succinoxidase system of the cerebral cortex of the fetal pig is inactive until after the middle of gestation owing to insufficiency of cytochrome c. Noth- ing is known about the presence of succinoxi- dase in the early stages of amphibian de- velopment, but in Amblystoma punctatum the enzyme has considerable activity at the first stage tested (Harrison stage 20), and it may reasonably be assumed that it is present even earlier (Boell, ’48). As will be shown below, there is good evi- dence that the Embden-Meyerhof system of phosphorylating glycolysis exists in the chick embryo (Novikoff, Potter, and Le Page, 48), and Barth and Jaeger’s series of studies C47a,b, *50a,b) give strong evidence of the presence of a mechanism in the amphibian embryo, from early cleavage on, for oxida- tive production of high-energy phosphate bonds. Alternate Pathways of Energy Release. Cya- nide-insensitive Respiration. Cyanide does not inhibit completely the oxygen uptake of most embryos, and in some few cases res- piration is almost completely resistant to this reagent. These facts have been mainly 542 ENERGY EXCHANGE AND ENZYME DEVELOPMENT DuRING EMBRYOGENESIS responsible for the suggestion that a portion of the total respiratory exchange does not go through the usual Warburg-Keilin system but proceeds instead through a system of “non-ferrous” catalysts. Although, as was mentioned previously, so-called cyanide in- sensitivity is sometimes the result of inad- equate care to insure that cyanide concentra- tion remains constant during an experiment, it is nevertheless true that a very substantial portion of the total respiration of the un- fertilized sea urchin egg and of the grass- hopper embryo in diapause cannot be depressed by cyanide or carbon monoxide. Both Lindahl (40) and Robbie (’46) have suggested that residual respiration in the presence of cyanide may be of a special kind that does not normally occur. That inhibitors may induce extraordinary respir- atory processes has been shown through the use of fluoride on the chick embryo, of dinitrophenol on the grasshopper embryo, and of azide on paramecium (Boell, ’46). Notwithstanding, there remains the pos- sibility that a small but definite fraction of respiratory exchange is mediated by a system of enzymes different from those that operate in the citric acid cycle and the Warburg- Keilin system.* The nature of such a pos- sible pathway is unknown, nor is it known whether it operates normally as a collateral to the main system of respiratory catalysts or whether it becomes functional only when the Warburg-Keilin system has been in- hibited. The Phosphogluconic Acid Shunt. Lindberg has adduced evidence to support the view that respiration in the sea urchin egg, dur- ing the period of exponential rise, proceeds by a pathway other than the Embden-Meyer- hof scheme (Lindberg, 43; Lindberg and Ernster, 48). He has suggested that glucose breakdown is an oxidative process which bypasses the usual glycolytic cycle through what has been termed the phosphogluconic acid shunt or the Warburg-Dickens scheme. In this, glucose-6-phosphate is oxidized to 6-phosphogluconic acid, and then, through several intermediate steps, to pyruvic acid. The reaction is catalyzed by a_ specific enzyme, glucose-6-phosphate dehydrogenase, and TPN is an obligatory cofactor (see * One need but recall the reported absence of cyto- chrome ¢ in sea urchin eggs, the early chick em- bryo, and the cerebral cortex, or the observation of Sanborn and Williams (’50) that cytochrome zx (which combines certain properties of cytochromes b and ¢ and succinic dehydrogenase) is destroyed during pupation in Cecropia. Fruton and Simmonds, 53, p. 459 ff.). Evi- dence for the operation of the scheme in the sea urchin egg stems from the observa- tion that the oxygen consumption of egg homogenates can be stimulated by glucose or hexosemonophosphate, but hexosediphos- phate has a negligible effect. Furthermore, Lindberg and Ernster have shown that iodoacetate, which blocks the Embden-Mey- erhof system by inhibiting phosphoglycer- aldehyde dehydrogenase, does not affect respiration. Additional support of rather in- direct nature has been provided by Hultin (53) and rests upon the demonstration that carbon dioxide fixation by sea urchin eggs, as measured by the uptake of C4O,, is maximal during the period when respiration rises exponentially. The pertinence of this observation will be apparent when it is recalled that the mechanism of carbon di- oxide fixation by pyruvic acid, in the for- mation of malic acid, is coupled through TPN with the oxidation of glucose-6-phos- phate to 6-phosphogluconic acid (see Fruton and Simmonds, 53, p. 474 ff.). Additional evidence, of still more peripheral nature, has been furnished by Horstadius and Gus- tafson (47, cited by Horstadius, ’49) that phosphogluconic acid exerts a marked an- imalizing influence on sea urchin eggs. Finally, although Cleland and Rothschild (52b) presented cogent arguments for the operation of the Embden-Meyerhof scheme in the sea urchin egg, they showed also that phosphogluconic acid increased the oxygen consumption of egg homogenates and have concluded that the enzymes for the phosphogluconic shunt are present. Perhaps a fraction of respiration proceeds normally through this pathway; on the other hand, as suggested by Hultin (753), the pathway may have functional significance only when the enzymes concerned with the glycolytic cycle have been inhibited by sulfhydryl reagents. The phosphogluconic acid shunt may thus be analogous to the cyanide-in- sensitive fraction of respiration. Non-phosphorylating Glycolysis. Much at- tention has been given to the elucidation of the mechanism of carbohydrate breakdown in the early chick embryo, and the claim has been made that glycolysis does not pro- ceed through the usual steps of the Embden- Meyerhof scheme, i.e., through a series of phosphorylated intermediates, but occurs through a non-phosphorylating mechanism (see Needham, °42, pp. 610-615, for a re- view of the literature and summary of the evidence). Needham and his co-workers ENERGY EXCHANGE AND ENZYME DEVELOPMENT DURING EMBRYOGENESIS 543 found that the chick embryo was deficient in four essential components of the Embden- Meyerhof system: (1) the embryo could not attack glycogen or a number of hexosephos- phates, although glucose could be broken down readily, and it was concluded that phosphorylase, the enzyme esterifying glyco- gen, was absent or present in too low con- centration to be effective; (2) since triose- phosphate could not be broken down, it was concluded that triosephosphate dehy- drogenase was absent; (3) ATP was present, but in quantities too small to be effective; and (4) DPN was absent. It was postulated, therefore, “that in the chick embryo there are two separate routes of carbohydrate breakdown: (1) a non-phosphorylating glu- colysis mechanism, very active and closely bound to the cell structure, and (2) a phos- phorylating system closely similar to that in muscle, dealing with glycogen and hexose- diphosphate, but of very low activity” be- cause of the deficiencies noted above. Novikoff, Potter, and Le Page (748) ob- tained results which were quite different from those of the Cambridge workers, for they found that embryo homogenates would glycolyze hexosediphosphate, fructose-6-phos- phate, and glucose-6-phosphate as well as glucose. Moreover, they extracted from em- bryos, three to ten days of age, various hexosephosphates and their breakdown prod- ucts, ATP, and DPN in amounts roughly equivalent to those in adult tissues. They concluded, therefore, that their results did not rule out of consideration a non-phos- phorylating pathway for glycolysis but that postulating one was unnecessary since all enzymes and intermediates in the Embden- Meyerhof scheme were present in adequate concentrations. There can be little doubt, after consider- ing the experimental findings reviewed in the foregoing pages, that in the sea urchin egg and in the chick embryo all of the com- ponents necessary for the operation of the scheme shown in Figure 202 are present. Unfortunately, much less is known for most embryos. To suggest that the scheme has general applicability to the other forms commonly used in embryological studies would be no more than to hazard a guess. Such a conclusion may be correct, but it cannot be justified from the experimental evidence available at present. The recon- struction of a complex and intricate mech- anism for energy release and energy storage from a few biochemical fragments that have come to hand is not unlike the restoration of a skull from a few chips of brain case, a jawbone, and perhaps a tooth or two. Museums throughout the world give evi- dence of successes along this line, but the re- cent disclosures concerning the Piltdown man point up the dangers inherent in such a procedure. ENZYMES IN ONTOGENESIS Synthesis of Respiratory Enzymes. It seems reasonable to expect that increased oxygen consumption during development would be accompanied by a corresponding change in the enzymes through which respiratory proc- esses are mediated. Because of its key posi- tion in respiration, cytochrome oxidase has been most extensively investigated, and the results obtained in the grasshopper (Bodine and Boell, ’36a; Allen, 40), the salamander (Boell, *45), and the chick (Albaum and Worley, °42; Albaum, Novikoff, and Ogur, 46; Levy and Young, ’48) agree that in- crease in respiration is paralleled by in- crease in the amount of cytochrome oxidase in the embryo.* In the sea urchin, Deutsch and Gustafson (752) have reported that cytochrome oxidase rises during the first four hours after fertilization, but then it falls, during the next twenty hours, to a level equal to or less than the initial value. The homogenates used in these experiments were prepared by subjecting eggs to a freez- ing mixture and then shaking them vigor- ously during thawing. It is well known that cytochrome oxidase and certain dehydro- genases as well are inactivated by such treatment; hence the differences in enzyme activity found by Deutsch and Gustafson are probably due largely to variations in *It is impossible to express the quantity of an enzyme by the usual metrical units employed for other chemical entities. In general, what is meas- ured in enzyme studies is the activity of the enzyme under optimal conditions, so that the rate of reac- tion, during a reasonable period of time, is limited only by the amount of enzyme present in the reac- tion system. If care is taken to insure such condi- tions, reaction rate is found to be proportional to the concentration of the enzyme in the system. Meas- urement of enzyme activity in tissue minces or homogenates gives only an indication of the total potential activity of the enzyme. It provides no information on the degree to which the enzyme functions in vivo. This is particularly well illus- trated in developing and diapause grasshopper em- bryos, where it is found that the amounts of cyto- chrome oxidase are identical. But the enzyme func- tions to only a slight extent during diapause, as is shown by the lower rate of respiration and the de- creased sensitivity to cyanide. 544 the degree of enzyme inactivation. As will be recalled, Gustafson has claimed that cytochrome oxidase is gradually taken up into mitochondria at the time when they increase in the mesenchyme blastula stage. These results may indicate, therefore, that the sensitivity of the enzyme to freezing and thawing is greater after incorporation into mitochondria than before. In the frog embryo, Spiegelman and Steinbach (745) failed to find any increase in cytochrome oxidase between stages 6 and 19 (Shumway), although respiration of intact embryos rose approximately 800 per cent during this 200 300 400 500 600 Hours Fig. 203. Developmental changes in total nitro- gen (JV), volume (V), respiration (R), and succin- oxidase (S) in the central nervous system of Am- blystoma punctatum. Only the dimensions of the log scale are shown on the ordinate. Initial logarithmic values for N=1.06; for V=4.81; for R=2.56; and for S=2.50. Abscissa denotes hours of development from fertilization. Volume data from Boell (48); data for nitrogen, respiration, and succinoxidase from Boell and Shen (unpublished). period. They have interpreted this result as showing that what changes during de- velopment is not the amount of enzyme but the spatial orientation of enzyme and substrate. It will be recalled that the same idea had been expressed earlier for the sea urchin egg by Runnstrém (’30). The conclusion that a change occurs during de- velopment in the relationship of enzyme with its normal substrate is reasonable, but its validity would not be impaired by the demonstration that enzyme synthesis occurs. Brachet (’49) has pointed out that determination of cytochrome oxidase ac- tivity is difficult in frog embryos owing to the large endogenous oxygen uptake of egg homogenates, but by means of a_ spectro- photometric method he has shown a synthesis of cytochrome oxidase similar to that ob- served in Amblystoma. In the grasshopper, salamander, and chick embryos, cytochrome oxidase increases rela- tively more rapidly than respiration. It is ENERGY EXCHANGE AND ENZYME DEVELOPMENT DuRING EMBRYOGENESIS clear, therefore, that although most of the respiratory exchange may proceed through cytochrome oxidase, the concentration of enzyme is not the factor that limits the rate of respiration or of respiratory increase. A number of suggestions have been made as to the cause for limitation of respiratory rate, such as the amount of substrate avail- able or the rate at which it is mobilized, affinity between enzyme and substrate, or limiting concentrations of essential inter- mediates such as cytochrome c or coenzyme. All of these factors are of importance, no doubt, in influencing the rate of respiration, but it would seem that the chief factor responsible for the respiratory rate at a par- ticular time in development is the energy requirement of the embryo. Changes in succinoxidase activity have been determined in the salamander (Boell, °48) and the chick (Albaum et al., ’46). In Amblystoma, the rate of increase during development is probably not significantly different from that of cytochrome oxidase. Shen (749) has obtained similar results with mitochondrial preparations from de- veloping rat muscle. Cytochrome oxidase and succinoxidase increase exponentially be- tween the fifteenth day of gestation and the second week after birth, and the slopes ot the developmental curves are identical. In the chick, however, the two enzymes in- crease at somewhat different rates between the second and sixth days of development. It has been suggested that enzymes con- cerned with general metabolic processes, such as respiration, develop in relation to the growth of the embryo (Boell, ’48). In Amblystoma, the developmental curves for respiration, cytochrome oxidase, succinoxi- dase, and alkaline phosphatase (Krugelis, Nicholis, and Vosgian, *52), as measured in homogenates of the entire embryo, have fairly similar slopes, and the curves seem best interpreted as reflecting the synthesis of metabolically active material from yolk reserves. The relationship between enzyme synthesis and growth in the central nervous system (brain plus spinal cord to the level of the sixth somite) of Amblystoma is shown in Figure 203. Here it can be seen that respiration and succinoxidase increase ap- proximately in parallel with size of the nervous system, as determined by total ni- trogen or volume. In the chick embryo, where the relation between enzyme synthesis and growth can also be tested directly, it appears that increase in cytochrome oxidase is proportional to increase in total nitrogen ENERGY EXCHANGE AND ENZYME DEVELOPMENT DURING EMBRYOGENESIS 545 (Albaum and Worley, ’42). Levy and Young (48) have contested this, however, and have reported a rise in cytochrome oxidase activity (per milligram dry weight) from the third to the fifth day of development. It may be noted that Levy and Young did not deter- mine dry weight themselves but related their enzyme data to Murray’s figures. Furthermore, they did not determine the developmental age of the embryos used in the enzyme assays but calculated them from measurements of wet weight. Too much emphasis should perhaps not be placed on the absolute values of enzyme activity during the earliest developmental stages when rela- tive errors in enzyme determination and in weight and age can be of considerable magnitude. Enzyme Development and Differentiation. In the development of the embryo, functional activities are at first of a general and un- specific type and are concerned primarily with such events as maintenance and growth. Later on, however, the so-called special functions of tissues and organs, such as con- traction, conduction, secretion, etc., make their appearance. In general, only after morphological development has proceeded to a definite extent do the special functions come into being. In the case of the nervous system, Harrison (735) has commented as follows: “The relative size of the nervous system in vertebrate embryos reaches its maximum before nervous function begins. Complicated neuromuscular mechanisms in higher vertebrates are essentially complete structurally before they become active, and in those forms, such as the Amphibia, where activity seems to develop concomitantly with the development of structure, nervous func- tion may be suppressed by chloretone with- out interfering at all with the development of correlated structures or with their normal functioning as soon as the effect of the anaesthetic passes off. The embryo carries on the ordinary functions of organisms, such as respiration and metabolism, but the pecul- iarly developmental processes are apart from or superimposed upon these and are for the most part continuously changing and irre- versible.”’ Many of the important reactions which underlie functional activities, both general and special, are catalyzed by enzymes. The immediate change from an inactive state to one of activity may be expected, then, to be concerned primarily with development of biochemical mechanisms, through which physiological activities become possible, rather than with marked alterations in the structural elements, through which they are made manifest. This idea has been more fully elaborated by Barron (’41), Herrmann (53), and Shen (’54). If this argument is valid, one would expect to find that those enzymes that are uniquely involved in a par- ticular function would be formed, or increase their activity, prior to or synchronously with development of functional capacity. The analysis of enzymes in the entire embryo generally provides limited informa- tion on the relation between enzyme changes and functional differentiation of specific structures. But the study of individual organs has furnished a great deal of evidence that development of function and development of enzymes are correlated events. Pioneer work along this line was done by Flexner and his co-workers. During differ- entiation of the pig cerebral cortex, two “critical periods” are noted. The first of these occurs about half-way through gesta- tion and is marked by changes in size and shape of neurons and in pattern of Nissl substance. At this time, the cytochrome-— cytochrome oxidase system increases in ac- tivity, and this appears as an abrupt change when enzyme activity is plotted against fetal crown-rump length. Tissue respiration re- mains constant, however, as does also the activity of cytochrome oxidase, and Flexner et al. (41) have concluded that the factor which changes during the first period of differentiation is the concentration of cyto- chrome. During the second critical period, neurons assume their adult characteristics, and, at the same time, respiratory rate, cytochrome oxidase, and succinic dehydro- genase activities begin to rise to their adult levels (Flexner and Flexner, 46). A similar rise in succinoxidase activity was found in the developing rat brain by Potter, Schnei- der, and Liebl (745), but owing to the greater immaturity of the rat at birth, the increase does not commence until the fifth to seventh day post partum. Since respiratory enzymes are ubiquitous and are concerned with the ordinary ac- tivities of all cells, it might be expected that enzymes more intimately connected with specific functions would show more marked correlations with functional development.* * Perhaps the most striking example of sudden enzyme development is that of xanthine oxidase in the chick embryo. Morgan (’30) has reported that the enzyme is present in the liver of a chick that has just made a hole in the shell; it is absent in the liver of a chick that has not yet made a hole. 546 ENERGY EXCHANGE AND ENZYME DEVELOPMENT DURING EMBRYOGENESIS Because of its unique relation to the mech- anism of muscular contraction, ATPase has been extensively investigated. Moog and Steinbach (745) showed that the ATPase activity of early chick embryos is low but rises on the sixth day, and an attempt was made to correlate this with the development of spontaneous muscular activity. Later, Moog (47) showed that the temporal pattern of development of the enzyme was different ENZYMIC ACTIVITY (ML. 0.1N NAOH FOR 30 MINUTES) PREMOTILE EARLY FLEXURE SINGLE COIL DOUBLE S-REFLEX COIL in concentration of actomyosin, with which it supposedly was associated, and this strong- ly suggested that some of the ATPase in muscle was not combined with actomyosin. That this was true was then shown by Herr- mann, Nicholas, and Vosgian (749). The demonstration by Kielley and Meyerhof (48) of two ATPases in muscle, one asso- ciated with actomyosin, the other with cell particulates, and the findings of Herrmann, EARLY SWIM STRONG SWIM Fig. 204. The relation between development of cholinesterase and neuromuscular activity in Amblystoma punctatum. Ordinate denotes enzyme activity; abscissa, sequence of behavioral manifestations. (After Sawyer, 743.) in brain, liver, and muscle, and this could to some extent be correlated with the time of functional differentiation of the various organs. In the case of muscle, “the enzyme reaches its maximum level on the very day that the muscles are . . . called upon to work in the business of hatching’ (Moog, *62)). In the developing rat, correlation of ATPase with functional differentiation is more clearly indicated. Herrmann and Nich- olas (48a) showed that ATPase was at a very low level until the sixteenth day when, coincident with the beginning of muscle contractility, it rose sharply and continued to increase for approximately three weeks after birth. They found in addition (Herr- mann and Nicholas, ’*48b) that the rise in ATPase did not run parallel with increase Nicholas, and Vosgian were taken as a point of departure by deVillafranca (’53) in a study of the course of development of the two enzymes. The two ATPases can be distinguished not only by their differential location, but also by differences in pH op- tima and ion activation. Although the two enzymes begin to develop in activity at the same time, that is, at the onset of contrac- tility, as shown by Herrman and Nicholas, the course of increase is different. Acto- myosin ATPase increases more slowly, and somewhat in parallel with the actomyosin curve of Herrmann and Nicholas, while particulate ATPase develops more rapidly. The function of the actomyosin ATPase seems quite clear; that of particulate ATPase is more obscure. It may be suggested that it operates for release of energy in general ENERGY EXCHANGE AND ENZYME DEVELOPMENT DURING EMBRYOGENESIS 547 cell functions in a manner similar to the ATPases in non-contractile tissue (Barth and Jaeger, °47a,b, °50a,b; Barth and Barth, 51; Flexner and Flexner, ’48; Potter, Schneider, and Liebl, °45; Moog, °47). The accumulation in tissues of alkaline phosphatase has been studied by Moog (44a, 50, °51, °53). Undifferentiated tissue is rela- tively rich in alkaline phosphatase, but it then declines during differentiation of those tissues in which it has no obvious functional significance. On the other hand, in tissues where the enzyme presumably functions in some way, alkaline phosphatase undergoes tremendous increase in activity. This is well illustrated in the chick duodenum. Phos- phatase activity is very low until the ninth day of incubation, but then it increases tre- mendously, reaching a peak on the day of hatching. In the mouse, there are two peri- ods of phosphatase increase; the first begins on the sixteenth day of gestation, and, by the time of birth, an increase of a hundred- fold has resulted. A second rise begins on the fourteenth or fifteenth day post partum, and this is of the order of twentyfold. Thus the total increase in activity is some 2000-fold. The first increase, Moog believes, is in preparation for the diet of the mouse during the nursing period; the second anticipates the varied solid diet of the mouse on wean- ing. A similar case of enzyme development in “anticipation of function” is seen in the case of tyrosinase in the grasshopper embryo. The enzyme develops to its full activity dur- ing the pre-diapause period, yet the first indication of its function in the formation of melanin pigment is not apparent until a short time before hatching (Bodine and Boell, ’35; Bodine, Allen, and Boell, ’37). Perhaps no better example of the intimate relationship between functional differentia- tion and enzyme development exists than that between cholinesterase and the nervous system. Youngstrom (738) and Nachmansohn (39) had shown that cholinesterase increases some time during the period when the nerv- ous system becomes active, but it was left for Sawyer (43) to demonstrate the precision with which neuromuscular development and increase in cholinesterase activity are cor- related. Sawyer showed that the concentra- tion of the enzyme, low in premotile stages of Amblystoma punctatum, rose significantly and progressively during development of the behavior sequence described by Coghill (29). These relationships are shown in Figure 204. Through determinations of cholinesterase in the major subdivisions of the nervous sys- tem, it was possible to show, as might be predicted from Coghill’s study of the ontog- eny of neuromuscular activity, that cholin- esterase appears first in the spinal cord, and then sequentially in hindbrain, midbrain, and forebrain. Increase of the enzyme in the spinal cord coincided exactly with the first ability of the embryo to respond neurogeni- cally to an external stimulus. In the midbrain, cholinesterase development was most marked at the time when motor activity of the cord began to be dominated by the mesencephalon, as shown by Detwiler’s studies (’46a,b). The relationship between differentiation of the nervous system and cholinesterase is further illustrated by the results of experiments in which the extent of neural differentiation was influenced by experimental means. It is well known that removal of the eye from one side of the amphibian embryo, before optic fibers have developed, leads to marked hypoplasia of the midbrain lobe on the side opposite to that of eye removal. That this involves a qualitative change in the neural tissue, i.e., reduction in number of differentiated neu- rons, has been shown by a number of workers. Determination of cholinesterase in the two halves of the midbrain, at various times after unilateral eye extirpation, reveals that cholinesterase activity (per ug. nitrogen) de- velops normally in the lobe that receives optic fibers. But in the opposite side (deprived of incoming fibers from the retina), cholin- esterase activity is much reduced, and the relative decrease is roughly proportional to the reduction in number of differentiated neurons. The decrease in cholinesterase ac- tivity appears to be confined to the midbrain, for the enzyme activities in the right and left sides of the diencephalon and hindbrain were found to be identical. Furthermore, the effect on cholinesterase seems to be specific, for the respiratory and succinoxidase activi- ties of the two lobes of the midbrain were also found to be the same (Boell and Shen, "49, 50). The question may be asked whether the changes in enzyme activity described above are simply general effects of increased devel- opment of an organ or whether they are specifically and directly associated with the development of function in that organ. If the former possibility is correct, one might suppose that many other enzymes and chem- ical entities would vary in the same way during functional differentiation; if the lat- ter is true, one would expect the ontogeny of specific enzymes to be independent of other more general biochemical changes. The 548 ENERGY EXCHANGE AND ENZYME DEVELOPMENT DuRING EMBRYOGENESIS fact that various enzymes, even in the same organ, appear at different times and develop at independent rates indicates very strongly that functional maturation in a particular organ is associated with the development of particular enzymes (Boell, ’45, ’48; Boell and Shen, 50; Dumm and Levy, 49; Moog, ’52; Gustafson and Hasselberg, ’51). It should be emphasized, of course, that several enzymes directly concerned with the same activity would more than likely develop somewhat in parallel. In the nervous system, for example, it would be reasonable to as- sume that the mechanism for acetylcholine synthesis, i.e., the cholinacetylase system, and that for hydrolysis would develop to- gether. No studies on the development of cholinacetylase have so far been made. How- ever, in the synthesis of acetylcholine energy from ATP is required, and it is of interest to note, therefore, that there seems to be a fair degree of parallelism in the development of ATPase and cholinesterase in the central nervous system. This may be seen in the rat brain by a comparison of data for ATPase and cholinesterase provided by Potter, Schneider, and Liebl (45) and Metzler and Humm (51), respectively. It is also shown in the chick brain by comparison of Moog’s (’47) curve for ATPase and Nachmansohn’s (’39) for cholinesterase (Moog, °52). Similar ex- amples are represented by dipeptidase and aminopeptidase in the chick embryo (Levy and Palmer, 43) or by dipeptidase and tri- peptidase in Amblystoma mexicanum (Lgv- trup, 4536). There is a substantial body of data which shows that enzyme development and func- tional differentiation during development go hand in hand. But it should by no means be concluded that the development of an enzyme is In any way causally linked with the differ- entiation process with which it appears to be associated. Statements may nevertheless be found in the literature of embryology to the effect that specific enzymes, or the reactions catalyzed by them, control such processes as determination, morphogenesis, or differen- tiation. So far this has not been proven for any enzyme. (See the discussions by Holter, 49, p. 73, and Weiss, *53, pp. 173-4, on this question. ) CONCLUSION In this review an attempt has been made to give an account of some of the features of energy exchange during development. It is clear that a great deal is known about the over-all energy metabolism of the embryo. Furthermore, the enzymes involved and their relationships to metabolic processes are gradu- ally being revealed. But the crucial problem of how energy liberated is used by the em- bryo for “developmental work” has not been penetrated. It is of little comfort to note that other biological disciplines are plagued by the same problem—the translation of energy released into work done. In spite of the tre- mendous advances that have been made in the biochemistry and physiology of muscle, there still exists a gap between the system concerned with the synthesis and breakdown of high-energy bonds and the change in muscle proteins from the extended to the contracted state. And on the question of how selective secretion is accomplished against a gradient, even less is known. The major problem in embryology, as in physiology, is to provide a bridge between events at a molecular level and the struc- tural components in which they find expres- sion. The common aim of students of develop- ment, no matter toward what aspect their special interests or energies direct them, is to describe as fully as possible the complex and intricately coupled reactions and inter- actions—involved at all levels, from molecular to organismic—in the transformation of the egg into the differentiated individual. Chem- ical studies of development contribute to this end by providing one of the dimensions of description. REFERENCES Albaum, H. G., Novikoff, A. B., and Ogur,M. 1946 The development of the cytochrome oxidase and succinoxidase systems in the chick embryo. J. Biol. Chem., 765:125-130. , and Worley, L. G. 1942 The develop- ment of cytochrome oxidase in the chick embryo. J. Biol. Chem., 144:697-700. Allen, T. H. 1940 Enzymes in ontogenesis. XI. Cytochrome oxidase in relation to respiratory ac- tivity and growth of the grasshopper egg. J. Cell. & Comp. Physiol., 76:149-163. Amberson, W. R. 1928 The influence of oxygen tension upon the respiration of unicellular or- ganisms. Biol. Bull., 55:79-81. , and Armstrong, P. B. 1933 The respira- tory metabolism of Fundulus heteroclitus during embryonic development. J. Cell. & Comp. Phys- iol., 2:381-397. Andresen, N., Holter, H., and Zeuthen, E. 1944 The respiration of syncytia formed by abnormal development of Ciona eggs. Compt. Rend. Lab. Carlsberg., ser. chim., 25:67-85. Atlas, M. 1938 The rate of oxygen consumption of frogs during embryonic development and growth. Physiol. Zool., 77:278-291. ENERGY EXCHANGE AND ENZYME DEVELOPMENT DURING EMBRYOGENESIS 549 Ball, E. G., and Meyerhof, B. 1940 On the occur- rence of iron-porphyrin compounds and succinic dehydrogenase in marine organisms possessing the copper blood pigment hemocyanin. J. Biol. Chem., 134:483-493. Ballentine, R. 1938 Reducing activity of ferti- lized and unfertilized Arbacia eggs. Biol. Bull., 75:368. 1940 Analysis of the changes in respira- tory activity accompanying the fertilization of marine eggs. J. Cell. & Comp. Physiol., 75:217- 250" Banga, I. 1937 Uber den Mechanismus der Milchsaurebildung im Muskel. Zeit. f. Physiol. Chem., 249:209-210. Barnes, M.R. 1944 The metabolism of the devel- oping Rana pipiens as revealed by specific inhib- itors. J. Exp. Zool., 95:399-417. Barron, D. H. 1941 The functional development of some mammalian neuromuscular mechanisms. Biol. Rev., 76:1-33. Barth, L.G. 1942 Regional differences in oxygen consumption of the amphibian gastrula. Physiol. Zool., 15:30-46. 1946 Studies of the metabolism of devel- opment. J. Exp. Zool., 103:463-486. , and Barth, L. J. 1951 The relation of adenosine triphosphate to yolk utilization in the frog’s egg. J. Exp. Zool., 176:99-121. , and Jaeger, L. 1947a Phosphorylation in the frog’s egg. Physiol. Zool., 20:133-146. , and Jaeger, L. 1947b The apyrase activ- ity of various protein fractions of the frog’s egg. J. Cell. & Comp. Physiol., 30:111-130. , and Jaeger, L. 1950a The role of aden- osine-triphosphate in phosphate transfer from yolk to other proteins in the developing frog egg. I. General properties of the transfer system as a whole. J. Cell. & Comp. Physiol., 35:413-436. , and Jaeger, L. 1950b The role of aden- osine-triphosphate in phosphate transfer from yolk to other proteins in the developing frog egg. II. Separation of the system into component en- zymes, phosphate donor and phosphate acceptor. J. Cell. & Comp. Physiol., 35:437—-460. Beams, H. W., and King, R. L. 1936 Survival of Ascaris eggs after centrifuging. Science, 84:138. Berenblum, I., Chain, E., and Heatley, N.G. 1939 The study of metabolic activities of small amounts of surviving tissue. Bioch. J., 33:68-74. Bialascewicz, K., and Bledovski, R. 1915 Cited from Needham, ’31. Bodenstein, D. A., and Kondritzer, A. A. 1948 The effect of nitrogen mustard on nucleic acids during embryonic amphibian development. J. Exp. Zool., 707:109-121. Bodine, J. H. 1929 Factors influencing the rate of respiratory metabolism of a developing egg (Orthoptera). Physiol. Zool., 2:459-482. 1946 Uric acid formation in the develop- ing egg of the grasshopper, Melanoplus differen- tialis. Physiol. Zool., 19:54—57. , Allen, T. H., and Boell, E. J. 1937 En- zymes in ontogenesis. III. Activation of naturally occurring enzymes (tyrosinase). Proc. Soc. Exp. Biol. & Med., 37:450-453. , and Boell, E. J. 1934a Action of carbon monoxide on respiration of normal and blocked embryonic cells. J. Cell. & Comp. Physiol., 4:475- 482. , and Boell, E. J. 1934b Respiratory mechanisms of normally developing and blocked embryonic cells. J. Cell. & Comp. Physiol., 5:97- Lills: , and Boell, E. J. 1935 Enzymes in onto- genesis. I. Tyrosinase. J. Cell. & Comp. Physiol., 6:263-275. , and Boell, E. J. 1936a Enzymes in onto- genesis. II. The indophenol oxidase. J. Cell. & Comp. Physiol., 8:213-230. , and Boeli, E. J. 1936b Respiration of em- bryo versus egg (Orthoptera). J. Cell. & Comp. Physiol., 8:357-366. , and Boell, E. J. 1937 The action of cer- tain stimulating and inhibiting substances on the respiration of active and blocked eggs and isolated embryos. Physiol. Zool., 10:245-257. , and Boell, E. J. 1938 The influence of some dinitrophenols on respiratory metabolism during certain phases of active development. J. Cell. & Comp. Physiol., 77:41-63. , Lu, K-H., and West, W. L. 1952 Reduc- tion of triphenyltetrazolium chloride by mitoti- cally active and blocked embryonic cells. Biol. Bull., 702:16-21. Boell, E. J. 1935 Respiratory quotients during embryonic development. J. Cell. & Comp. Phys- iol., 6:369-385. 1942 Biochemical and physiological anal- ysis of organizer action. Growth (Suppl.), 7: 37-53. 1945 Functional differentiation in em- bryonic development. II. Respiration and cyto- chrome oxidase activity in Amblysioma puncta- tum. J. Exp. Zool., 100:331-352. 1946 The effect of sodium azide on Para- mecium calkinsi. Biol. Bull., 917:238-239. 1948 Biochemical differentiation during amphibian development. Ann. N. Y. Acad. Sci., 49:773-800. , Koch, H., and Needham, J. 1939 Mor- phogenesis and metabolism. IV. Respiratory quotient of the regions of the amphibian gastrula. Proc. Roy. Soc. London (B) 127:374— 387. , Needham, J., and Rogers, V. 1939 Mor- phogenesis and metabolism. I. Anaerobic gly- colysis of the regions of the amphibian gastrula. Proc. Roy. Soc. London (B), 127:322-356. , and Nicholas, J. S. 1948 Respiratory metabolism of the mammalian egg. J. Exp. Zool., 109:267-282. , and Poulson, D. F. 1939 The respiratory metabolism of normal and genetically deficient eggs of Drosophila melanogaster. Anat. Rec., 75 (Suppl.): 65-66. ,and Shen, S.C. 1949 Experimental mod- ification of cholinesterase development in the midbrain of Amblystoma punctatum. Anat. Rec., 105:490. , and Shen, S. C. 1950 Development of cholinesterase in the central nervous system of 550 ENERGY EXCHANGE AND ENZYME DEVELOPMENT DURING EMBRYOGENESIS Amblystoma punctatum. J. Exp. Zool., 113:583- 600. Bohr, C., and Hasselbalch, K. A. Needham, ’31. Boivin, A., Vendrely R., and Vendrely C. 1948 L’acide désoxyribonucléique du noyau cellulaire, dépositaire des caractéres héréditaires; arguments d’ordre analytique. Compt. Rend. Acad. Sci. Paris, 226:1061-1063. Booth, V.H. 1935 The identity of xanthine oxi- dase and the Schardinger enzyme. Bioch. J., 29: 1732-1748. Borei, H. 1948 Respiration of odcytes, unferti- lized eggs and fertilized eggs from Psammechinus and Asterias. Biol. Bull., 95:124-150. 1950 Cytochrome c in sea urchin eggs. Acta Chem. Scand., 4:1607—1608. Brachet, J. 1934 Etude du métabolisme de l’oeuf de grenouille (Rana fusca) au cours du dével- oppement. 1. La respiration et la glycolyse de la segmentation a l’éclosion. Arch. de Biol., 45:611- 727. 1903 Cited from 1938 The oxygen consumption of artifi- cially activated and fertilized Chaetopterus eggs. Biol. Bull., 74:93-98. 1939 Etude du métabolisme de l’oeuf de grenouille (Rana fusca) au cours du développe- ment. V. Le métabolisme protéique et hydrocar- boné de Voeuf en relation avec le probleme de Vorganizateur. Arch. de Biol., 50:233-267. 1947 Nucleic acids in the cell and em- bryo. Symp. Soc. Exp. Biol., 7:207-224. 1949 Discussion of paper by Holter. Publ. Staz. Zool. Napoli, 27 (Suppl.): 73. 1950 Chemical Embryology, translated by L. G. Barth. Interscience Publishers, New York. , and Needham, J. 1935 Etude du métab- olisme de l’oeuf de grenouille (Rana fusca) au cours du développement. IV. La teneur en glyco- gene de loeuf de la ségmentation a léclosion. Arch. de Biol., 46:821—835. Bragg, A. N. 1939 Observations upon amphib- ian deutoplasm and its relation to embryonic and early larval development. Biol. Bull., 77:268-283. Brody, S. 1945 Bioenergetics and Growth. Rein- hold Publishing Corp., New York. Buchanan, W. J. 1938 Developmental accelera- tion following inhibition. J. Exp. Zool., 79:109- 1972 1940 Developmental rate and alternating temperature. J. Exp. Zool., 83:235-248. Butler, J. A. V. 1946 Life and the second law of thermodynamics. Nature, 758:153-154. Caspersson, T. 1947 The relation between nu- cleic acids and protein synthesis. Symp. Soc. Exp. Biol., 7:127-151. Cleland, K. W., and Rothschild, Lord 1952a The metabolism of the sea-urchin egg. Anaerobic breakdown of carbohydrate. J. Exp. Biol., 29:285- 294. , and Rothschild, Lord 1952b The metab- olism of the sea-urchin egg. Oxidation of carbo- hydrate. J. Exp. Biol., 29:416-428. Clowes, G. H. A., Keltch, A. K., Strittmatter, C. F., and Walters, C. P. 1950 Action of nitro- and halophenols upon oxygen consumption and phos- phorylation by a cell-free particulate system from Arbacia eggs. J. Gen. Physiol., 33:555-561. , and Krahl, M. E. 1934 Action of dinitro compounds on sea urchin eggs. Science, 80:384— 385. , and Krahl, M. E. 1936 Studies on cell metabolism and cell division. I. On the relation between molecular structures, chemical proper- ties, and biological activities of the nitrophenols. J. Gen. Physiol., 20:145-171. Coghill, G. E. 1929 Anatomy and the Problem of Behavior. Cambridge University Press, Cam- bridge, England. Crescitelli, F. 1935 The respiratory metabolism of Galleria mellonella (bee moth) during pupal development at different constant temperatures. J. Cell. & Comp. Physiol., 6:351-368. Daniel, J. F., and Yarwood, KE. A. 1939 The early embryology of Triturus torosus. Univ. Calif. Publ. Zool., 43:321-356. Daniel, R. L. 1947 Distribution of glycogen in the developing salmon (Salmo salar L.). J. Exp. Biol., 24:123-144. Davidson, J. N., and Leslie, I. 1950 A new ap- proach in the biochemistry of growth and devel- opment. Nature, 765:49-53. Detwiler, S. 1946a Experiments upon the mid- brain of Amblystoma embryos. Am. J. Anat., 78: 115-138. 1946b A quantitative study of locomotion in larval Amblystoma following either midbrain or forebrain excision. J. Exp. Zool., 102:321-332. Deutsch, H. F., and Gustafson, T. 1952 The changes in catalase and cytochrome oxidase in de- veloping sea urchin eggs. Arkiv Kemi, 4:221-231. deVillafranca, G. W. 1953 An investigation of the distribution and development of adenosine- triphosphatase activity in developing rat muscle. Doctoral dissertation, Yale University. Dickens, F., and Greville, G. D. 1933a Metab- olism of normal and tumour tissue. VIII. Respira- tion in sugar-free media. Bioch. J., 27:832-841. , and Greville, G. D. 1933b Metabolism of normal and tumour tissue. IX. Ammonia and urea formation. Bioch. J., 27:1123-1133. , and Simer, F. 1930 The metabolism of normal and tumour tissue. II. The respiratory quotient, and the relationship of respiration to glycolysis. Bioch. J., 24:1301-1326. , and Simer, F. 1931 The metabolism of normal and tumour tissue. IV. The respiratory quotient in bicarbonate-media. Bioch. J., 25:985- 993. Dixon, M. 1951 Manometric Methods, 3d ed. Cambridge University Press, Cambridge, Eng- land. Dumm, M. E., and Levy, M. 1949 Chemistry of the chick embryo. VII. The accumulation of solids, nitrogen, lipids, and peptidase by the gizzard and liver of the chick embryo. J. Cell. & Comp. Physiol., 33:373-382. Elhott, K. A. C., and Greig, M. E. 1938 The dis- tribution of the succinic oxidase system in animal tissues. Bioch. J., 32:1407-1423. Fischer, F. G., and Hartwig, H. 1938 Vergleich- ENERGY EXCHANGE AND ENZYME DEVELOPMENT DuRING EMBRYOGENESIS 551 ende Messungen der Atmung des Amphibien- Keimes und seiner Teile wahrend der Entwick- lung. Biol. Zentralbl., 58:567-589. Fisher, K. C. 1941 The fractionation of respira- tion by the use of narcotics. Biol. Bull., 87;:282. , and Henry, R. J. 1940 The use of ure- thane as an indicator of “‘activity” metabolism in the sea urchin egg. Biol. Bull., 79:731-732. ,and Henry, R.J. 1944 The effects of ure- thane and chloral hydrate on oxygen consump- tion and cell division in the egg of the sea urchin, Arbacia punctulata. J. Gen. Physiol., 27:469-481. , Henry, R. J., and Low, E. 1944 The ef- fects of sulfanilamide and azide on oxygen con- sumption and cell division in the egg of the sea urchin, Arbacia punctulata. J. Pharmacol. & Exp. Therap., 87:58-66. , and Stern, J. R. 1942 The separation of an “activity” metabolism from the total respira- tion of yeast by the effects of ethyl carbamate. J. Cell. & Comp. Physiol., 79:109-122. Fiske, C. H., and Boyden, E. A. 1926 Nitrogen metabolism in the chick embryo. J. Biol. Chem., 70:535-556. Fitzgerald, L. R. 1949 The alkaline phosphatase of the developing grasshopper egg. J. Exp. Zool., 110:461-487. Flexner, J. B., and Flexner, L. B. 1948 Bio- chemical and physiological differentiation during embryonic development. VII. Adenylpyrophos- phatase and acid phosphatase activities in the de- veloping cerebral cortex and liver of the fetal guinea pig. J. Cell. & Comp. Physiol., 37:311-320. , and Flexner, L. B. 1951 Biochemical and physiological differentiation during embry- onic development. XIV. The nucleic acids of the developing cerebral cortex and liver of the fetal guinea pig. J. Cell. & Comp. Physiol., 38:1-16. , Flexner, L. B., and Straus, W. L., Jr. 1944 The oxygen consumption, cytochrome and cyto- chrome oxidase activity and histological structure of the developing cerebral cortex of the fetal pig. J. Cell. & Comp. Physiol., 78:355-368. Flexner, L. B., and Flexner, J.B. 1946 Biochem- ical and physiological differentiation during em- bryonic development. III. Succinic dehydro- genase and succinoxidase in the cerebral cortex of the fetal pig. J. Cell. & Comp. Physiol., 27: 35-42. Friedberg, F., and Eakin, R. M. 1949 Studies in protein metabolism of the amphibian embryo. I. Uptake of radioactive glycine. J. Exp. Zool., 110: 33-46. Fruton, J. S., and Simmonds, S. 1953 General Biochemistry. John Wiley & Sons, New York. Goldinger, J. M., and Barron, E.S.G. 1946 The pyruvate metabolism during the process of cell division. J. Gen. Physiol., 30:73-82. Gray, J. 1925 Cited from Needham, ’31. 1926 The growth of fish. I. The relation between the embryo and yolk in Salmo fario. J. Exp. Biol., 4:215-225. 1927 The mechanism of cell division. III. The relationship between cell division and growth in segmenting eggs. J. Exp. Biol., 4:313-321. 1929a The growth of fish. II. The growth-rate of the embryo of Salmo fario. J. Exp. Biol., 6:110-130. 1929b The kinetics of growth. J. Exp. Biol., 6:248-274. Gregg, J. R. 1948 Carbohydrate metabolism of normal and hybrid amphibian embryos. J. Exp. Zool., 109:119-134. , and Ballentine, R. 1946 Nitrogen metab- olism of Rana pipiens during embryonic develop- ment. J. Exp. Zool., 103:143-168. , and Lovtrup, S. 1950 Biochemical gradi- ents in the axolotl gastrula, Compt. Rend. Lab. Carlsberg, ser. chim., 27:307-324. , and Ornstein, N. 1952 Anaerobic am- monia production by amphibian gastrulae ex- plants. Biol. Bull., 702:22-24. , and Pomerat, C. M. 1942 The glycogen content of the embryo of Rana pipiens during de- velopment. Growth, 6:231-234. Greig, M. E., Munro, M. P., and Elliott, K. A. C. 1939 The metabolism of lactic and pyruvic acids in normal and tumour tissues. Bioch. J., 33:443- 453. Gustafson, T. 1952 Nitrogen Metabolism, En- zymic Activity, and Mitochondria Distribution in Relation to Differentiation in the Sea Urchin Egg. Almqvist and Wiksells, Uppsala. , and Hasselberg, I. 1951 Studies on en- zymes in the developing sea urchin egg. Exp. Cell Res., 2:642-672. , and Lenicque, P. 1952 Studies on mito- chondria in the developing sea urchin egg. Exp. Cell Res., 3:251-274. Harris, D. L. 1946 Phosphoproteinphosphatase, a new enzyme from the frog egg. J. Biol. Chem., 165:541-550. Harrison, R. G. 1935 The origin and develop- ment of the nervous system studied by the meth- ods of experimental embryology. Proc. Roy. Soc. London (B), 778:155-196. Hayes, F. A. 1938 The relation of the fat changes to the general chemical embryology of the sea urchin. Biol. Bull., 74:267-277. , and Hollett, A. 1940 The carbohydrate metabolism of developing salmon eggs. Canad. J. Res., 78 (D) :53-65. , Wilmot, I. R., and Livingstone, D. A. 1951 The oxygen consumption of the salmon egg in relation to development and activity. J. Exp. Zool., 116:377-395. Heatley, N. G., and Lindahl, P. E. 1937 Studies on the nature of the amphibian organization cen- ter. 5. Distribution and nature of glycogen in the amphibian embryo. Proc. Roy. Soc. London (B), 122:395-402. Henry, R. J., and Henry, M.D. 1945 The effect of penicillin on eggs of the sea urchin, Arbacia punctulata. J. Gen. Physiol., 28:405—-413. Herrmann, H. 1953 Biochemistry of organo- genesis. Arch. Néerl. de Zool., 70 (Suppl.):127- 143. , and Nicholas, J. S. 1948a Enzymatic liberation of organic phosphate from adenosine- triphosphate in developing rat muscle. J. Exp. Zool., 107:177-181. , and Nicholas, J. S. 1948b Quantitative 552 ENERGY EXCHANGE AND ENZYME DEVELOPMENT DURING EMBRYOGENESIS changes in muscle protein fractions during rat development. J. Exp. Zool., 107:165-176. Herrmann, H., and Nicholas, J.S. 1949 Nucleic acid content of whole homogenates and of frac- tions of developing rat muscle. J. Exp. Zool., 112: 341-359. , Nicholas, J. S., and Vosgian, M. E. 1949 Liberation of inorganic phosphate from adeno- sinetriphosphate by fractions derived from de- veloping rat muscle. Proc. Soc. Exp. Biol. & Med., 7 2:454457. Hill, D. L. 1945 Carbohydrate metabolism dur- ing embryonic development (Orthoptera). J. Cell. & Comp. Physiol., 25:205-216. Horstadius, S. 1949 Experimental researches on the developmental physiology of the sea urchin. Publ. Staz. Zool. Napoli, 27 (Suppl.): 131-172. Hoff-Jorgensen, E., and Zeuthen, E. 1952 Evi- dence of cytoplasmic deoxyribosides in the frog’s egg. Nature, 169:245-246. Holter, H. 1949 Problems of enzyme localiza- tion and development. Publ. Staz. Zool. Napoli, 21 (Suppl.): 60-76. Holtfreter, J. 1946 Experiments on the formed inclusions of the amphibian egg. I. The effect of pH and electrolytes on yolk and lipochondria. J. Exp. Zool., 101:355—-405. Hopkins, H. S., and Handford, S. W. 1943 Re- spiratory metabolism during development in two species of Amblystoma. J. Exp. Zool., 93:403—-414. Horowitz, N. H. 1940 The respiratory metabo- lism of the developing eggs of Urechis caupo. J. Cell. & Comp. Physiol., 75:229-308. Huff, G. C., and Boell, E.J. 1936 Effect of ultra- centrifuging on oxygen consumption of the eggs of Ascaris suum, Goeze. Proc. Soc. Exp. Biol. & Med., 34:626-628. Hultin, T. 1953 Metabolism and determination. Arch. Néerl. de Zool., 70 (Suppl.) :76-91. Hutchens, J. O., Kopac, M. J., and Krahl, M. E. 1942 The cytochrome oxidase content of cen- trifugally separated fractions of unfertilized Ar- bacia eggs. J. Cell. & Comp. Physiol., 20:113-116. » Keltch, A., Krahl, M. E., and Clowes, G. H. A. 1942 Studies on cell metabolism and cell division. VI. Observations on the glycogen con- tent, carbohydrate consumption, lactic acid pro- duction, and ammonia production of eggs of Ar- bacia punctulata. J. Gen. Physiol., 25:717-731. Jaeger, L. 1945 Glycogen utilization by the am- phibian gastrula in relation to invagination and induction. J. Cell. & Comp. Physiol., 25:97-120. Jandorf, B. J., and Krahl, M. E. 1942 Studies on cell metabolism and cell division. VIII. The di- phosphopyridine nucleotide (cozymase) content of eggs of Arbacia punctulata. J. Gen. Physiol., 25:749-754, Keltch, A. K., Strittmatter, C. F., Walters, C. P., and Clowes, G. H. A. 1950 Oxidative phos- phorylation by a cell free particulate system from unfertilized Arbacia eggs. J. Gen. Physiol., 33: 547-554. Kielley, W. W., and Meyerhof, O. 1948 Studies on ATPase of muscle. Il. A new Mg-activated adenosinetriphosphatase. J. Biol. Chem., 176:591- 601. Kitching, J. A., and Moser, F. 1940 Studies on a cortical layer response to stimulating agents in the Arbacia egg. IV. Response to chemical and physical agents in the absence of oxygen and ob- servations of the effects of low oxygen tensions and high hydrostatic pressures upon amoeboid eggs. Biol. Bull., 78:80-91. Krahl, M.E. 1950 Metabolic activities and cleav- age of eggs of the sea urchin, Arbacia punctulata. A review, 1932-1949. Biol. Bull., 98:175-217. , Keltch, A. K., Neubeck, C. E., and Clowes, G. H. A. 1941 Cell metabolism and cell divi- sion. V. Cytochrome oxidase activity in the eggs of Arbacia punctulata. J. Gen. Physiol., 24:597- 617. Krugelis, E. J., Nicholas, J. S., and Vosgian, M. E. 1952 Alkaline phosphatase activity and nucleic acids during embryonic development of Amblys- toma punctatum at different temperatures. J. Exp. Zool., 7217:489-504. Kutsky, P. B. 1950 Phosphate metabolism in the early development of Rana pipiens. J. Exp. Zool., 115:429-460. Laser, H., and Rothschild, Lord 1939 The metab- olism of the eggs of Psammechinus miliaris dur- ing the fertilization reaction. Proc. Roy. Soc. Lon- don (B), 126:539-557. Lennerstrand, A. 1933 Aerobe und anaerobe Glykolyse bei der Entwicklung des Froscheies (R. temporaria). Zeit. vergl. Physiol., 20:287- 290. Levy, M. 1952 Metabolic patterns in embryonic development. Ann. N. Y. Acad. Sci., 55:51-56. , and Palmer, A. H. 1943 Chemistry of the chick embryo. IV. Aminopeptidase. J. Biol. Chem., 750:271-279. , and Young, N. F. 1948 Chemistry of the chick embryo. V. Accumulation of cytochrome oxidase. J. Biol. Chem., 175:73-77. Lindahl, P. E. 1939 Zur Kenntnis der Entwick- lungsphysiologie des Seeigeleies. Zeit. vergl. Physiol., 27:233-250. 1940 Uber die CN-resistante Atmung des Seeigeleies. Arkiv for Kemi, Mineral. och Geol., 14A:1-31. 1953 Somatic reduction division in the development of the sea urchin. Nature, 1771:437. Lindberg, O. 1943 Studien iiber das Problem des Kohlehydratabbaus und der Saurebildung bei der Befruchtung des Seeigeleies. Arkiv fér Kemi, Mineral. och Geol., 76A:1-20. , and Ernster, L. 1948 On carbohydrate metabolism in homogenized sea urchin eggs. Bioch. Biophys. Acta, 2:471-477. Linderstrom-Lang, K. 1946 Periodic metabolism and diffusion. Compt. Rend. Lab. Carlsberg, ser. chim., 25:229-272. Loeb, J. 1895 Untersuchungen iiber die physiol- ogischen Wirkungen des _ Sauerstoffmangels. Pfliig. Arch., 62:249-294. Lovtrup, S. 1953a Energy sources of amphibian embryogenesis. Compt. Rend. Lab. Carlsberg, ser. chim., 28:371-399. 1953b Utilization of reserve material during amphibian embryogenesis at different temperatures. Compt. Rend. Lab. Carlsberg, ser. chim., 28:400-425. 1953c Changes in the content of pep- ENERGY EXCHANGE AND ENZYME DEVELOPMENT DURING EMBRYOGENESIS 553 tidases during amphibian embryogenesis at dif- ferent temperatures. Compt. Rend. Lab. Carls- berg, ser. chim., 28:426-443. Ludwig, D. 1950a The metabolism of starved nymphs of the grasshopper, Chortophaga viridi- fasciata, De Geer. Physiol. Zool., 23:41-47. 1950b Changes in the distribution of nitrogen during starvation in the grasshopper, Chortophaga viridifasciata, De Geer. Physiol. Zool., 23:208-213. Lyon, E. P. 1904 Cited from Needham, ’31. Metzler, C. J.. and Humm, D. G. 1951 The de- termination of cholinesterase activity in the whole brains of developing rats. Science, 1/3:382-383. Meyerhof, O. 1911 Untersuchungen iiber die Warmeténung der vitalen Oxydationsvorgange in Eiern. Bioch. Zeit., 35:279-315, 316-328. Mirsky, A. E., and Ris, H. 1949 Variable and constant components of chromosomes. Nature, 163:666-667. Moog, F. 1943 Cytochrome oxidase in early chick embryos. J. Cell. & Comp. Physiol., 22:223-231. 1944a The chloretone sensitivity of frogs’ eggs in relation to respiration and development. J. Cell. & Comp. Physiol., 23:131-155. 1944b Localizations of alkaline and acid phosphatase in the early embryogenesis of the chick. Biol. Bull., 86:51-80. 1947 Adenyl pyrophosphatase in brain, liver, heart, and skeletal muscle of chick embryos and hatched chicks. J. Exp. Zool., 705:209-220. 1950 The functional differentiation of the small intestine. I. The accumulation of al- kaline phosphomonoesterase in the duodenum of the chick. J. Exp. Zool., 775:109-129. 1951 The functional differentiation of the small intestine. II. The differentiation of al- kaline phosphomonoesterase in the duodenum of the mouse. J. Exp. Zool., 778:187-208. 1952 The differentiation of enzymes in relation to the functional activities of the develop- ing embryo. Ann. N. Y. Acad. Sci., 55:57-66. 1953. The functional differentiation of the small intestine. III. The influence of the pitu- itary-adrenal system on the differentiation of phosphatase of the suckling mouse. J. Exp. Zool., 124:329-346. , and Steinbach, H. B. 1945 Adenylpyro- phosphatase in chick embryos. J. Cell. & Comp. Physiol., 25:133-144. Moore, J. A. 1941 Developmental rate of hybrid frogs. J. Exp. Zool., 86:405-422. 1946 Studies in the development of frog hybrids. J. Exp. Zool., 107:173-219. Morgan, E. J. 1930 Xanthine oxidase in the avian embryo. Bioch. J., 24:410—-414. Nachmansohn, D. 1939 Cholinesterase dans le systeme nerveux central. Boll. Soc. Chim. Biol. (Paris), 21:761-796. Needham, J. 1931 Chemical Embryology. Cam- bridge University Press, Cambridge, England. 1932a On the true metabolic rate of the chick embryo and the respiration of its mem- branes. Proc. Roy. Soc. London (B), 110:46-74. 1932b A manometric analysis of the metabolism in avian ontogenesis. II. The effects of fluoride, iodoacetate, and other reagents on the respiration of blastoderm, embryo, and yolk sac. Proc. Roy. Soc. London (B), 772:114-138. 1933. The energy sources in ontogenesis. VII. The respiratory quotient of developing crustacean embryos. J. Exp. Biol., 70:79-87. 1942 Biochemistry and Morphogenesis. Cambridge University Press, Cambridge, Eng- land. , and Nowinski, W. W. 1937 Intermedi- ary carbohydrate metabolism in embryonic life. I. General aspects of anaerobic glycolysis. Bioch. J., 31:1165-1184. Negelein, E. 1925 Uber die glykolytische Wir- kung des embryonalen Gewebes. Bioch. Zeit., 165:122-1 34. Nolf, L.O. 1932 Experimental studies on certain factors influencing the development and viability of the ova of the human Trichurus as compared with those of the human Ascaris. Am. J. Hyg., 16:288-322. Novikoff, A. B., and Potter, V. R. 1948 Changes in nucleic acid concentration during the develop- ment of the chick embryo. J. Biol. Chem., 773: 233-238. , Potter, V. R., and Le Page, G. A. 1948 Phosphorylating glycolysis in the early chick em- bryo. J. Biol. Chem., 173:239-252. Ohman, L. O. 1940 Uber die Veranderung des respiratorischen Quotienten wahrend der Friih- entwicklung des Seeigeleies. Arkiv fdr Zool., 32A:1-9. Parnas, J. K., and Krasinka, Z. 1921 Ueber den Stoffwechsel der Amphibienlarven. Bioch. Zeit., 116:108-137. Philips, F. S. 1940 Oxygen consumption and its inhibition in the development of Fundulus and various pelagic eggs. Biol. Bull., 78:256-273. 1941 The oxygen consumption of the early chick embryo at various stages of develop- ment. J. Exp. Zool., 86:257-289. 1942 Comparison of the respiratory rates of different regions of the chick blastoderm during early stages of development. J. Exp. Zool., 90:83- 100. Pollister, A. W. 1954 Cytochemical aspects of protein synthesis; in Dynamics of Growth Proc- esses, edited by E. J. Boell, pp. 33-67. Princeton University Press, Princeton, New Jersey. Porter, K. R. 1954 Cell and tissue differentiation in relation to growth; in Dynamics of Growth Processes, edited by E. J. Boell, pp. 95-110. Princeton University Press, Princeton, New Jersey. Potter, V. R., and DuBois, K.P. 1942 The quanti- tative determination of cytochrome c. J. Biol. Chem., 742:417—426. , Schneider, W. C., and Liebl, G. J. 1945 Enzyme changes during growth and differentia- tion in the tissues of the new born rat. Cancer Res., 521-24. Raven, C. P. 1935 Experimentelle Untersuch- ungen iiber den Glykogen-Stoffwechsel des Or- ganizationszentrums in der Amphibiengastrula. Proc. Konin. Akad. Wetensch., 38:1107-1109. Reddy, D. V. N., Lombardo, M. E., and Cerecedo, 554 ENERGY EXCHANGE AND ENZYME DEVELOPMENT DuRING EMBRYOGENESIS L.R. 1952 Nucleic acid changes during devel- opment of the chick embryo. J. Biol. Chem., 198: 267-270. Robbie, W. A. 1946 The effect of cyanide on the oxygen consumption and cleavage of the sea urchin egg. J. Cell. & Comp. Physiol., 28:305- 324, , Boell, E. J., and Bodine, J. H. 1938 A study of the mechanism of cyanide inhibition. I. Effect of concentration on the egg of Melanoplus differentialis. Physiol. Zool., 11:54-61. Romanoff, A. L. 1941 The study of the respira- tory behavior of individual chicken embryos. J. Cell. & Comp. Physiol., 78:199-214. 1943 Differentiation in respiratory ac- tivity of isolated embryonic tissues. J. Exp. Zool., 93:1-26. Rothschild, Lord 1939 Effect of phlorizin on the metabolism of cytolyzing sea urchin eggs. J. Exp. Biol., 16:49-55. 1948 The physiology of Echinus escu- lentus spermatozoa. J. Exp. Biol., 25:15-21. , and Cleland, K. W. 1952 The physiology of sea urchin spermatozoa: The nature and local- ization of the endogenous substrate. J. Exp. Biol., 29:66-71. Runnstrom, J. 1930 Atmungsmechanismus und Entwicklungserregung bei dem Seeigelei. Proto- plasma, 70:106-173. 1932 Die Beeinflussung der Atmung und Spaltung im Seeigelei durch Dimethylparaphen- ylenediamine und Hydrochinon. Protoplasma, 15:532-565. 1933 Zur Kenntnis der Stoffwechselvor- gange bei der Entwicklungserregung des Seei- geleies. Bioch. Zeit, 258:257-279. 1934 Stoffwechselvorgange wahrend der ersten Mitose des Seeigeleies. Protoplasma, 20: 1-10. Sanborn, R. C., and Williams, C. M. 1950 The cytochrome system in the cecropia silkworm, with special reference to the properties of a new component. J. Gen. Physiol., 33:579-588. Sawyer, C. H. 1943 Cholinesterase and the be- havior problem. I. The relationship between the development of the enzyme and early motility. II. The effects of inhibiting cholinesterase. J. Exp. Zool., 92:1—29. Scholander, P. F., Claff, C. L., Sveinson, S. L., and Scholander, S. I. 1952 Respiratory studies on single cells. III. Oxygen consumption during cell division. Biol. Bull., 702:185-199. Shen, S.C. 1949 Development of respiratory en- zymes in rat muscle. Anat. Rec., 105:489. 1954 Enzyme development as ontogeny of specific proteins (in press). Sholl, D. A. 1954 Regularities in growth curves, including rhythms and allometry; in Dynamics of Growth Processes, edited by E. J. Boell, pp. 224-241. Princeton University Press, Princeton, New Jersey. Sliffer, E. H. 1930 Insect development. I. Fatty acids in the grasshopper egg. Physiol. Zool., 3: 503-518. 1932 Insect development. III. Blastoki- nesis in the living grasshopper egg. Biol. Zen- tralbl., 52:223-299. Smith, S. 1946 Studies in the development of the rainbow trout (Salmo irideus). I. The heat pro- duction and nitrogen excretion. J. Exp. Biol., 23: 357-378. 1952 Studies in the development of the rainbow trout (Salmo irideus). Il. The metab- olism of carbohydrates and fats. J. Exp. Biol., 29: 650-666. Spiegelman, S., and Moog, F. 1945 A compari- son of the effects of cyanide and azide on the de- velopment of frogs’ eggs. Biol. Bull., 89:122-130. , and Steinbach, H. B. 1945 Substrate- enzyme orientation during embryonic develop- ment. Biol. Bull., 88:254—268. Spratt, N. T., Jr. 1948 Development of the early chick blastoderm on synthetic media. J. Exp. Zool., 107:39-64: 1950 Nutritional requirements of the early chick embryo. II. Differential requirements for morphogenesis and differentiation of the heart and brain. J. Exp. Zool., 1174:375-402. 1952 Reducing enzyme (dehydrogenase) systems in unincubated and one day old chick embryos. Anat. Rec., 772:434. Stefanelli, A. E. 1937 A new form of micro- respirometer, with a note on the effect of cleav- age on the respiration of the eggs of Rana. J. Exp. Biol., 74:171-177. Stotz, E. 1939 The estimation and distribution of cytochrome oxidase and cytochrome c in rat tissue. J. Biol. Chem., 737:555-565. Taylor, I. R., and Crescitelli, F. 1937 Measure- ment of heat production of small organisms. J. Cell. & Comp. Physiol., 70:93-112. Taylor, K. M., and Schechtman, A. M. 1949 In vitro development of the early chick embryo in the absence of small organic molecules. J. Exp. Look. 114-927-253: Ten Cate, G. 1953 The intrinsic development of amphibian embryos. Doctoral dissertation. North- Holland Publishing Co., Amsterdam. Thompson, D’Arcy W. 1942 On Growth and Form. Cambridge University Press, Cambridge, England. Tuft, P. 1953 Energy changes in development. Arch. Néerl. de Zool., 70 (Suppl.): 59-75. Tyler, A. 1942 Developmental processes and energetics. Quart. Rev. Biol., 77:197-212, 339- 353: , and Horowitz, N. H. 1938a The activ- ities of various substituted phenols in stimulating the respiration of sea urchin eggs. Biol. Bull., 75: 209-223. , and Horowitz, N. H. 1938b On the ener- getics of differentiation. VII. Comparison of the respiratory rates of parthenogenetic and ferti- lized Urechis eggs. Biol. Bull., 74:99-107. Vendrely, R., and Vendrely, C. 1949 La teneur de noyau cellulaire en acide désoxyribonucléique a travers les organes, les individus et les espéces animales. Experientia, 5:327—329. Vles, F. 1922 Cited from Needham, ’31. Warburg, O. 1908 Beobachtungen iiber die Oxy- ENERGY EXCHANGE AND ENZYME DEVELOPMENT DURING EMBRYOGENESIS 555 dations-Prozesse vom Seeigelei. Zeit. Physiol. Chem., 57:1-16. Warburg, O. 1910 Uber die Oxydationen in lebenden Zellen nach Versuchen am Seeigelei. Zeit. Physiol. Chem., 66:305—340. , Posener, K., and Negelein, E. 1924 Uber den Stoffwechsel der Carcinomzelle. Bioch. Zeit., 152:309-344. Weiss, P. 1953 Summary comments at the Sym- posium on the Biochemical and Structural Basis of Morphogenesis. Arch. Neéerl. de Zool., 70 (Suppl.) :165—-176. Whitaker, D. M. 1933 On the rate of oxygen con- sumption by fertilized and unfertilized eggs. V. Comparisons and interpretations. J. Gen. Physiol., 16:497-528. Wills, I. A. 1936 The respiratory rate of devel- oping amphibia with special reference to sex dif- ferentiation. J. Exp. Zool., 73:481-510. Woerdeman, M. W. 1933 Uber die chemischen Prozesse der embryonalen Induktion. Proc. Konin. Akad. Wetensch., 36:842-849. Yaoi, H. 1928-29 Glutathione, cytochrome and hydrogen ion concentration in developing chick embryos. Jap. J. Exp. Med., 7:135-143. Youngstrom, K. A. 1938 On the relationship be- tween cholinesterase and the development of be- havior in amphibia. J. Neurophysiol., 7:357—363. Zeuthen, E. 1946 Oxygen uptake during mitosis. Experiments on the eggs of the frog (Rana pla- tyrrhina). Compt. Rend. Lab. Carlsberg, ser. chim., 25:191-228. 1950a_ Respiration during cell division in the egg of the sea urchin Psammechinus miliaris. Biol. Bull., 98:144-151. 1950b Respiration and cell division in the egg of Urechis caupo. Biol. Bull., 98:152-160. 1951 Segmentation, nuclear growth and cytoplasmic storage in eggs of echinoderms and amphibians. Publ. Staz. Zool. Napoli, 23:47-69. 1953 Biochemistry and metabolism of cleavage in the sea urchin egg, as resolved into its mitotic steps. Arch. Néerl. de Zool., 70 (Suppl.): 31-58. Section 1X ONTOGENY OF IMMUNOLOGICAL PROPERTIES* ALBERT TYLER ANTIGENS IN DEVELOPMENT SOME GENERAL REMARKS THE vARrIous tissues and fluids of an adult organism contain a large assemblage of anti- gens. Of these the serum proteins and red blood corpuscles of mammals have been most extensively studied immunologically, but a large number of investigations have also been performed with soluble and insoluble constituents of various tissue cells. As a gen- eral statement one may say that the diverse antigens exhibit properties of species-specific- ity and tissue-specificity to various degrees. Certain serum or tissue-proteins show a high degree of species-specificity in that the anti- sera that are prepared against these materials of one species of animal fail to react or give weaker reactions with similar preparations from other species, depending somewhat on the degree ot phylogenetic relationship. Certain antigens termed Forssman or hetero- genetic antigens, however, are widely and irregularly distributed in various tissues and species of animals. Similarly, immunization with extracts of one kind of tissue yields antisera that react to various degrees with corresponding preparations from other tissues of the same animal. Where extensive cross- reactions occur specific components of the tissue may nevertheless be demonstrated by absorption of the antiserum with the prepara- tion from the heterologous tissue. This pro- cedure removes the antibodies for such anti- gens as may be common to various tissues. It is often supposed that if the chemically different proteins of the different tissues were prepared in pure form no _ cross-reactions would be obtained. This has, however, not as yet been established experimentally. The difficulties here are due to the ability of * A presentation of the fundamentals of immunol- ogy is beyond the scope of this chapter. The unin- formed reader should, therefore, consult a recent text such as Boyd’s (747) or Kabat and Mayer’s (48). 556 relatively small amounts of antigen to give rise to considerable amounts of antibody, and to the uncertainties involved in specifying the degree of purity of various proteins and other antigenic substances. As an illustration of this, some recent work by Cohn, Wetter and Deutsch (49) on the proteins of chicken egg-white may be cited. They studied the precipitation of ovalbumin and conalbumin by antibodies produced in rabbits and horses. The ovalbumin was recrystallized six times and found to be of homogeneous molecular size upon sedimentation-velocity and diffu- sion tests. However, it showed two main components upon electrophoresis. The conal- bumin preparation was _ ultracentrifugally and electrophoretically homogeneous. The electrophoretic patterns (Fig. 205) showed neither conalbumin in the ovalbumin prepa- ration nor the reverse. Nevertheless, the antisera prepared against ovalbumin reacted also with conalbumin. On the other hand, the two ovalbumin components detected electrophoretically were indistinguishable immunologically. From quantitative studies of the various reactions of these preparations these workers conclude that there are anti- genically active impurities in these prepara- tions that are not revealed by the various physicochemical tests. It is, of course, well known that there are no absolute criteria of purity even for simple chemical substances and that crystallizability does not necessarily mean molecular homo- geneity. For protein preparations modern methods of determining homogeneity, in regard to different properties, include elec- trophoresis, sedimentation, diffusion and sol- ubility, and each has certain limits of sen- sitivity. Experiments of the type cited above tend to show that immunological methods may be more sensitive for the detection of impurities in protein preparations. However, there is some uncertainty that impurities are really being detected by these methods. It would be necessary to show in the first place ONTOGENY OF IMMUNOLOGICAL PROPERTIES 557 that the maximum amount of impurity that might be present in a particular preparation and still be physicochemically undetected is sufficient to induce the formation of the cross-reacting antibodies that are obtained A Fig. 205. Electrophoretic patterns of highly puri- fied conalbumin (A) and of six times recrystallized ovalbumin (B and C) (from Cohn, Wetter and Deutsch, *49). upon immunization. No clear-cut demonstra- tion of this has been made in studies of cross- reactions as far as the present author is aware. For this purpose it would presumably be necessary to run parallel immunizations H. R H Protein Hz 0; As N=N+Cl + Ht Diazotized Arsanilic Acid Va nina roniaroial to Negi te | portant to test for possible competition of antigens and for adjuvant action. In experi- ments by Vaughan and Kabat (’53), showing the presence of cross-reacting antibodies after immunization with minimal amounts of highly purified ovalbumin, the quantities of impurities that might be present would not have been presumed to induce antibody for- mation. However, they interpret their results as due to trace contaminants in their oval- bumin antigen. Another way of interpreting cross-reactions is on the basis of similar determinant, or combining, groups on chemically different proteins or other antigenic substances. This can best be illustrated by the fundamental experiments of Landsteiner (’17-°46) in which proteins are coupled with small molecular, chemically well defined substances. Thus horse serum protein can be coupled with diazotized arsanilic acid (Atoxyl) as illus- trated in Figure 206, the union being pre- sumably with certain amino acids of the protein that have benzene or heterocyclic rings, such as tyrosine, histidine or trypto- phane. When this is injected into a rabbit the antiserum that is obtained is found to react not only with the immunizing antigen but also with other proteins, such as chicken serum protein, that have been similarly coupled with diazotized Atoxyl. On the other hand it will not precipitate chicken serum Diazotized Arsanilic Acid Fig. 206. Illustration of the manner in which a hapten may be coupled with a protein. with the maximum possible amounts of the suspected impurities and determine whether or not the amounts of antibody produced corresponded to those indicated by the cross- reactions. At the same time it would be im- protein that has been coupled with other sumple substances, such as diazotized sul- fanilic acid. The chemically introduced group is, then, a determinant of specificity. It is also termed a hapten. A simple hapten 558 ONTOGENY OF IMMUNOLOGICAL PROPERTIES cannot by itself induce antibody formation may depend upon relatively small determi- or precipitate antibodies produced against the nant groups rather than on the whole struc- conjugated protein. * However, it can combine ture of the molecule, and that the molecule with the antibodies to form soluble com- may contain more than one kind of determi- plexes and thereby inhibit the antibodies nant group. Unfortunately very little is from precipitating with the conjugated pro- known about the size, number and kind of tein. While the specificity of these reactions determinant groups of natural proteins. Ex- is so great that stereo-isomers can be dis- periments by Landsteiner (42) with hydrol- tinguished, cross-reactions are also obtained ysates of silk protein have shown that pep- with various haptens. Extensive studies of tides of molecular weight of 600 to 1000 can these have shown that the extent of cross- specifically inhibit the precipitation of the reaction is dependent upon the degree of intact protein by its antiserum. For the pres- similarity in size, shape and constitution of ent we can only conclude by analogy with the “working end” of the hapten. the results of experiments on the coupled When an azoprotein is used for immuniza- proteins that natural proteins may induce tion the antiserum generally contains anti- the formation of more than one kind of anti- bodies directed against the uncoupled protein body. Some of the antibody molecules would as well as those against the hapten. It is be specific for one determinant group, some Antigen Antibodies A ( \ B B anti- AB A Fig. 207. Diagram illustrating the three kinds of antibody molecules that may be produced in response to immunization with an antigen containing two different kinds of determinant groups. clear, then, that a single kind of protein can for another, and some for combinations of give rise to more than one kind of antibody. determinant groups. If two different proteins This is illustrated further when two different possess a common determinant group, then haptens (A and B, see Fig. 207) are coupled the antiserum induced by one of the proteins with a protein. Thus Haurowitz and will cross-react with the other. Schwerin (’43) obtained distinct anti-A and When antisera for two natural proteins of anti-B antibodies upon injection of such an diverse origin cross-react there are, then, at antigen. They did not find a third possible least three possible interpretations. One is type, anti-AB. However, Dodd (’52) presents that, of the various kinds of determinants strong evidence for this type of antibody in of the two proteins, one or more may be group O humans immunized with mixed A identical. Another is that various regions of and B antigens, as well as for its occurrence the two proteins may possess some degree in normal human group O sera (see also of structural similarity. The third is the pos- Bird, °53). A less satisfactory interpretation sible presence of one of the proteins as a is that some anti-A and anti-B antibodies in chemically undetected contaminant in prep- group O sera are more cross-reacting than arations of the other. This uncertainty as to are others. interpretation must be taken into account in The work with haptens serves to emphasize attempting to analyze investigations on the that in natural proteins antigenic specificity | antigens of adult and embryonic tissues, since cross-reactions are very commonly ob- * A complex hapten, formed by coupling more 3 i : 3 2 Don aati! tained with such material. In other words, than one molecule with a simple substance such as resorcinol, can form a precipitate with thé anti- it is important to know what is being detected bodies produced by injection of the azo-protein (see by the antiserum that is employed. To some Landsteiner and van der Scheer, ’32, ’33; Pauling extent the difficulties are overcome by the et al., 41). use of the methods of absorption whereby ONTOGENY OF IMMUNOLOGICAL PROPERTIES 559 cross-reacting antibodies are removed from an antiserum by precipitation with the heter- ologous antigen. It should also be emphasized again that antigens that behave alike serologically are not necessarily identical in over-all chemical constitution. Similarity in serological be- havior can be attributed to identity or close structural resemblance of the so-called de- terminants or combining groups of the anti- gens. For convenience in the following dis- cussion the terms determinants or combining groups will, in general, be omitted, it being understood that when two antigens are desig- nated identical or similar the designation refers primarily to these groups. RECENT INVESTIGATIONS ON SALINE EXTRACTS OF EMBRYOS The work of various early investigators gave rather inconsistent results concerning the antigens of embryos and adults. Accord- ing to some (e.g., Graham-Smith, ’04; Braus, 06; Dunbar, ’10; Uhlenhuth et al., ’10, ’39), embryo and adult seemed to possess no anti- gens in common, whereas others (e.g., Réssle, 05; Kritchewsky, 714, ’23; Wilkoewitz and Ziegenspeck, ’°28) found more or less exten- sive cross-reactions. The divergent results were evidently not due to species differences but were probably largely due to technical difficulties. Much more consistent results have been obtained by the recent investiga- tors in this field. The work of Burke et al. (44), Cooper (746, °48, °50), Schechtman (47, °48, °52), Nace and Schechtman (48), Perlmann and Gustafson (748), Maculla (48a,b), Ebert (50, °51, 52), Ten Cate and van Doorenmaalen (750), Flickinger and Nace (752), Nace (53), Spar (53), Clayton (53), Perlmann (753), Telfer and Williams (53) and Telfer (54) is in agreement in show- ing that eggs, embryos and adults possess cer- tain antigens in common. In addition, mainly by means of absorption technique, these work- ers find antigenic differences and have studied to some extent the changes in antigenicity during development. Some of their results are presented here. Burke et al. prepared antisera in rabbits against saline extracts of various organs of adult and embryonic chickens. With the ex- ceptions of the lens the extracts of the organs studied (brain, testis, ovary, kidney, liver) showed cross-reactions with the various anti- sera. The cross reactivity of the antisera could be more or less completely removed by absorption with heterologous extracts (sedi- mentable constituents obtained after removal of large tissue fragments were used for ab- sorption). When an antiserum against adult brain was thus absorbed and tested with embryo brain extracts, reactions were ob- tained only with embryos of 300 hours (pre- cipitin test) and 260 hours (complement fixation test) of incubation or older. When an antiserum against 312-hour brain was absorbed with heterologous extracts of other 312-hour organs, it gave reactions (precipitin tests) with brain extract of embryos of 160 hours or older. After absorption of this anti- serum with adult brain extract (sediment) it still reacted with the 160-hour extract. Ac- cording to these experiments, then, antigens of the adult brain first arise at about 260 hours. Earlier embryonic brain seems to possess antigens common to later stages but not to the adult brain. Somewhat similar results were obtained by Burke et al. on the lens (see Table 22). The antiserum against adult lens proved to be rather organ specific and was, therefore, not absorbed. When tested with lens extracts of embryos of various stages, starting with 72 hours, reactions (complement fixation) were obtained only after 160 hours and reached adult intensity at 330 hours. With a 300-hour lens antiserum reactions were obtained at 120 hours. With a 160-hour antiserum reactions were obtained at 96 hours. Thus, again, there is indication of antigens disappearing while new ones arise during the development of an organ. From these and similar tests with other or- gans Burke et al. specify the approximate time at which adult organ antigens appear as follows: lens, 146 hours; erythrocytes, 100 hours; kidney, 220 hours; brain, testis and ovary, 260 hours. It would appear, then, that adult organ specificity does not arise until a considerable time after the initial morpho- logical differentiation. This conclusion of Burke et al. is contra- dicted by the results of Schechtman (’48) and Ebert (50) indicating the presence of organ (brain, heart, spleen) antigens in the early chick blastoderm, and by the findings of Ten Cate and Van Doorenmaalen (50), who have detected adult lens antigen in lens vesicles of 60-hour embryos, i.e., before spe- cific morphological differentiation (see Table 23). The latter suggest, also, that the ap- parent disappearance of a specific embryonic lens antigen during development (Burke et al.) may be attributed to the presence of yolk in the preparations of embryonic anti- 560 gens. The various experiments seem to be in accord in that the younger the embryo the less organ antigen is detectable, as illustrated in Table 23. It is, however, clear that the complete absence of an antigen cannot be proved by serological methods. So the pos- sibility remains open that “adult” antigens ONTOGENY OF IMMUNOLOGICAL PROPERTIES therefore, absorbed with frog serum. When the absorbed antisera were tested with the saline extracts of brain, liver, kidney and heart, their precipitin titers (dilution of antigen) were found to be greatly reduced. It could be concluded, then, that the presence of serum or of tissue substance with “‘serum- Tas_e 22. Reactions of Brain- and Lens-Antisera with Saline Extracts of Brain and of Lens, Respectively, of Chick Embryos of Vari.us Ages (after Burke et al., ’44) AGE IN ADULT BRAIN ANTI- 312-HOUR AGE IN ADULT ADULT 300- 160- HOURS OF SERUM (ABSORBED WITH BRAIN ANTI- HOURS OF LENS LENS HOUR HOUR EMBRYO TESTIS, OVARY, KIDNEY SERUM EMBRYO ANTI- ANTI- LENS LENS SUPPLYING AND LIVER EXTRACTS) (SIMILARLY SUPPLYING SERUM SERUM ANTI- ANTI- BRAIN ABSORBED LENS SERUM SERUM EXTRACT WITH 312- EXTRACT HOUR ORGANS) Comple- Precip- Precip- Comple- Precip- Comple- Precip- ment itin itin ment itin ment itin fixation test test fixation test fixation test 96 0 0 0 V2 0 0 0 0 120 0 0 0 96 0 0) 0 oe 144 0 0 0 110 0) Dini 160 0 0 1+ 120 0 0 2+ 2 168 0 0 134 0 4+ 184 0 0 ioe 144 0 2+ 192 0 0) 148 0 4+ 208 0 0 1+ 160 1+ 0 2+ 4+ 216 0) 0 180 1+ 0 4+ 232 0 0 2+ 200 2+ 0 4+ 240 0 0 224 2+ 0 4+ 256 0 0 2+ 250 2+ 1+ 4+ 260 Se 0 272 3+ 1+ 4+ 280 Pita 0 2+ 300 3+ 1+ 4+ 300 2+ 1+ 330 4+ 2+ 4+ S112) 2+ 1+ 3+ Adult 4+ 4+ 4+ 312 (unabs.) 4+ Adult (abs.) 2+ 2+ Adult (unabs.) 4+ 4+ Taste 23, Precipitin Reaction of Adult Lens Antiserum with Saline Extracts of Embry- onic Chick Lens (after Ten Cate and Van Doorenmaalen, ’50) 168 3200 Agelofembryo, Hours) .-9 ere 192. Last reactive dilution of extract. . . .6400 120 96 74 72 60 54 Sil 48 1600 800, 9400 say eee ct a Eee eee may be present in the earliest stages of development, or even in the unfertilized egg. Cooper (’46) demonstrated that antisera against saline extracts of the brain of adult frogs react with extracts of the eggs, embryos and larvae and of adult liver, kidney, heart and serum. Although the brain extract used for immunization, as well as the other organ extracts used in the tests, had been well washed it was suspected that serum might be present in these and be responsible for the cross-reactions. The anti-frog-brain sera were, like” determinants largely accounted for the cross-reactions. Since some reactivity for the organ extracts remained after the absorption, this indicated that there may also be anti- gens other than those with serum-protein determinants that are common to brain and the other organs tested. When the absorbed antisera were tested with saline extracts of the eggs, embryos and larvae the titers were also found to be greatly reduced, as is illus- trated in Table 24. These results again show that antigens with serum specificity are ONTOGENY OF IMMUNOLOGICAL PROPERTIES present in the extracts. For the larval stages this can be interpreted on the basis of the ere presence of blood at these stages. For the ee i) 2 earlier stages the results mean that, even Ey ae os on before blood serum develops, antigens are 2 fs cial Gee ace a ae present that have determinants that are bats eae Aces similar to at least some of those of constitu- = é ents of adult serum. Similar results have been g 2 obtained by Flickinger and Nace (’52). ae aS In order to obtain further information se elves = & concerning the nature of these antigens m a = 2 St Ey Cooper (’48) prepared antisera against frog s Bs? Bs serum, egg-brei supernatant and egg-yolk a a platelets. The antiserum against adult serum 5 5 2 cross-reacted with saline extracts of the eggs BS ae ee y EEE (unfertilized or fertilized), embryos and =o 3 Bae S 5 alge larvae of various stages, confirming the above- ce § < eran lee: 3 a9 described results. Also the antisera vs. egg oe S rake a # Ss “supernatant” and egg “yolk” cross-reacted SIS a E with adult serum, thereby demonstrating that = = . 8 the egg substances with serum-like reactivity S & = e ae could act as immunizing antigens. In addi- Sy ae 5 s S/S ¢e isis tion the three kinds of antisera (vs. serum, = 5 fs Sican eco E : vs. egg “supernatant” and vs. egg “yolk’”’) all S 5 & F = reacted with washed, and subsequently dis- : A < ae solved, yolk platelets as well as with a clear as 5 ‘a a § solution prepared by high speed centrifuga- I3755 : Oe Ae tS ies! tion of the egg “supernatant.” The serum- Seale a 2 S SS 36 like determinants are, then, evidently present a S Be eo¢]- 5 g : ~ n =] DS in both the formed elements and the ground 2. regu) a & 3 substance of the egg. Tests of the antisera mp a g & with fractions of the frog serum and of clari- > 5 an = e fied egg “supernatant” prepared by repeated eS) S = é =) © o salting out by 34, 50 and 62 per cent satura- 8 ne 2 G = =o Se tion with ammonium sulfate gave cross- is 5 ie fe e re ‘as reactions in all cases. Although these frac- ae 5 2 8 tions are very likely not completely homo- S eal SE geneous, they may well be sufficiently so to Sale rm permit the conclusion that determinants of capa 4 Quel 2 a both albumins and globulins of the serum Sa Sy line 0 e a A a & are represented in various substances of the S 5 Fi ithe haan o g egg. Cooper presents additional evidence, ayia ‘me S from heat inactivation, that these cross-reac- s =| 2 6g 2 tive substances of the egg are proteins and SS) $s K suggests also on the basis of the details of = 6) ae Se 3 the serological reactions that they are not a Decal ike chemically identical with any of the serum e Sala seer ha lite proteins. Tests by the Oudin (’48) technique Z 9 indicated the presence of five to seven such e Ss antigens (Cooper, °50). More recently Clay- ae ton (753) and Spar (’53) report that some - ES ° antigens appear and others disappear during iS Ae £5 gastrulation and neurulation in amphibians. a of © e s 23 Clayton further reports that ectoderm and at se ee 23 archenteric roof contain fractions specific to a 2% os themselves. aa E ae In chickens Schechtman (’47) has likewise S32 ¢ * 5 demonstrated the presence of serum-like anti- y< 7m gens in egg yolk and in extracts of embryos 561 stages but only between absorbed and unabsorbed antiserum. 562 at various developmental stages, using anti- sera against the euglobulin, pseudoglobulin and albumin fractions, as well as against whole serum. The antisera against the serum fractions cross-reacted with the respective fractions and with egg yolk. After absorption with the heterologous fractions the antisera still reacted with the homologous fractions and with egg yolk. After additional absorp- tion with egg yolk the euglobulin antiserum still reacted with the homologous antigen but the pseudoglobulin and albumin antisera failed to do so. Apparently the latter two fractions possess no antigens that are not ONTOGENY OF IMMUNOLOGICAL PROPERTIES of incubation, and more slowly subsequently. These workers suggest that the vitelloid constituent may be a pseudoglobulin pos- sibly identical with livetin, whereas the non-vitelloid constituent may be represented by serum albumin. However, this does not seem to be quite consistent with the pre- viously reported (Schechtman, °47) results showing that when antisera against pseudo- globulin and antisera against albumin are absorbed with yolk they no longer react with the homologous antigen (see above). In another set of experiments, Schechtman (48) prepared antisera against a saline ex- TasiE 25. Reactions (Ring Tests) of Rabbit Anti—19-day-Chick-Embryo-Brain Sera with Extracts of Blood and Organs of 19-day Embryos and of Early Embryos (from Schechtman, ’48) TESTED WITH SALINE EXTRACTS OF: ANTI-BRAIN SERA ABSORBED WITH BLOOD BRAIN LIVER HEART Unabsorbed + + + + Blood — + + + Brain = — _ - Liver — ae _ se Heart _ — - _ Muscle —_ =e _ — also present in the heterologous fractions and in yolk. The tests with embryo extracts were made with anti-whole serum and with anti-euglob- ulin (absorbed with pseudoglobulin and albumin). Both of these antisera reacted with extracts of embryos (free of visible yolk), from the primitive streak stage to the 15- to 17-somite stage. However, after absorption with yolk the antisera no longer reacted with these extracts, but still reacted with adult serum and with the blood and also the per- fused (liver and brain) or washed (heart and muscle) organs of the 19- 20-day embryo. Thus, of the antigens detectable by these anti- sera, the extracts of the early embryos con- tain only those that are common to yolk, and that may be termed “yolk-like” or ‘‘vitel- loid.”” Although blood is present in the 7- to 8-somite and 15- to 17-somite stages that were tested, the “non-vitelloid” antigens of adult serum were not detectable. Further detailed tests by Nace and Schechtman (748), with yolk-absorbed an- tisera vs. adult serum, showed that “non- vitelloid” antigens are first detectable in the blood of 5-day embryos. These increase rap- idly in relative amount up to the ninth day PRIMITIVE EARLY 4705 MUSCLE YOLK STREAK NEURULA SOMITES =F ar AF =F a5 AF =F AF a ae tract of perfused brain of chicks of 19 to 20 days of incubation. These antisera cross- reacted with extracts of liver, heart, muscle and blood of chicks of the same stage, with yolk and with extracts of yolk-free embryos at the primitive streak, early neurula and 4- to 5-somite stages. Upon absorption with blood the antisera no longer reacted with yolk but still reacted with the various organ and early embryo extracts (see Table 25). It appears, then, that antigens of the brain extract that are not present in blood or yolk are present in the early embryos and the other organs tested. These antigens are evi- dently not specific for brain, since absorption of the anti-brain sera with either liver or heart removes reactivity for brain and for the early embryos. Ebert (750) partly con- firms these results, but Burke et al. (see above) obtained a brain-specific antiserum after absorption with heterologous organs and found that the absorbed antiserum re- acted with extracts of late stages but not of early stages. The basis for this discrep- ancy is not clear. In the sea urchin Paracentrotus, Perlmann and Gustafson (’48) find common antigens to be present at stages from the unfertilized ONTOGENY OF IMMUNOLOGICAL PROPERTIES 563 eggs to the 48-hour pluteus. From the results of absorption experiments they conclude that there are, in addition, antigens present in the 48-hour embryos that are not detect- able in the earlier stages tested (unfertilized eggs, 4-hour and 12-hour embryos). On the other hand, none of the antigens detectable in eggs and early embryos appears to be lost in the later embryo. These workers also examined vegetalized 48-hour embryos, pro- duced as a result of lithium treatment, and found no antigenic difference from the nor- mal plutei. By means of the Ouchterlony (49, see also Jennings and Malone, 54) technique, of diffusion in agar plates, Perlmann (753) finds that extracts of the different develop- mental stages of the sea urchin possess mostly common antigens that are also of similar con- centration. Harding et al. (54), using this method with hybrid embryos, report the ap- pearance of “paternal” antigens at a stage be- fore the morphological traits can be easily detected. In the Cecropia silkworm Telfer and Wil- liams (53) describe, in addition to five per- sistent antigens, one that appears in the blood of fifth instar larvae and disappears during adult development. Telfer (54) finds an anti- gen in adult female blood that is undetectable in larvae and almost so in adult males but present in the yolk of unfertilized eggs. The various investigations described above are more in accord in regard to antigenic resemblances than in regard to differences at various stages in development. They do not rule out the possibility that all antigenic structures are represented in the uncleaved ege and only changes in quantity and loca- tion (i.e., distribution to different chemical substances and to different tissues) occur during development. This question will be considered further after presentation of ex- periments on blood cell antigens. ANTIGENS OF VERTEBRATE BLOOD CELLS In humans the A and B isoagglutinogens have been detected in the erythrocytes of the month-old fetus and the M and N ag- glutinogens at the second month (see Wiener, °43, for references). Presumably these may be present in earlier stages that have not been tested. In rabbits the agglutin- ogens H, and H., have been found as early as the 4 mm. stage (Keeler and Castle, ’34), at which time the erythrocytes are still nucleated. According to the early work it would seem that all the blood cell antigens appeared on the cells as soon as they were formed. However, there are recent experi- ments in which some cellular antigens have been found to appear rather late. Briles, McGibbon and Irwin (748) studied chicken red cell antigens the inheritance of which is determined by two series of multiple al- lelic genes. The antigens determined by one series were all found in embryos of about 3 days. In the other series one of the antigens was detected in 4-day embryos but the other three antigens did not appear until after hatching. Ycas (49) studied antigens of sheep red cells, nine of which are detected by use of immune sera and one (R) by means of an isoagglutinin present in the serum of certain sheep (not possessing the R antigen). The former were found to be present at birth, but the latter did not usually appear until two or three weeks later. While the red cells of the newborn R lambs did not possess the antigen, it was found to be pres- ent in the serum of the animal at that stage as well as in the adult. This antigen, then, appears in the serum before it can be de- tected in the red cells of the animals. It would be of further interest to learn whether or not the antigen is detectable in various tissues of the fetus and early embryos. In the chick Witebsky and Szepsenwol (734) found Forssman antigen to be present in, and equally extractable from, first to twelfth day embryos. In humans fetal hemoglobin is re- ported to be antigenically different from that of the adult (Darrow et al., 40). The extensive genetic studies that have been made on the blood groups of man and lower vertebrates have clearly established the fact that these cellular antigens are gene- determined. In fact the antigens seem to be rather direct products of their causative genes, each antigen being produced by the action of a specific gene with no influence, in general, of other genes or of the environ- ment. Since all of the cells of an organism are supposed to contain the full complement of genes, the appearance of these antigens on the surface of particular cells raises the same sorts of questions as are involved in the development of any particular gene- determined character in particular tissues. Certain antigens, such as the A and B char- acters of humans, are found in most other tissues of the body, whereas others, such as the human M and N antigens, are restricted to the blood cells. Thus, certain genes seem to be active in a variety of tissues in im- pressing a certain specificity on the tissue substances, while others act to an appreciable 564 ONTOGENY OF IMMUNOLOGICAL PROPERTIES extent only in a particular tissue. The work on the development of the blood cell antigens also points to a difference in the time at which various genes may come into action in the same tissue. This work does not rule out the possibility that the differences involve primarily rates of activity in different tissues and concomitant differences in distribution of the antigens in various parts of the cell. DEVELOPMENT OF NATURAL ANTIBODIES AND OF COMPLEMENT While in humans the A and B agglutino- gens are demonstrable in the early fetus, the corresponding isoagglutinins are usually not found in the serum until some time after birth (see Wiener, ’43). In fact it has been shown that when agglutinins are present in the serum of the newborn infant these are derived from the mother by transfer through the placenta. The maternal agglu- tinins disappear during the first two weeks of postnatal life while new ones character- istic of the infant begin to appear in the serum. In ruminants transfer of antibodies does not occur across the placenta, which is of the syndesmochorial type with four or five layers of cells separating fetal and maternal bloods instead of the two layers of the hemo- chorial type of placenta of primates. How- ever, as is well known since the experiments of Orcutt and Howe (’22), antibodies (both immune and natural) are acquired from the mother by newborn ruminants as a result of ingestion of the colostrum. In the recent experiments of Ycas (49) the normal anti- R antibody could not be detected in the serum of a lamb at birth but was present after ingestion of the colostrum and _ per- sisted for a variable period of several weeks. The earliest time of appearance of the lamb’s own anti-R was found to be 15 weeks. The transfer of proteins by ingestion of the colo- strum has been studied in calves by Hansen and Phillips (47), who find that gamma- globulin is absorbed only during the first day after birth when the gut wall is appar- ently permeable to large molecules. While it has been generally believed that, where transfer of antibodies from mother to fetus occurs, the route is through the placenta, it has been shown recently by Brambell et al. (49) that this is not neces- sarily the case. These workers demonstrated in rabbits the presence of maternally derived antibodies in the yolk-sac cavity of embryos at a stage prior to establishment of the embryonic circulation. They have also shown, by ligaturing the yolk-sac stalk of 24-day embryos, that the maternal anti- bodies do not pass through the placenta, but rather by way of the uterine lumen and yolk sac into the fetal circulation. To what extent these results with rabbits may apply to other Mammals remains to be determined, but it seems unlikely that the situation would be the same in species with a very rudimentary yolk sac. In chickens a naturally occurring hemo- lysin for sheep erythrocytes is not found un- til about 5 days after hatching (Pickering and Gladstone, ’25). Apparently this anti- body is not transferred from the hen to the egg, although other antibodies produced as a result of immunization of the adult hen can evidently accumulate in the egg. For example, such transfer has been shown for diphtheria and tetanus antitoxins (Jukes et al., ’°34; Fraser et al., °34; Ramon, ’28) and for antibodies against Newcastle disease vi- rus (Brandly et al., ’46). Complement (alexin) is apparently formed rather early in mammals. It has been re- ported (Solling, °37) to be present in low titer in the serum of the 14-week human fetus and to reach full strength at 28 weeks. In chickens, according to one investigation (Sherman, °19), it is present in 17- to 21-day embryos, while according to another (Polk et al., 38) it is not detected until two days after hatching. DEVELOPMENT OF ANTIBODY-FORMING CAPACITY It is generally assumed that the ability to form antibodies is lacking or poorly de- veloped in the embryo, the fetus and the newborn (see, for example, Needham, °42; Beveridge and Burnet, 46; Topley and Wil- son, 46). Experiments concerning this have been performed with mammals and birds. In mammals the fetus is not readily avail- able for such studies. The problem is also complicated by placental transmission of antibodies and by uncertainties as to whether or not antigens that are introduced into the mother will reach the fetus. However, ex- periments have been performed on newborn mammals and at various ages after birth. Thus, Freund (730) injected rabbits with sheep red cells, typhoid bacilli, horse serum and egg white and found that those of one to 20 days of age produced either no anti- bodies or only very small amounts of anti- bodies against these antigens. With increas- ONTOGENY OF IMMUNOLOGICAL PROPERTIES 565 ing age the titers of antibody obtained increased. A detailed study of antibody production in chickens of various ages from hatching to 12 weeks has been made by Wolfe and Dilks (48). These workers used bovine serum as antigen and obtained very weak, or no, precipitins from the newly hatched chicks. The titers increased up to 5 weeks of age and remained approximately the same to 12 weeks (see Table 26). That the embryo chick also fails to pro- duce detectable antibodies has been shown by several investigators using a variety of antigens. For example, no antibody response showed that such grafts begin to regress at the eighteenth day of incubation, and sim- ilar results have been subsequently obtained by others (see Waterman, ’36). It should be noted, however, that this is the time when the chorioallantois itself starts to regress. It is also now well known, since the original experiments of Born (1897) and Harrison (1898), that tissue grafting homo- plastically or heteroplastically succeeds read- ily in embryonic stages, whereas it generally fails in the adult. Other chapters of this book deal with the work along this line, and detailed discussion of the greater part of the literature on this subject is presented in TaBLE 26. Precipitin Titers (Ring Tests) of Antisera vs. Bovine Serum Pro- duced in Chickens Given Three Alternate-Day Injections Starting at Various Ages and Bled 6 and 9 Days after Last Injection (from Wolfe and Dilks, ’48) Age in weeks ay A Z, 3 4 Av. titers as tube* numbers Number of antisera tested Dey, PAS 2) 2A) Bl LO 230 S69 Os2B Wall 5 6 7 8 9 10 11 12 Os Fs is Et! Fe Bn Ws Ue Co a Fa Da i Un ns Wn 9 a a? 31 42 59 36 31 18 18 18 * Tube 1 is a 121% fold dilution and succeeding tubes are serial twofold dilutions. was obtained with such good antigens as diphtherial and tetanal toxoids (Grasset, 29), B. sporogenes and Vibrio septique (Weinberg and Guelin, *36) and with a bacteriophage and influenza virus (Beveridge and Burnet, ’46). There is a report (Gebauer- Fuelnege, °32) of antibodies being found in the egg white at 14 days of incubation after injecting the embryos with sheep serum four or five times on alternate days starting at the third day of incubation. This result was obtained in only 5 out of 126 eggs used. The ability of many viruses and rickettsiae to grow readily on the membranes and tissues of the chick embryo, as first demon- strated by Rous and Murphy (’11) and by Goodpasture and his collaborators (38-44), seems to correlate with lack of antibody- forming capacity (see Beveridge and Burnet, 460). For example, Brandly et al. (46) find that neutralizing antibodies against New- castle disease virus are not detected in chick embryos earlier than 15 days of incubation. Similarly the ability of various normal and tumor tissues of the same or of other species to grow on the chorioallantois, as shown initially by Murphy (’13), Willier (24) and Hoadley (’24) appears to depend upon in- ability of the embryo to form antibody (see Needham, 42; Loeb, *45). Murphy (14) the recent book by Loeb (745). Here we will consider briefly the question of whether or not antibody formation may be responsible for tissue incompatibility. Loeb and Wright (27) established the genetic basis of tissue incompatibility in their classic demonstration, with inbred lines of guinea pigs, that the reaction was due to the absence in the host of factors present in the grafted tissue. Kozelka (33) showed, in chickens, that the factors responsible for the incompatibility of homografts are not rep- resented by the agglutinogens of the blood cells. In rabbits, too, the red cell antigens do not appear to be involved (Medawar, ’46b). However, in mice incompatible skin homografts stimulate the formation of anti- bodies that react with red cells and leuko- cytes (Amos et al., 54). That an antibody mechanism of some sort is involved in graft- incompatibility is evident from many ex- periments showing immunity to second trans- plants. For example, in experiments of Me- dawar (’46a), inhibitory effects are obtained in rabbit skin that is transplanted to a rabbit that has previously been “immunized” by grafting skin from the same donor or, to a less extent, from another rabbit. Injection of leukocytes also is found to induce this type of immunity to subsequent skin grafting. Skin 566 ONTOGENY OF IMMUNOLOGICAL PROPERTIES can grow well in the presence of serum and mesenchymal tissues of an immunized rabbit (Medawar, *48). However, prolonged treat- ment (Billingham and Sparrow, 54) can pre- vent dissociated Malpighian cells from form- ing epithelium upon grafting, thus indicating a specific, in vitro, antibody effect. In tissue culture it has been shown (Harris, ’43) that mouse and rat tissues are quite compatible. Even when mouse tissues are taken from ani- mals previously immunized against tissues of the rat, guinea pig or chicken there is no in- compatibility exhibited in mixed cultures with each of the latter (Grobstein and Young- ner, 49). It appears, then, that tissues grown in vitro behave much like embryonic tissues in regard to the absence of incompatibility re- action, and this, it seems, should provide a clue to the analysis of the incompatibility problem on the basis of features common to both. One feature common to both is that they are actively growing. Taking this into consideration it would seem reasonable to suppose that any antibodies that might be produced by the tissues of the host embryo or by one of the tissues in culture, in re- sponse to the presence of the foreign anti- gens, would be incorporated into the growing cells. In this location, intimately bound in or on the cells, the antibodies would not be available for action on the foreign tissue. On this basis it is not a necessary conclusion, from the experiments reported in the first part of this section, that the embryo is in- capable of antibody formation. Those ex- periments do not exclude the possibility that any antibodies that are formed may be simply used as part of the building blocks of the tissue. Later, when growth slows down and material is released from the tissue, such antibodies may be liberated and then initiate the incompatibility reaction. However, there are cases in which cells may survive well into the adult life of a genetically different recipient embryo. This is illustrated by the persistence in the adult chicken of melanoblasts transplanted between embryos (Willer and Rawles, *40, 44) and by the occurrence of erythrocyte mosaicism in adult dizygotic cattle (Owen, ’45) and hu- man twins (Dunsford et al., 53), due most likely to reciprocal transfer of the fetal blood through the common placenta. In addition, experiments by Billingham, Brent and Meda- war (753) now show that adult animals may be induced to tolerate foreign skin grafts if they are inoculated during fetal life with cells from the donor strain. Thus mice of CBA strain injected in utero, at the 15th-16th day, with tissue (isolated cells and clumps from chopped-up adult testis, kidney and spleen) of A strain can tolerate A-strain skin transplanted in adult life. In chickens trans- fusion of blood between embryos of different strains renders the recipient, at two weeks after hatching, similarly tolerant to skin erafts received from the donor strain. A some- what related investigation (Buxton, ’54) has shown that the inoculation of chick embryos with killed Salmonella pullorum results in a marked decrease in the capacity of the hatched chickens to produce antibodies in their sera when challenged again with the bacterial antigen. Billingham et al. (53) sug- gest that the failure of the host’s immuno- logical response in their experiments may be related to a specific immunological paralysis provoked by high doses of antigen (see Fel- ton, 49). Another possibility is that “type- transformation” (see below) has been in- duced in those host cells responsible for the immune response. These highly interesting experiments will undoubtedly stimulate ex- tensive investigation leading to their inter- pretation, so that one need not speculate fur- ther at this time. THE CONCEPT OF NATURAL AUTO-ANTIBODIES The results of certain experiments on the interacting substances of eggs and sperm (see Section IV, Chapter 1), along with con- sideration of various findings of others, re- ported in immunological literature, led Tyler (40—48) to propose a view of cell structure and growth termed an ‘“‘auto-anti- body” concept. This concept states that each of the various macromolecular substances of which cells are constructed bears the same sort of relationship to another of these sub- stances as do antigen and antibody, and that their mode of origin is analogous to that of antibody formation. This view has been applied to the interpretation of certain as- pects of the problems of differentiation (Ty- ler, 47). A brief account of this is presented here along with some of the background of the concept. The experiments with the fertilizins and the antifertilizins that are derived from the surface layer of eggs and of sperm, respec- tively, have shown that these substances interact in a manner analogous to that of antigen with antibody (see Section IV, Chap- ter 1). In the course of this work it was found (Tyler, °40) that upon removal of ONTOGENY OF IMMUNOLOGICAL PROPERTIES 567 the gelatinous coat (which is constituted of fertilizin) of the sea-urchin egg an anti- fertilizin could be extracted from the naked eggs. This antifertilizin, derived from within the egg, was capable of agglutinating intact eggs or of forming a precipitate with the surface coat substance (fertilizin). Thus, from one and the same cell a pair of sub- stances are obtainable that interact in sero- logical manner, and which may be termed complementary substances. Tests with verte- brate blood cells and with bacteria have also yielded such auto-agglutinins, but meth- ods for obtaining them consistently have not as yet been fully worked out. The diffi- culties here appear to involve interaction and precipitation of the complementary sub- stances in extraction procedures that cause destruction of the cell before surface sub- stance can be sufficiently removed. Also, some may be “univalent” (Tyler, ’45, °54) and, thus, not readily detectable by direct testing. There have, however, been a suffi- cient number of reports in the literature (see Tyler, °47, for references) in which auto-antibodies have been evidently ob- tained, so that it seems safe to conclude that the situation is a general one for all kinds of cells. Tests have also been made of the possibility that such auto-antibodies might have protective action against toxins, ven- oms, etc., and evidence for such action has been obtained (Tyler, 46) in the case of a venom. Thus, it has been found that the lethal action on mice of the venom of the Gila monster can be neutralized by serum or by an extract of liver of the same animal. According to the auto-antibody concept the formation of immune antibodies by an animal, in response to the injection of a foreign antigen, is a special case of the gen- eral type of process involved in the synthesis of the macromolecular constituents of cells. The now fairly generally accepted view of the manner of formation of immune antibody is that which was proposed by Breinl and Haurowitz (’30), Alexander (32), and Mudd (32) and which has been extended by Paul- ing (40). This view proposes that foreign antigen becomes incorporated in the site of synthesis of serum globulin so that, as the polypeptide chains of the new globulin that is being formed fold up, the molecules now bear regional surface configurations that are complementary to certain structures on the antigen. The auto-antibody concept in fact may be inferred from this view, if one con- siders the situation in the absence of foreign, introduced, antigen. Under such conditions the normal globulin that is formed should bear regional structural configurations that are complementary to chemical structures of the normal site of synthesis. The concept is not, however, restricted to the special case of the formation of serum globulin but ap- plies to any of the macromolecular con- stituents of cells. Since it is the formation of such substances that is involved in the proc- ess of growth, that process, then, may be considered to result from the operation of the same sort of mechanism exemplified in the formation of immune antibodies. Growth also involves an increase in self-duplicating entities, such as genes, and for the formation of these one may assume structures that are both complementary and identical, as Paul- ing and Delbriick (40) suggest, or the pro- duction of an intermediate template, as Em- erson (45) proposes as an alternative. RELATION TO SPECIFIC ADHESION OF CELLS This general point of view has been ap- plied (Tyler, ’47) to two aspects of the prob- lem of differentiation. One is the question of the nature of the forces that are re- sponsible for the specific adhesion or non- adhesion of cells and tissues. The other is the phenomenon of induction. In regard to the former Loeb (22) had early suggested that specific agglutination is the factor in- volved in binding cells into tissues. That some mechanism analogous to antigen-anti- body interaction is involved is suggested by the specificity of the tissue affinities. Thus species specificity is exhibited, for example, in experiments on the reconstitution of a sponge from cells that have been dissociated from one another by forcing the organism through fine bolting cloth (Wilson, ’07, °32: Galtsoff, 29). When the dissociated cells of two species are mixed, coalescence is found to occur only between those of the same species. The degree of cell-type spec- ificity that is exhibited in these experiments is not entirely clear, but the evidence is to the effect that the archeocytes are mainly involved and that other cells, such as collar cells, that are on hand may be incorporated. Experiments by Holtfreter (43—48) with amphibian embryos offer illustration of the specificity of association of cells within the species. He has shown that isolated cells or clumps of cells from blastulae or early eastrulae will fuse regardless of their pro- spective significance. While the cells of the same germ layer remain fused, separation 568 ONTOGENY OF IMMUNOLOGICAL PROPERTIES later occurs between ectodermal and en- todermal cells. At this time, too, initial in vitro combinations of entodermal and ecto- dermal cells fail to unite. However, meso- dermal cells are found to be capable of uniting with either of the others and are thus capable of tying together cells of the other two germ layers. The various inter- actions occur only as a result of contact and there is no indication of the operation of forces acting at a distance. These and other experiments of Holtfreter and of others serve to emphasize the antigen-antibody analogy. On this basis the specific adhesion or non- adhesion of cells and tissues would depend upon the extent of complementariness of chemical structure of the respective cell surfaces, and the changes that these undergo during development. A similar point of view has been developed in more detail by Weiss C41, °47). Through the auto-antibody concept this can be related to genic action. It is known, especially from the work on the antigens of vertebrate blood cells, that the antigenic composition of the cell surface is gene- determined (see reviews by Irwin, ’47, °49). The antigens appear to be rather direct prod- ucts of their causative genes, each of which manifests itself independently of its allele. Cells belonging to different tissues do not, as a rule, possess the same set of surface antigens despite the presumed identity of chromosomal constitution of all the somatic cells of the body. Certain antigens, such as the A and B structures of humans, are found on most other cells and tissue fluids, whereas other antigens, such as M, N, and Rh, are restricted to the red blood cells. According to the auto-antibody concept each gene would serve as a mold or template upon which is formed a complementary structure. This may represent a terminal step or be merely an intermediate step in a series of such syntheses. The antigenic differences between the surface of cells of different tissues can be interpreted in various ways. The simplest, perhaps, would be the assumption that all of the genes are active, but at different rates in different tissues, so that the surface in each case would be composed of a different assortment of the products of activity of the genes. Each antigenic structure on the sur- face of the cell may, then, be regarded as representing a specific structural aspect of the determining gene or the complement of that structure. The extent to which the cells of one tissue have surface substances that are complementary to those of another tissue would determine the degree of association of the two tissues. Within one tissue the ad- hesion of the cells may be regarded as in- volving combination of surface substance of one cell with complementary subsurface substance of another, in a manner previously suggested (Tyler, 40) for auto-agglutination phenomena. RELATION TO INDUCTION For the other aspect of the problem of differentiation that we wish to consider, namely the phenomenon of induction, there is, as yet, only slight indication that the antigen-antibody type of interaction may be involved. One indication is that inductive activity is associated more with protein materials than with other types of extracts of inductor tissue (see Holtfreter, ’33, ’48b; Barth and Graff, ’43). Inductive action re- quires contact with the inducing tissue and it is evidently not dependent upon special metabolic properties of the latter, since dead tissues possess inductive capacity. This ca- pacity may depend, then, upon the presence of a specific type of protein, or other macro- molecular substance, on the surface of the natural inductor. The fact that certain tissues which lack inductive action in the living state acquire this capacity when killed may be attributed to the exposure of inductor substance, previously located in a subsurface position. For speculation along this line to acquire any significance it would be important to know whether specific morphological changes can be brought about in cells by the action of antibodies. It is, of course, well known that cells of various types can be lysed by treatment with immune antibodies in the presence of complement. This, however, is a lethal effect that would certainly not warrant any consideration unless non-lethal steps in the process were demonstrable. There are two somewhat similar lines of work that bear on this point. One is the work with the so-called reticulo-endothelial immune serum of Bogomolets (see reviews by Pomerat, ’45, ’46; Straus, ’46). This serum is cytotoxic or inhibiting in high concentra- tion. In low doses it has been claimed to stimulate cellular growth and activity. How- ever, careful in vitro tests of this by Pomerat have not revealed any very marked stimula- tory action. The other work consists in ex- periments by Weiss (’47) in which hen’s eggs, at 60 hours to 8 days of incubation, were injected with antisera prepared against ONTOGENY OF IMMUNOLOGICAL PROPERTIES 569 autolyzed suspensions of liver, kidney or pectoral muscle of adult chickens. Upon examination at 20 days of incubation the total weights of the injected embryos were found to average considerably less than those of the controls. However, the organ corresponding to the antiserum employed showed rather marked relative increase in size. Thus the liver weights averaged 10 per cent above the normal controls and 29 per cent more than the livers of embryos that had been injected with anti-kidney serum. Kidney size in the embryos treated with anti-kidney serum was found to average 28 per cent higher than in the embryos treated with anti-liver serum. Experiments by Ebert (51, °54) on the enlargement of host-embryo spleen by chorioallantoic grafting of adult spleen can be related to this, as he suggests, on the assumption of a release from the eraft of certain of the complementary sub- stances which then become incorporated in the host spleen and serve as templates for the synthesis of further splenic substance. There is, then, some evidence of a specific stimulating effect of antibodies. This en- courages further examination of the pos- sibility that specific morphological alterations may be induced in cells by means of anti- bodies. In the above cited experiments of Weiss no alteration of cell type seems to be involved, and in no experiments of others does this possibility seem to have been directly tested. Perhaps the nearest approach to this is the work on changes in morpho- logical and antigenic structure in micro- organisms. It is well known (see Dubos, ’46, for references) that certain transformations, such as the smooth to rough change, can be brought about in many kinds of bacteria by growth in immune serum. Cultures of encapsulated organisms, such as the pneumo- cocci or Friedlander’s bacilli, growing in media containing the homologous, specific anticapsular antibody, transform to a non- specific, non-encapsulated type. The reverse transformation can also be obtained by growth in an antiserum directed against the non-encapsulated organisms. Motile variants of B. subtilis can be transformed to non- motile, and vice versa, by growth in the corresponding antisera. Reversible changes described as phase variation, as for example in the specificity of the flagellar antigens of various Salmonella species, can be induced by means of homologous antisera. Changes in antigenic specificity have been described also in Paramecium (Sonneborn, 48). The manner in which the antibodies act has not, as yet, been fully elucidated in any of these experiments. The various changes also occur, although generally more slowly, in response to other environmental changes or sometimes spontaneously in the absence of controlled or readily detectable environmental change. In the bacteria it is often uncertain to what extent the antibodies act by selectively in- hibiting the types with which they react and thus permit a more rapid growth of the variants in the culture. Selective action of this type does not appear to be involved in the experiments on directed transformation of pneumococcal types (see McCarty, Taylor and Avery, °46). This tranformation consists in the conversion of a non-encapsulated (R) variant derived from one specific serological type into a serolog- ically different encapsulated (S) type. The conditions for transformation involve sev- eral factors, including anti-R serum and desoxyribonucleic acid derived from organ- isms of the type into which the change is to be directed. The latter is evidently the direc- tive agent in this transformation. The R- antibodies can be replaced by other agents, such as normal serum and agar semisolid medium, that cause a colonial type of growth of the R organisms. The action of the R- antibodies is interpreted as inducing a type of growth in which local conditions sur- rounding the organisms so modify the cells as to permit absorption or entrance and action of the specific desoxyribonucleic acid. There is, then, evidence that antibodies can induce changes in cells, but as far as present evidence is concerned, they do not appear to be directive agents. If the antigen- antibody type of reaction is involved in the phenomena of embryonic induction it could conceivably operate in some manner such as is indicated in the experiments on the trans- formation of pneumococcal types. It is of interest to note, in this connection, that nu- cleic acids, as Brachet (’47) has emphasized, can act as inductors. This might mean that certain specific nucleic acids are liberated from the surface of living cells having spe- cific inductive action in the embryo. This possibility has not, as yet, been examined, but seems worth investigating. On the other hand, the present status of experimentation along this line does not exclude the possibil- ity that natural auto-antibodies may act rather directly as inductive agents, and var- ious schemes could be devised whereby this would entail inactivation of different sets of genes in cells of different tissues. This be- comes similar to the proposals of Sturtevant 570 ONTOGENY OF IMMUNOLOGICAL PROPERTIES (44) and Emerson (’44) for the induction of mutations by means of antisera. The pres- ent analysis suggests, then, the desirability of investigation of the occurrence of com- plementary substances in inductor and re- actor tissue and tests of inductive action of such substances as well as of immunologi- cally produced antibodies. REFERENCES Alexander, J. 1932 Some intracellular aspects on life and disease. Protoplasma, 74:296-306. Amos, D. B., Gorer, P. A., Mikulska, Barbara M., Billimgham, R. E., and Sparrow, Elizabeth M. 1954 An antibody response to skin homografts in mice. Brit. J. Exper. Pathol., 35:203-208. Barth, L. G., and Graff, S. 1943 Effect of protem extracts of neural plate plus chordamesoderm on presumptive epidermis. Proc. Soc. Exp. Biol. & Med., 54:118-121. Beveridge, W. I. B., and Burnet, F. M. 1946 The Cultivation of Viruses and Rickettsiae in the Chick Embryo. Medical Research Council, Spe- cial Report Series No. 256, pp. 1-92. His Majesty’s Stationery Office, London. Billingham, R. E., and Sparrow, Elizabeth M. 1954 Studies on the nature of immunity to homologous skin grafts, with special reference to the use of pure epidermal grafts. J. Exp. Biol., 31:16-39. , Brent, I, and Medawar, P. B. 1953 ‘Actively acquired tolerance’ of foreign cells. Nature, 772:603-606. Bird, G. W. G. 1953 Observations on haemag- glutinin “linkage” in relation to iso-agglutinins and auto-agglutinins. Brit. J. Exper. Pathol., 34: 131-137. Bogomolets, A. A. 1943 Antireticular cytotoxic serum as a means of pathogenic therapy. Am. Rey. Soviet Med., 7:101-112. Born, G. 1897 Uber Verwachsungsversuche mit Amphibienlarven. Roux’ Arch. Entw.-mech., 4: 349-465. Boyd, W. C. 1947 Fundamentals of Immunol- ogy. Interscience Publishers, New York. Brachet, J. 1947 Embryologie Chimique. Edit. Desoer, Liege. Brambell, F. W. R., Hemmings, W. A., Henderson, M., Parry, H. J., and Rowlands, W. T. 1949 The route of antibodies passing from the maternal to the foetal circulation in rabbits. Proc. Roy. Soc., London, Ser. B., 736:131-144. Brandly, C. A., Moses, H. E., and Jungherr, E. L. 1946 ‘Transmission of antiviral activity via the egg and the role of congenital passive immunity to Newcastle disease in chickens. Am. J. Veter. Res., 7:333-342. Braus,H. 1906 Uber das biochemische Verhalten von Amphibienlarven. Roux’ Arch. Entw.-mech., 22:564-580. Breinl, F., and Haurowitz, F. 1930 Chemische Untersuchungen des Prazipitates aus Hamoglo- bin und Anti-Hamoglobin-Serum und Bemer- kungen iiber die Natur der Antikérper. Z. phys- iol. Chem., 192:45-57. Briles, W. E., McGibbon, W. H., and Irwin, M. R. 1948 Studies of the time of development of cell- ular antigens in the chicken. Genetics, 33:97. Burke, V., Sullivan, N. P., Petersen, H., and Weed, R. 1944 Ontogenetic change in antigenic spe- cificity of the organs of the chick. J. Infect. Dis- eases, 74:225-233. Buxton, A. 1954 Antibody production in avian embryos and young chicks. J. Gen. Microbiol., 10:398-410. Clayton, R. M. 1953 Distribution of antigens in the developing newt. J. Embryol. Exp. Morphol., 1:25-42. Cohn, M., Wetter, L. R., and Deutsch, H. F. 1949 Immunological studies on egg white proteins. I. Precipitation of chicken-ovalbumin and conal- bumin by rabbit- and horse-antisera. J. ILmmunol., 61:283-296. Cooper, Ruth S. 1946 Adult antigens (or speci- fic combining groups) in the egg, embryo and larva of the frog. J. Exp. Zool., 107:143-172. 1948 ae pearance of acini and colloid. Moreover, the xt z = first appearance of organic iodine is accom- es a © = - = a a a) = Zz 5 > ~ Z = re n wo so} o) io] a8 o 3 oY oO To ce hye (ef os Pen |e o § Re | A 2) ue ess || eas vas oS s| 6 258 6 = es) wa iS G9 Ou Soot ae oS 2 g a 0 ies Os Sed es Bag er} Oe oO os Ge . ston q S25 ous gs eo) Get do! pretica|ieu = Oa: tf Sy] 2» TR B33 oHe| 3% a9 elo) le tone . 8g or ae fo) as Foe o ie} & om S w oO ov = o | ma Ae) = wn & be 2 a0 ,, S re) 5} q ©) &= (o) ah to 520 ze mS O00 oA S| 2 UO az B.S A) | an) ins ¢ 8 one, © Tw wr : q— © = ~~ > 249 o> 8 el SG} i 0 8U 5 oc 4 Ey v 1 ee ol | aa cele es me 8 oo & =) Ne + S mm Ox ) Sta Fig. 208. Section of thyroid of a normal 11-day 26 chick embryo showing marked sinusoidal vascular- ar fet 41 5 : y 5 8 ° : eee | ity at time of rapid formation of discrete follicles i=) Oo 6 P - ne z ons with colloid. Carotid artery to the left. Initial x fe) = ie) Yi f= iota) = 2S 135. oe Simro n A — © : ; : 4 20 ~ = ° ww anied by marked increase in vascularity of oY a = us & : 3 See £2 € 3 the gland (cf. Fig. 208). In the bovine fetus $a9 woo from 53 days to term there is a progressive a js : js aA < increase in thyroxine-like compounds of iodine and non-thyroxine iodine. These io- a4 foc dine fractions increase proportionately with 3) = a 5 a ; zz & = increasing fetal size and age (Wolff et al., .@) ’°49). The rate at which each of these sub- stances accumulates in the fetal thyroid shows an exponential relationship to body weight and length as well as to age of the fetus. Moreover, the increase in their ac- cumulation cannot be accounted for solely on the basis of increase in thyroid growth, since the percentage growth rate of the thyroid declines steadily with increasing age. Apparently as the growth rate declines with age there is a progressive increase in the Parathyroid 582 capacity of the fetal thyroid gland to store iodine and its compounds. The question next arises as to the time when the thyroid primordium in the course of its development is responsive to the thyro- trophic hormone. Shortly before and at the time of follicle formation the thyroid is known to be responsive to anterior lobe stimulation. That it behaves as an effector organ long before colloid storage begins is brought out in frog embryos by grafting the primordium of either the thyroid or the anterior pituitary so that the two are in closer proximity than in their normal sites (Etkin, ’39). Irrespective of whether one or the other gland is shifted or an extra gland grafted, such proximity leads to a precocious activation of the thyroid with consequent precocious metamorphosis. There is no pre- cocity if either gland is grafted to a site outside the neighborhood of the other, i.e., apparently when the thyroid is beyond the range of influence of the anterior lobe (see p. 583). Furthermore, an extra thyroid grafted near the pituitary is activated, whereas the host thyroid at its normal site is not. The acti- vation of the thyroid is first visibly evident at approximately seven days after the operation, at which time the follicles are greatly en- larged and contain abundant colloid, con- trasting with the control thyroid where these processes are much less advanced. The acti- vated gland shows a marked increase in size (volume), reaching at 19 days a size over twenty times that of the normal thyroid. It is evident, therefore, that the thyroid primordium is in a responsive state for a considerable length of time prior to its normal histological differentiation and is apparently as responsive as the gland in which follicle formation is about to begin or has begun. At what time in the course of development of the anterior lobe does the secretion of thyrotrophic substance begin? Is its time of onset prior to or coincident with the histo- logical differentiation of cell types? The early responsiveness of the thyroid analyzed above indicates that the anterior pituitary is con- currently active in producing a thyrotrophic substance. Since the activation of a thyroid in proximity to an anterior lobe is similar in nature to activation by the thyrotrophic hormone known to be produced at later stages, it is inferred that the production of the hormone begins very early in pituitary development of anurans. It apparently be- gins not later than seven days after the tail- bud stage, which is long before it is func- tionally effective in normal anuran meta- ONTOGENY OF ENDOCRINE CORRELATION morphosis (Etkin, ’39). Surely at the tail- bud stage the cells of the hypophyseal pri- mordium, which has just invaginated to a position below the infundibulum, are not visibly differentiated into cell types (so reported for R. pipiens by Kleinholz, ’40). The cell types apparently differentiate con- siderably later. According to Kerr (39), whose study of histogenesis of the anterior lobe in relation to the thyroid in the anurans (R. temporaria and Bufo bufo) is the most detailed and complete, the acidophiles ap- pear prior to metamorphosis, i.e., at the stage when the larvae are first completely free- swimming (11 to 13 mm. in length), whereas the basophiles first appear during the most active period of metamorphosis, i.e., of rapid growth of hind legs, appearance of fore-legs, and tail absorption. Furthermore, the histo- logical differentiation of acidophiles and the first occurrence of minute droplets of colloid in the thyroid coincide in time. This, coupled with a steady increase in the number of acidophiles and in the amount of colloid during the early stages up to the beginning of rapid hind-leg growth and with the ab- sence of basophiles, shows a close corres- pondence between acidophiles and thyroid activity. Thus Kerr was led to infer that thyrotrophic function and general body growth as well are associated with the acido- philes. In the amniotes but few attempts have been made to correlate the initiation of thyro- trophic activity with the histogenesis of the anterior lobe. Of interest here is the pioneer work of Rumph and Smith (’26). These in- vestigators reported that intraperitoneal in- jections of extracts of the anterior pituitary from fetal pigs of 140 mm. and 160 mm. stages fail to call forth a response in hypo- physectomized frog tadpoles. However, at 260 to 280 mm. stages (near time of birth) a definite although slight response is elicited as indicated by a clear-cut stimulation of hind-limb growth and some activation of the thyroid. This capacity to evoke meta- morphic changes appeared to be correlated with the time (260 mm. stage) in histogenesis of the cell types that the acidophiles became nearly as numerous as in the adult pig gland. (See Table 27.) However, that the acidophiles have a spe- cific functional significance must be ques- tioned in the light of later detailed studies on the cytological differentiation of the ante- rior pituitary in the fetal pig. Nelson (33) reports that, although acidophiles are present at the 70 to 100 mm. stages, the population ONTOGENY OF ENDOCRINE CORRELATION of basophiles greatly predominates. During successively later stages the acidophile popu- lation gradually increases until the 160 to 170 mm. stage, at which time a marked rise in their number is exhibited. Although at this time the acidophile population is mark- edly increased, the functional test as noted above does not indicate thyrotrophic activity. It is possible that although the thyrotrophic substance is produced by the anterior lobe it is not present in sufficient quantity in the extracts thereof to elicit a positive response in hypophysectomized tadpoles. By this method of bio-assay the hormone appears to be present in sufficient quantities only near the time of birth. Nevertheless, its presence as early as 90 mm. is indicated, since at that stage but not earlier (1) minute droplets of colloid appear within follicles of the thyroid and (2) thyroid extracts stimulate metamor- phosis in the hypophysectomized tadpole (Rumph and Smith, ’26). It must be recog- nized, however, that (a) the thyroid may attain an initial functional state independ- ently of thyrotrophic secretions, (b) the bio- assay method is not sufficiently sensitive for detecting minute quantities of thyrotrophic hormone, and (c) the hormone may be sup- plied to the fetus by the mother. The evidence is inadequate for deciding whether one type of cell or the other is related to the produc- tion of thyrotrophin in the pig fetus. In the chick embryo a fairly close correla- tion exists between the time of differentiation of the thyroid and of the cell types of the anterior pituitary. The thyroid begins to store colloid in follicles on the tenth day (Hopkins, °35), at which time the acido- philes are first apparent in the anterior lobe. The acidophiles gradually increase in num- ber with advance in age, becoming the dominant type of cell by the eighteenth day (Rahn, ’39) or by the third day after hatch- ing (Payne, °46). Although relatively few basophile-like cells appear to be present as early as the twelfth day, they do not attain a fully differentiated state until about ten days or more after hatching. In the chick the presence of acidophiles and basophile-like cells seems to precede by one day the onset of thyrotrophic secretion, as is revealed by erafting thyroid glands from hypophysecto- mized donors to the chorio-allantois of nor- mal and hypophysectomized chick hosts. From the differences in thyroid response given it may be inferred that the thyro- trophic hormone is produced in small yet significant amounts as early as the eleventh day and in still greater amounts from the 583 twelfth to the eighteenth day of incubation (Martindale, ’41). To summarize, it becomes apparent on the basis of functional tests that the anterior pituitary during embryonic development produces a thyrotrophic hormone. Regardless of whether basophiles or acidophiles arise first, a predominance of acidophiles usually occurs according to the majority of investi- gators during some phase of pituitary devel- opment of the growing embryo. Although it is difficult to relate the specific cells to specific secretions, the body of evidence sug- gests that the acidophiles may be associated with the production of the thyrotrophic hormone and, as will be seen below, of the growth hormone as well. Since it is evident that a thyrotrophic sub- stance is produced sooner or later by the developing anterior pituitary, it is important to know whether the hormone content in- creases quantitatively with developmental age. This appears to be the case, as several lines of evidence indicate. From a study of the effects of approximating the primordia of the anterior pituitary and thyroid in anuran embryos, Etkin (739) concluded that a “field of thyrotrophic substance” surrounds the normal pituitary in quite early stages of development. The activation effect on the thyroid is limited in its range. As noted above, an extra thyroid grafted near the pituitary of a host is activated, whereas the thyroid of the host is not. This suggests that the manner of dispersal of the thyrotrophic substance is by diffusion through the tissues rather than through the blood stream.* Furthermore, it is probable that the anterior pituitary at the stages tested is producing * The concept of a direct and local action of hor- mones in actuating the expression of the intrinsic potentialities of endocrine receptor glands and tis- sues is well substantiated by three recent discov- eries. (1) Sex hormones of an embryonic gonad in contact with another one of opposite sex constitu- tion or with a terminal sex receptor elicit specific responses of the tissue components in accordance with their pre-existing reaction capacities (for a more extensive treatment of the subject see p. 594). (2) Hormones from a thyroid graft act locally by direct contact in accelerating metamorphic events in the neuronal elements of the hindbrain of the anuran tadpole (Weiss and Rossetti, 51). (3) Hor- mones from fragments of the anterior pituitary (34%-month cockerel) cultivated for 9 days in vitro in close contact with a thyroid gland of the pre- follicular stage (814-day chick embryo) greatly stimulate the formation of intracellular colloid droplets in the parenchymal tissue of the thyroid; follicles, however, fail to form (Gaillard, *53). 584 the hormone in amounts insufficient to have an activation effect via the vascular circula- tion. Grafts of the anterior lobe from anuran tadpoles of different ages vary in their effectiveness in bringing about precocious metamorphosis in host tadpoles of immature stages. Anterior pituitaries from animals in metamorphic climax are highly potent as grafts in stimulating metamorphosis, whereas those from donors before the onset of meta- morphosis are impotent (Allen, 738). These results may be interpreted as indicating that the normal anterior pituitary undergoes a sudden increase in the production of thyro- trophic substance at the time of metamorpho- sis (Etkin, °38). This marked change in activity appears to be the primary stimulus at metamorphosis, since it is known that the receptor gland is responsive to the pituitary long before the normal time of metamorpho- sis. This peak is soon followed by a decline in thyrotrophic activity, since grafts of glands from tadpoles in stages beyond the more active phases of metamorphosis give a re- tarded response in premetamorphic tadpole hosts. During the subsequent growth of the young or juvenile frog no information is available as to the time that thyrotrophic activity of the anterior lobe again attains a high level. It is highly potent in adult frogs as measured by its capacity to induce pre- cocious metamorphosis. Since the quantity of thyrotrophic sub- stances is apparently low at first, it should be expected that an excess number of anterior pituitary grafts would produce a precocious metamorphosis. In Rana sylvatica this ap- pears to be the case (Etkin, ’35). As grafts, a single anterior pituitary primordium is in- sufficient; but three extra primordia (total of four acting) are sufficient to induce meta- morphosis in 7.8 + 1.4 days as compared with 17.7 + 0.6 days required for the un- operated controls. Furthermore, the animals with four primordia are significantly smaller than the controls at the time of metamorpho- sis. Thus the difference in time required for and maximum size attained at metamorpho- sis give a rough measure of the quantity of hormone produced. The results support the interpretation that the anterior pituitary primordium produces the hormone in smal] quantities in the anuran embryo. Little or no information is available as to whether the thyrotrophic substance is pro- duced in increasing amounts during the ontogeny of amniotes. As previously noted, the presence of thyrotrophic activity of the ONTOGENY OF ENDOCRINE CORRELATION anterior lobe of the fetal pig is indicated near the time of birth but not earlier. The experiments were designed to ascertain the time of onset of secretory activity and not the quantity of hormone produced. More re- fined quantitative methods of extracting and of bio-assay of the thyrotrophic hormone from the anterior lobe from fetal and post- fetal stages are needed in order to ascertain the quantity necessary for thyroid activation and consequent effects on the metabolism of growth. Finally, the question is raised as to whether the thyroid has a reciprocal action on the anterior pituitary during ontogeny. Does the thyroid hormone released into the blood by the activated thyroid regulate or depress the quantity of thyrotrophic substance pro- duced or released by the anterior pituitary, the trophic gland? Is the secretory output of trophic and effector glands regulated by a balance between the quantities of thyro- trophic and thyroid hormone in the blood circulation? That the thyroid has an effect on the development of the anterior pituitary is brought out by the study of thyroidecto- mized anuran embryos (Hoskins and Hos- kins, °19; Smith, ’20). By removing the thyroid primordium from embryos at an early stage (R. sylvatica embryos 5 to 8 mm. in length) thyroidless larvae are obtained. At the expected time of metamorphosis of such tadpoles, the anterior lobe is definitely larger and the number of acidophiles greater than in control tadpoles of the same size and age. The validity and significance of these findings must await further investiga- tion. The increase in number of acidophiles is puzzling, in view of the fact that thyroid removal in certain species of mammals dur- ing postnatal life commonly leads to a de- crease in number or complete loss of acido- philes and an increase of basophiles in the anterior lobe (Severinghaus, ’37). A large body of evidence indicates in general that thyrotrophic potency of the anterior pituitary varies in different species and within the same species at different ages (or phases of reproductive activity). Such potency changes, particularly those following thyroidectomy, lend support to the commonly accepted theory that a high level -of thyroid hormone suppresses the production and release of the thyrotrophic substance and, conversely, a low level of the thyroid hor- mone stimulates the production and release of thyrotrophic substances by the pituitary (for a review of the evidence see Adams, ’46). In keeping with this theory is the significant ONTOGENY OF ENDOCRINE CORRELATION discovery of Courrier (’51) that radioactive thyroxine, when injected intravenously into normal male rabbits, is concentrated selec- tively in the posterior pituitary. ANTERIOR PITUITARY-ADRENAL RELATIONS Another functional relationship established during the course of development of the em- bryo is that between the anterior pituitary and the adrenal glands. In the main the evi- dence for this relationship is derived from a study of the effects of (a) extirpation or de- struction of the primordia of the anterior lobe or of the adrenal gland and (b) grafting of the anterior pituitary or administration of its extracts from adult animals. ADRENOCORTICOTROPHIC EFFECTS ON DEVELOPMENT OF ADRENAL CORTEX That the development of the adrenal glands is altered after removal of the in- growing primordium of the anterior pituitary in the anuran embryo was first demonstrated by Smith (’20). The cortical or interrenal tissue (in the form of profusely branched cords of cells) is greatly diminished in quantity as compared with that of the normal or thyroidectomized tadpole of the same age (in the thyroidectomized larvae the cortical tissue is hypertrophied). The lipid granules of the cortical cells show a change in re- action to osmic acid, being less intensely blackened as compared to the normal or to the thyroidless tadpole. The diminution in the cortical elements and alteration of the lipid granules may be attributed to the ab- sence of some essential trophic substance, since injection of an extract of fresh bovine anterior lobe into hypophysectomized tad- poles restores the scanty adrenal cortex to normal (Smith and Smith, ’23). That the anterior pituitary exerts a trophic action on the adrenal cortex of the mam- malian fetus and the chick embryo is like- wise clearly indicated. Destruction of the hypophysis of the 13-day mouse fetus by x-irradiation (Raynaud and Frilley, 50) or decapitation of the rabbit fetus at 19 to 22 days (Jost, °48, ’51a), of the rat fetus 2-4 days before term (Wells, ’48) and of the chick embryo at 40 to 50 hours (Case, ’52), usually leads to a marked diminution in volume of the adrenal cortex, and according to Jost equally so in both sexes. Histologically the cortex is atypical. The cortical cells are 585 not only smaller and fewer in number as compared with the normal but also show a diminution in lipid granules and of ascorbic acid as well in the chick. Similarly, a reduc- tion in size of the adrenal cortex occurs in the human fetus in cases of spontaneous anencephaly and cyclopia, but only when the anterior lobe is distinctly smaller than nor- mal or completely absent (see Edmunds, ’50). From the above observations it is clear that the effects of removing the pituitary, irrespec- tive of the means employed, lead to retro- gressive changes in the adrenal cortex of the embryo or fetus. These changes may be attributed to the absence of a specific pitui- tary hormone, since administration of adre- nocorticotrophin (ACTH) to the hypophysec- tomized embryo or fetus restores the adrenal cortex more or less completely to its normal volume and histology (Jost, *51a,b; Kitchell and Wells, 52a; Case, 52). The results not only indicate that the anterior pituitary of the mammalian fetus and chick embryo at least during the later stages of development produces a hormone comparable in action to ACTH, but also suggest that this specific hormone is essential for the normal histo- genesis and growth of the cortical component of the adrenal gland. EFFECTS OF ADRENAL CORTEX ON ANTERIOR PITUITARY Does the adrenal cortex in turn have a reciprocal action upon the anterior pituitary of the developing embryo? According to Tobin (’39) total or partial destruction of the adrenal glands of the rat fetus (17 days and older) by electric cautery results in cellular changes in the pituitary. In comparison with normal controls, the number of acidophiles is decreased, the basophile cells show de- granulation, and the number of chromophobe cells appears to be increased. These results, although suggestive of a reverse influence, need confirmation and extension. That the anterior pituitary is reactive to secretions of the adrenal cortex during late fetal stages of the rat is indicated by the recent investiga- tions of Kitchell and Wells (52b). After unilateral adrenalectomy of the rat fetus on the twentieth day, the intact adrenal under- goes a compensatory hypertrophy that can be prevented by subcutaneous implants of corti- sone. Implants of cortisone to the normal fetus of the same age fail to produce any significant change in the volume of the adrenal or the histology of the cortex. These results seem to indicate that the liberation 586 of ACTH, which appears to be essential for compensatory hypertrophy of the adrenal, is suppressed by the administered cortisone. INTERRELATIONS IN TIME OF ONSET OF FUNCTIONAL ACTIVITY As to the time of onset of functional activ- ity in the anterior pituitary and adrenal cortex during the course of their develop- ment no precise information is available. However, certain inferences may be drawn from the data at hand (see Table 27). The preceding analysis indicates that the primor- dium of the adrenal gland is responsive to ACTH in the rat and mouse during the last third and in the rabbit during the last fourth of the period of gestation. Apparently the period before term is one in which the constituent cortical and medullary cells are assuming step by step an orderly arrange- ment or oriented cell pattern simulating that characteristic of the adult gland. It is a period of active morphogenesis and to some extent histogenesis. Using the mouse fetus as an example, it is seen according to Waring (35) that on about the thirteenth day the cortical and medullary primordia are both constituted and closely adpressed. By the fourteenth day the sympathochromaffin ele- ments have begun to migrate into the cortical component, the immigration con- tinuing gradually until the day of birth, when they are concentrated at the center of the gland. The cells of the cortex at 14 days are all more or less alike (finely granular and highly eosinophilic cytoplasm), and their arrangement is haphazard. Between the sixteenth day and term an oriented pattern of cortical cells arises. The cells at the periph- ery of the cortex become less eosinophilic between the sixteenth and eighteenth days, and at birth have an arrangement which merely foreshadows that characteristic of the glomerular and fasciculate zones of the per- manent cortex. Internal to this apparent fore- runner of the permanent cortex and inter- locking with the medullary cells is a zone of more highly eosinophilic cortical cells, the beginning of the so-called “fetal cortex,” “x-zone” of Miller, or “androgenic zone.” It may be inferred, therefore, that at the time of experimental removal of the pitui- tary in the mammalian fetus the adrenal cortex is clearly responsive to adrenocortico- trophic hormone after the cortical and med- ullary primordia are already constituted and combined, and especially so during the sub- sequent critical stages of morphogenesis and ONTOGENY OF ENDOCRINE CORRELATION histogenesis of the two component tissues. The nature of the effect appears to be that of retarded or arrested morphogenesis, rather than degeneration. No evidence is available as to whether the adrenal at still earlier stages can respond to ACTH. This problem might be resolved by growing the primordia of anterior pituitary and adrenal in juxta- position, either by grafting or by in vitro cul- tures, since activation by local diffusion of hormones appears to be a more sensitive test than activation by hormones that pass through the regular blood vessels of the embryo (see p. 583). Also, since ACTH definitely stimulates adrenocortical growth in the rat as early as the fourth day after birth (Moon, ’40), the time and degree to which adrenocortical growth is stimulated might be ascertained by injection of ACTH into fetuses at successively earlier stages. Whether the growth of the primordia of the permanent cortex or the “fetal cortex’”’ is equally or differentially affected by the ab- sence of anterior pituitary is a problem of much significance. An inhibition of the growth of the permanent cortex is to be ex- pected, since in the dwarf mouse, which has an endocrine-deficient anterior pituitary, the characteristic zoning of the cortex is absent or indistinct (Smith and McDowell, ’30). Possible effects on the growth of the fetal cortex are of special interest in many mam- mals (such as mouse, rabbit, cat, human, certain carnivores and ungulates), since it is proportionally very large, constituting the main bulk of the cortex. In some species its growth is relatively enormous, taking place during either fetal or early postnatal life. This hypertrophy apparently is responsible for the relatively large size of the adrenal gland at birth (Hill, ’30). In man at birth, for instance, the adrenals constitute 0.2 per cent of the entire body weight, contrasting with 0.1 per cent for the adult (Scammon, °30). The only clues that pituitary secretions possibly regulate the growth of the fetal component have been reported by Elliot and Armour (711), who attributed the small size of the adrenal of an anencephalous human infant (at birth) to an almost complete ab- sence of the fetal cortex, and by Deanesley (38), who demonstrated that in pituitary- deficient dwarf mice the x-zone is absent in spite of normal reproductive activity. More- over, in early postnatal life the fetal cortex often shows a distinct sex difference, per- sisting longer and attaining a larger size in females than in males. Although often re- garded as having possible androgenic func- ONTOGENY OF ENDOCRINE CORRELATION tions in sex development, this has not been substantiated for the mouse (Howard-Miller, 27, °39; Howard, °46). According to Gersh and Grollman (739a,b), the function of the x-zone and the more peripheral fascicular and glomerular zones is identical; both respond to stimulation but not equally, the x-zone being less responsive. Furthermore, the wide- meshed capillary plexus of the x-zone is altered by excessive activity so as to resemble more closely the capillaries of the fasciculate zone (Gersh and Grollman, ’41). (For a thor- ough analysis of the relation of the adrenal cortex to the reproductive system in fetal and young postnatal rats, see Moore, 53.) The changes in vascular pattern of the growing cortex offer a possible index of the onset of functional activity (see Figs. 209A, B, and C). During the early stages of adrenal growth in the embryo the arrangement of the cortical cells is haphazard, at which time the capillaries of the cortex are in the form of a loosely woven and irregular plexus. \t later stages with the growth of the cortex, portion of the cortical cells becomes ar- ranged into parallel cords or columns of the fascicular zone, the commonly supposed secretory zone of the adrenal of juvenile and adult mammals. With this change in cellular arrangement the earlier unoriented type of capillary circulation is gradually recon- structed into a more or less longitudinally oriented pattern of vessels, i.e., in rather sim- ple lines the plan of the adult circulation (Flint, ’00; Whitehead, 733). The significance of the changes in vascular pattern may be interpreted in terms of function (Gersh and Grollman, *41). The change is from a type which permits a slow rate of blood flow to one which favors a much more efficient flow of blood, thus assuring a richer supply of blood to the cortex. The fact that the more efficient mode of blood transportation arises as the cortical cells assume an organization with distinctive properties seems to imply a beginning functional relation and one which, furthermore, may reflect the need of the growing organism for the cortical hor- mone in its metabolism. At least, such a view is in keeping with findings on postnatal mammals that in response to increased needs of the organism for cortical hormone the cortex hypertrophies and the vascular pattern of a hyperactive cortex is altered so as to increase blood flow. The question next arises as to when in the course of adrenal development the ori- ented type of circulation arises. According to Flint (’00) the beginning of an orderly 587 niet Fig. 209. Three stages in development of the vas- cular pattern of the adrenal of the pig embryo (adapted from Flint, ’00). A, Stage of 3 cm. showing the simple irregular plexus of capillaries of the cortex, which at this stage is comprised of irregular cords of cortical cells. Arrow indicates direction of blood flow from the fibrous capsule (uppermost surface of figure) to the central venule (v). Precursor medullary cells ap- parently le between cortex and capsule. B, Stage of 8 cm. showing the capillary plexus of two groups of medullary cells (note resemblance to glomeruli), which have invaded the cortex. Each is separately vascularized by an arterial capillary from the capsule. Capillary plexus of cortex is still irregular. C, Stage of 22 cm. (near birth) showing the ori- ented pattern of the capillary plexus of the cortex, thus reflecting the initial zonation of the cortical tis- sue. The definitive topographical relation of cortex (cort.) and medulla (7med.) is initially established at about the 12 cm. stage, when the invading groups of medullary cells reach their terminal position adjacent to the central vein (v). 588 arrangement of the capillaries is first noted in the pig fetus (12 to 15 cm.) at a stage when the definitive topographical relations of cortex and medulla are first established. During subsequent stages the capillary plexus assumes gradually a more orderly arrangement, and at a stage just before birth (22 cm. fetus) it has a distinct longitudinal orientation, although still of a much simpler pattern than in the adult (see Fig. 209C). Species variations in the time of origin of an oriented pattern of circulation are ex- pected. In the mouse, for example, the per- manent cortex is much less well organized than it is in the pig at birth. Subsequent to birth in the mouse the cortical cells (pe- ripheral to x-zone) gradually assume the form of parallel cords in the fasciculate zone, only becoming distinct at about 25 days after birth. On the assumption that column for- mation and orientation of circulation are causally related, the capillary plexus should gradually change to a longitudinal orienta- tion. By the fourth week, at any rate, the pattern of circulation is clearly an oriented one (Gersh and Grollman, ’41). A further criterion of beginning functional activity is the time of appearance, rate of accumulation, and identification of the kinds of lipids of the cortex during development. It has commonly been inferred that the lipids of the cortical cells may be the ve- hicle for cortical hormones, since they are obtained primarily from the lipid fraction of extracts of adult adrenal glands. In the hu- man embryo lipids in the form of fine drop- lets are already present in the cortical cells at the stage (14 mm.) of initial penetration of the future medullary cells into the cortical primordium. Generally speaking, during sub- sequent stages in adrenal development the amount of sudanophilic lipids seems to in- crease in the permanent cortex and to dimin- ish in the fetal cortex. In the latter the lipids gradually diminish during the later weeks of intra-uterine life, disappearing altogether at full term (in the postnatal mouse the homologous x-zone prior to onset of its de- generation is likewise free of sudanophilic lipids according to Howard-Miller, ’27). Con- comitantly with the active growth and or- ganization of the layers of the permanent cortex the amount of lipid increases. An increase in lipids over earlier stages is in- dicated at birth when the cortical cells ‘show a beginning arrangement into columns, a further increase at 3 weeks after birth when the fascicular zone is better defined, and a well-marked increase in the glomer- ONTOGENY OF ENDOCRINE CORRELATION ular and fascicular zones of a 3-month-old infant (see Keene and Hewer, ’27; Uotila, ’40). Although highly suggestive, this increase in sudanophilic lipid, which apparently at- tains its highest concentration at about the time the permanent cortex acquires a well- defined zonal organization, cannot be re- garded by itself as a reliable indicator of cortical hormone production. Sudan III does not discriminate among the various kinds of lipids. Neither does osmic acid, another com- monly used reagent for detecting lipid sub- stances in tissues. Nevertheless such histo- chemical tests, although not useful alone for a differential discrimination of lipids, may be valuable indicators when applied to the adrenal in which changes in cortical activity have been induced experimentally. For example, in the postnatal rat a definite correlation exists between cortical activity and reduction of osmic acid. The reduction of osmic acid is increased by stimulation (in- duced by unilateral adrenalectomy, cold, etc.) and decreased by depression (induced by injection of variable amounts of cortical hormone) of cortical activity (Flexner and Grollman, ’39). Furthermore, the same type of dynamic change in sudanophilic lipids (possibly cholesterol esters) occurs in the adrenal cortex of rats following administra- tion of ACTH or subjection to stress. These changes parallel the alterations in cholesterol content (Sayers et al., ’44). Still another possible key to the onset of adrenal cortical activity in the fetus is the time and rate of accumulation of the vita- min, ascorbic acid, which along with choles- terol esters occurs in high concentration in the functionally active adrenal of juvenile and adult mammals. Only a few attempts have been made to determine the presence of ascorbic acid in the developing adrenal of any vertebrate. By means of a colorimetric method Case (752) found that the quantity of ascorbic acid rapidly increases in the chick adrenals from the twelfth to the nineteenth day of incubation, rising from a value of 0.95 to 1.67ug. per milligram of adrenal tissue during this period. The further observation that the total amount of ascorbic acid increases relatively more rapidly than does the net weight suggests that with ad- vancing age the adrenals acquire an increas- ing capacity to accumulate ascorbic acid. Unfortunately the initial time and rate of accumulation have not yet been determined for adrenals of embryos earlier than the twelfth day (cf. Dawson, 53). The changes ONTOGENY OF ENDOCRINE CORRELATION in ascorbic acid content during the period under consideration apparently constitute a reliable indicator of functional activity, since in pituitaryless embryos the quantity of as- corbic acid does not rise but remains at a level characteristic of the adrenals of the normal 12-day embryo. Treatment of such pituitaryless embryos with ACTH brings about an apparent increase in sudanophilic lipids of cortical tissue, but whether ascorbic acid is likewise increased has not been de- termined. The lack of information on the possible occurrence of ascorbic acid in the adrenal of the mammalian fetus is a challenge. In juvenile and adult mammals the concentra- tion of ascorbic acid (as well as of choles- terol) is regulated by the anterior pituitary. The administration of ACTH and _ stress (several varieties) both bring about a marked reduction or depletion of ascorbic acid (for a recent consideration and references see Sayers and Sayers, *49, and Long, ’49). In all probability ascorbic acid accumulates along with sudanophilic lipids which, as noted above, occur in the developing adrenal cortex. Microchemical and _ histochemical tests for the presence of ascorbic acid, es- pecially if correlated with alterations in its content under conditions of increased acti- vation, should prove to be valuable methods for the evaluation of the degree of cortical activity, if any, in the fetus or during early postnatal life. Progressive changes in the mechanism of response are indicated in in- fant rats of increasing age. According to Jailer (50) the ascorbic acid content of the adrenal declines (24 to 37 per cent) in re- sponse to the administration of ACTH to animals 4 to 6 days of age, but no decline occurs upon exposure of the animals to a temperature of 5°C. (stress) until the six- teenth day; also, the administration of adren- aline (epinephrine) causes no increase in ascorbic acid until the eighth day. Further, using fall in ascorbic acid content as an index of activation, no ACTH activity in the anterior pituitary of infant rats is de- tectable until the eighteenth day of age post partum (Jailer, °51). Such combined methods of attack now seem feasible for determining cortical activ- ity in the fetus, since at least for a short period prior to birth the adrenal cortex is re- sponsive to adrenocorticotrophic hormone as noted above. In a similar way the changes in birefringent substances may furnish addi- tional clues (Yoffey and Baxter, ’47). Special attention should be directed to the histo- 589 chemical detection of cholesterol esters, cor- tical sterones, and ascorbic acid, substances known to be present in the functionally ac- tive cortex of the adult mammalian adrenal (Yoffey and Baxter, ’49). Whether the cortex of the mammalian embryo at any stage con- tains these substances or chemically related ones has not yet been established. Tests for the occurrence of these substances would have a twofold purpose: (1) to ascertain the period when the growing cortex acquires the capacity to select steroid mother substances from the blood and synthesize them into cortical sterones or allied compounds, and (2) to determine whether such capacity is a key to the approximate time when adreno- corticotrophic substance is released by the anterior lobe in quantities sufficient for cor- tical activation. Progress along these lines has been made by Dawson (’53) in a study of the time course in histochemical differ- entiation of the adrenal gland of the chick embryo. As judged by their reactions to given chemical agents or by birefringent proper- ties, several kinds of lipid compounds appear in some cords of the gland as early as the seventh day but are not common to all cords until the eleventh or twelfth day; on the basis of the silver reaction, ascorbic acid, although first detectable on about the tenth day, is not present in appreciable quantities until the twelfth day. Turning next to the functional activity of the anterior pituitary, very little is known concerning the time when adrenocorticotro- phic substance begins to form, the rate of its increase, or whether its production is cor- related with the differentiation of cell types (see Table 27). That the anterior pituitary is active in producing and releasing ACTH during the last third or quarter of fetal life in the rat and rabbit and of the last third of the embryonic period of the chick is clearly indicated, since within that period the growth of the adrenal cortex is definitely arrested in the absence of the pituitary and can be restored to normal or partially so by administration of adrenocorticotrophin. Apparently, then, for a short period prior to birth or hatching the trophic hormone in question is present in sufficient concentration to be physiologically effective. This may mark the peak production during embryonic life of the species under consideration. It is of interest to note parenthetically that at about this time signs of thyroid activity are first noted in the rat fetus. In this species the thyroid does not begin to form follicles and store colloid until 3 to 4 days prior to birth, becoming more 590 marked immediately before term (Kull, ’26). It is within this period, i.e., at 18 to 19 days of gestation, that the thyroid first acquires the capacity to con- centrate radioactive iodine; the capacity is consid- erably stronger at birth (Gorbman and Evans, 43). Also, between the 1914 day and term, the thyroid of normal and pituitaryless rat fetuses is reactive to subcutaneous injections of thyrotrophin (Sethre and Wells, 51). From the presumptive evidence at hand, it would appear that in the rat fetus the endo- crine functions of the thyroid, anterior lobe, and adrenal cortex begin relatively late, 1.e., at a time when nearly 90 per cent of the intra-uterine life is completed. Whether the anterior pituitary at early stages has ACTH activity is not known but might be tested in the manner suggested on pp. 582 and 583. Likewise, no information is available on whether the production of ACTH increases quantitatively with develop- mental age. It probably increases rapidly after birth, judging from the marked changes that take place in the growth and vascularity of the adrenal cortex in the rat, mouse, and man. No systematic attempt has been made to correlate the onset of adrenocorticotrophic activity with the time course of differen- tiation of the cell types in the anterior pituitary. Of those species in which trophic activity is indicated, only in the anuran, chick, and rat embryos has the occurrence of cell types been reported. In the anuran no relationship is ascertainable since the anterior lobe is removed at a very early stage, long before the histogenesis of cell types begins. Although in the rat, acidophiles, basophiles, and chromophobe cells are al- ready differentiated in fetuses of 18 to 23 days (Tobin, ’39), the time order of their differentiation and numerical frequency have not been investigated. In the chick a close correspondence is seen between the indicated period of adrenocorticotrophic ac- tivity and active histogenesis of the cell types. Subsequent to the tenth day, when they are first apparent, the acidophiles rap- idly increase in number, becoming the pre- dominating type of cell by the eighteenth day (Rahn, ’39) or at most by the third day after hatching (Payne, ’46). Although baso- phile-like cells seem to be present from the twelfth day on, typical basophiles are not found until after hatching (cf. Wilson, ’52). That the cellular source of the hormone in the embryonic development of the pitui- tary is by no means clear is not surprising, since even in postnatal mammals at various ages it has been difficult to assign a specific hormonal secretion to any particular cell ONTOGENY OF ENDOCRINE CORRELATION type. Reports as to the cellular source of ACTH are diverse and often contradictory, as the following selected examples show. Presumptive evidence that at least one cell type plays a role in the secretion of adreno- corticotrophin is found in the hereditary dwarf mouse (homozygous for a _ recessive dwarfing gene), in which adrenal cortical aplasia is associated with the complete ab- sence of typical acidophiles in the anterior pituitary (Smith and McDowell, ’30). Al- though these authors were uncertain as to the identification of the other cell types, later studies by Francis (’44) on the fully grown dwarf show that the basophiles are likewise absent and the number of typical chromophobes is markedly reduced (small pyknotic acidophiles and chromophobes com- prise the predominating types of cell). More recently, Finerty and Briseno-Castrejon (49) suggested that the acidophiles secrete ACTH since a marked increase in the percentage of these cells occurs in the anterior pituitary, following unilateral adrenalectomy in the immature male rat. Still more recently, Mar- shall (51) was able to show by means of a fluorescent antibody technique adapted to the localization of antigens in cells and tissues that a solution of fluorescent antibody (to pig ACTH) stains selectively the cyto- plasm of basophile cells of the pig pituitary. The reaction is highly specific, since cells of the sheep or bovine pituitary do not stain. In any investigation of the problem of the cellular source of the hormone in pituitary development special attention should be given to the time order of differentiation of the cell types as well as to their progressive changes in numerical frequency and in cyto- logical characteristics. Such an approach, when combined with experimental pro- cedures designed to bring about an imbal- ance in the normal functional interaction of the adrenal cortex (and other receptor glands) and anterior pituitary, should prove of value in elucidating the problem. For the specific localization of ACTH in cells during the histogenesis of the anterior pi- tuitary the fluorescent antibody technique is undoubtedly the most promising of all. ON THE ROLE OF ADRENALINE AS A CORRELATING HORMONE IN THE PITUITARY-ADRENAL SYSTEM As may be seen from an inspection of Table 27, the substance adrenaline* arises * According to Shepherd and West (751). in fetal (near term) and early postnatal stages of various ONTOGENY OF ENDOCRINE CORRELATION as a rule very early in embryonic life. Both the chromaffin reaction of Henle and the intestinal strip method agree in showing that adrenaline is first detectable at a time when small groups of future medullary cells begin eM islands of cells proceeds progressively through the cortex toward their definitive topographical position, the quantity of adren- aline gradually increases. In some species the relative quantity present in late fetal Fig. 210. Portion of adrenal cortex with interlacing cords free of intervening medullary tissue. Developed in a chorio-allantoic graft from a small isolate just back of the primitive pit of a chick blastoderm of the early head-process stage. Initial < 450. to enter the cortex from the contiguous sym- patho-chromaffin primordium (cf. Figs. 209B, 210, and 211). As the migration of these mammals (cat, guinea pig, and rabbit) the adrenal glands contain a high proportion of noradrenaline and a small quantity of adrenaline, and, further, the relative amounts of these two substances change with advancing developmental age. The observa- tion that the large amounts of noradrenaline present in the adrenals of fetal and young postnatal mam- mals generally diminish as the adrenaline con- tent increases in still older animals led to the suggestion that the amine noradrenaline is not only a precursor of adrenaline but indeed may be the hormone of the adrenal gland “in the early days of life.” by hormonal substances? life may be distinctly greater than in the adult adrenal (McCord, °15; Fenger, °12). This is in keeping with the observation that in many species of mammals the relative volume of the medulla to the whole adrenal is high at birth, rapidly decreasing shortly thereafter until puberty, when a stationary condition is reached (Donaldson, °19). What is the significance of such an early appearance and quantitative increase in adrenaline? Could they possibly reflect the “needs” of the growing embryo for adren- aline in the operation of mechanisms, es- pecially of those dependent on or controlled Paralleling this 592 appearance and increase of adrenaline is a peculiar and striking vascular change asso- ciated with the differentiation and migration of the future medullary cells. As shown in Figure 209B, the migrating spherical islands of chromaffin cells in the cortex are provided with a distinctive arrangement of the capil- Case BPs 4 ONTOGENY OF ENDOCRINE CORRELATION a <) sae PIVayoy 3y € wae (5 Dine Bh ji eee gradually increases, a relationship of a func- tional nature. Nothing is definitely known as to the function of adrenaline during fetal life. It is significant to note, however, that almost from the beginning the development of the adrenal medulla and of the sympathetic Fig. 211. Section of an adrenal gland showing partial invasion of medullary cells from a mass of sympa- thoblasts (s). Differentiated from an isolate of the mesonephros-gonad-adrenal complex taken from a 4-day chick embryo and grown for 9 days on the chorio-allantoic membrane of a male host embryo. c, Group of cortical cords free of intervening medullary cells; m, medullary cord; w, portion of mesonephros. Initial X< 200. (From Willier, *30.) laries, which, although continuous with those of the cortex, have a smaller meshwork. In both instances the blood is derived from the arterioles in the capsule. As the islands increase in size and move inward toward the central vein the capillary network of each of them is provided with a separate arteriole. For a short time the capillaries of the cortex act as venules from the capillaries of the island (Flint, 00). These changes in vascular pattern are of a kind that permit a more efficient flow of blood to the islands of chromaffin cells. Apparently as the mode of blood transportation becomes more efficient the adrenaline content of the adrenal gland nervous system is interlocked. In the course of their functional development the former becomes in effect an integral part of the latter (constituting the so-called “‘sympatho- adrenal system’). It would be of much sig- nificance to determine at the time of first appearance of adrenaline, as well as during its increase in the fetal adrenal, whether splanchnic nerve fibers (preganglionic) al- ready terminate around the individual chromaffin cells, as they do in the adult. The possibility, however, that the chromaffin cells may form and store adrenaline as chromophile granules before innervation is established must be recognized, since these ONTOGENY OF ENDOCRINE CORRELATION processes can continue for a while in the denervated adrenal of the adult. On the other hand, the rate of formation and release of adrenaline into the blood in the adult are partly regulated by the nervous system. The approximate time of release including quan- tity present in the blood stream of the em- bryo or fetus could be readily determined, since bio-assay methods sensitive to concen- trations of adrenaline of 1 in 500,000, or less, are available. Once adrenaline is liberated into the blood stream, which seems likely to occur toward the end of embryonic life, the question arises as to its interrelations with other substances within the embryo. (See p. 609.) It is well known that, as a hormone, adrenaline has a highly selective action on effectors innervated by the sympathetic nervous system. Of partic- ular interest here, however, is that on the basis of recent investigations (see Long, *49, for evidence and theoretical considerations) it plays a role in the endocrine activity of the adrenal cortex. The administration of adrena- line as well as its liberation following stimu- lation of the sympathetic nervous system causes an increase in adrenal cortical activity. Although adrenaline acts as a potent stimulus to adrenal cortical activity, its mechanism of action is unknown. According to the theory of Long its action is indirect, involving the correlation and interplay of anterior pitui- tary, adrenal cortex, and medulla. The par- ticular merit of this theory is that it combines in one comprehensive picture a variety of factors or conditions, such as activity of the sympathetic nervous system (adrenaline re- lease), adrenocorticotrophic hormone, and stress, all of which are known to call forth secretion by the cortex. Their interaction constitutes an adaptive mechanism enabling the organism to adjust itself to stresses which may arise within or without it. It is con- ceivable, therefore, that such an interplay of anterior pituitary—adrenal functions as they become established in the embryo creates an internal milieu essential for maintaining a normal state of balance, and in preparing the organism for exigencies appertaining to birth and early postnatal life. ANTERIOR PITUITARY-GONAD RELATIONS The gonad is the third component of the endocrine system with which the anterior pituitary becomes interrelated during the course of ontogeny. The concept of anterior pituitary—-gonad relations is based almost 393 entirely on extensive researches on immature and adult vertebrates. Experimental proce- dures designed to determine the effects of removal of one gland upon the other and/or in combination with implantation of either gland or the administration of their active principles as well as the effects of anterior lobe implants on sexually immature animals have yielded evidence which clearly estab- lished for many species that specific gonado- trophic secretions of the anterior lobe are essential to the maintenance of structure and function of the gonads, which in turn by secreting sex hormones affect receptor sex structures such as the gonoducts and acces- sory glands. Moreover, sex hormones re- leased by the gonads have a reciprocal effect on the anterior pituitary, regulating release of gonadotrophic hormones. (For supporting evidence and citations see Engle, ’39; Fevold, °39; Moore and Price, 32; Smith, ’39.) The main problem for consideration here is the sequence of events in the establishment of functional relations between anterior pi- tuitary and gonad. Particular attention will be given to such specific problems as (a) the time and degree to which the entire course of gonad development is dependent upon gonadotrophic secretions, and (bd) whether the onset of production of sex hor- mones and their effects on receptor sex struc- tures precedes or coincides with beginning anterior lobe activity. ONSET OF SEX-HORMONE ACTIVITY OF THE GONAD As a basis of approach to the over-all] problem of the development of functional interaction between the gonad and the an- terior pituitary, the time course in the dif- ferentiation of functional activity of the gonad will be examined first. At what period are sex hormones initially produced? When are they released into the blood circulation in quantities sufficient to act on such ter- minal receptors as the seminal vesicle, gono- duct, and phallus, sex structures which react selectively to sex hormones? These questions can be best answered by a consideration of the following several lines of evidence (also refer to Table 27). 1. As is known from the work of Lillie (17), the vascular connections of the fetal membranes of twin embryos in cattle usually become suitable for the intermingling of the blood of the co-twins either prior to or at about the time that structural sex differ- entiation of the gonad sets in (25-mm. 594 stage). If the embryos are of opposite sex and vascular connections are established between them, the female partner is modified in the male direction, forming the freemar- tin. Since the structural changes are limited to the production of male sex characters in the female co-twin, it has been inferred that the gonads of the male co-twin not only produce but release male sex hormones into the common blood stream at a very early stage. (For certain shortcomings in the purely hormonal interpretation of the free- martin condition see Owen et al., 46; Moore, 47; and Anderson et al., ’51.) 2. In the chick, by implanting fragments of a sexually differentiated gonad (6- to 11-day embryo) into the coelom of a_ host embryo of 50 hours (preceding by about two days the time of origin of the gonad primordia), Wolff (’47) succeeded in bring- ing about an approximation of grafted tissue and host gonads. If of opposite sex, the grafted tissue modifies the direction of sex development of the reproductive organs of the host. Ovarian tissue causes the formation of ovarian cortex on the left testis; testis tissue has little effect on the left ovary but inhibits or completely suppresses the Miiller- ian ducts. The degree of sex transformation is dependent upon the volume of the graft and the distance between graft and host gonads. In general, the closer the graft, the greater is the degree of effect. Similar re- sults are obtained by cultivating in vitro duck gonads of sexually indifferent stages in juxtaposition. If left gonads of opposite sex constitution are in more or less intimate contact, the testis is modified into an ovo- testis by the ovary (Wolff and Haffen, ’52). Since the effects produced in hosts by grafts of embryonic testicular or ovarian tissue and in explants of associated left gonads of oppo- site sex are identical in character with effects produced by the injection of synthetic sex hormones, it is apparent that the embryonic differentiated gonads produce sex hormones (for further details and argument see Wil- Iiters; 2395752). 3. In the rat, according to Jost (’50), a fetal testis (15 to 16 days old) grafted upon the atrophied seminal vesicles of an adult castrate produces within five days an in- tense but local activation of the seminal epithelium identical in character to the effects of injecting testosterone, Tissues other than testis fail to produce the effect. Al- though androgenic activity of the fetal testis is shown, the activity seems to depend upon gonadotrophic stimulation, particularly since ONTOGENY OF ENDOCRINE CORRELATION an embryonic testis graft has little or no effect on the seminal epithelium in hypo- physectomized and castrated adult rats. 4. Castration by localized x-irradiation of a duck or chick embryo during the sex- ually indifferent stage indicates that hor- mone secretions of the early differentiated gonad are essential in determining the di- rection of differentiation of other sex struc- tures. An asexual or neutral type is obtained following such precocious castration (Wolff and Wolff, 51). In both males and females the genital tubercle and syrinx (duck only) assume the male shape and oviducts persist. These results imply that in the normal em- bryo (a) the female form of the genital tubercle and syrinx is conditioned by the presence of ovarian hormone and (b) the retrogression of the oviducts is dependent upon the presence of testicular hormone. These findings are in agreement with the results of injecting sex hormones, male hor- mones suppressing oviducts in genetic fe- males and female hormones stimulating ovi- ducts in genetic males. (For a more complete analysis of the time and manner of release of sex hormones in bird embryos, see Willier, 62.) Similarly, the effects of surgical castration of the rabbit fetus indicate that the direction of differentiation of gonoducts and accessory sex structures is influenced by hormonal secretions of the embryonic gonads (Jost, 47, °50). If the gonads are removed from fetuses of either sex before the initiation of structural sexual differentiation (18 to 19 days of gestation) the entire genital tract is feminine in form. In males the develop- ment of the Wolffian ducts and _ prostatic buds is entirely suppressed, the oviducts per- sist, and the external genitalia are feminine in form. The Wolffian body retrogresses as in normal females. However, castration of male fetuses at successively later stages (20 to 24 days of gestation) brings out the ad- ditional point that sex structures, already acted upon by testicular hormones, retain the male type although usually in an unde- veloped state. Synthetic androgens admin- istered to castrates have a reparative influ- ence on sex structures which simulates that of the embryonic testis hormones. In cas- trated females, on the other hand, the genital tract retains a feminine form essentially like that of control females of the same age. Apparently, then, the embryonic testis pro- duces male sex hormones which contro] the development of the male form of the ex- ternal genitalia and Wolffian bodies, the ONTOGENY OF ENDOCRINE CORRELATION formation of the prostate, the persistence of the Wolffian ducts, and the retrogression of the oviducts. In genetic females the further development of the oviducts is conditioned by ovarian sex hormones. More recently Wells and Fralick (’51) have shown for the fetal rat that surgical removal of the testes on the twentieth day of gestation selectively retards the growth of the seminal vesicles and_ bulbo-urethral glands and, further, that such specific cas- tration effects are prevented by the adminis- tration of testosterone propionate. From the growth responses thus given by the accessory sex organs it is at once clear that for at least a few days prior to birth the testes are active in the production and release of androgenic hormone. In summary, the embryonic gonads from the time of onset of morphological sex differ- entiation are active in the production and/ or release of specific sex hormones. The hor- mones are liberated in quantities at least sufficient for effective action through the vascular circulation as judged from the na- ture of the responses, especially those of the terminal sex receptors. (For a further treat- ment of the subject see Burns, °49; Jost, 50; Moore, *50.) ONSET OF GONADOTROPHIC ACTIVITY OF ANTERIOR PITUITARY The evidence just presented for the exist- ence and action of sex hormones early in embryonic development naturally poses a two-fold problem of (1) the extent to which their production and release are dependent upon gonadotrophic stimulation and (2) the time that the anterior pituitary begins to secrete gonadotrophic substances as well as to release them into the blood circulation of the embryo (see Table 27). Whether functional activity of the gonad during em- bryonic life is dependent upon gonadotrophic hormones is a problem which is as yet only beginning to be resolved. The most complete and satisfactory analysis yet made is that of Jost (48, ’51b) on the rabbit fetus. If the male fetus is decapitated at the time of initial structural sexual differentiation (about the nineteenth day of gestation) the inter- stitial tissue of the testis is distinctly reduced as compared with normal controls of the same age. Moreover, testicular hormones are quantitatively deficient, as is indicated by the feminization of the external genitalia and by the arrested development of the prostatic buds as shown in Figure 212. Since 395 similar effects on the prostate and external genitalia are produced in males castrated at 21 to 22 days, it is inferred that the re- moval of the hypophysis affects the produc- tion of male hormone by the testis. The effects of hypophysectomy are attributed to the absence of specific hormones, since if at the time of decapitation gonadotrophic hormone is administered no signs of in- sufficient testicular secretions are found. These results are interpreted as indicating that the anterior pituitary of the rabbit fetus is active in producing and releasing gonado- trophic hormones either at the time of or shortly after the onset of testis differentia- tion. They are apparently released in suffi- cient quantities to be physiologically effective in maintaining testicular hormone production at a level adequate for regulating the devel- opment of such sex characters as the prostate and external genitalia during their initial stages of growth and differentiation. In other species of vertebrate embryos the evidence thus far available is inconclusive on whether secretions of the anterior lobe are essential for the early development and func- tion of the gonad. In general, so far as results on different species can be compared, re- moval of the anterior lobe has little or no effect on the early morphological develop- ment of the gonads. Two examples suffice to illustrate this point. In the anuran embryo removal of the anterior lobe at the time of its ingrowth has no apparent effect on the erowth of the gonads; their development continues a normal course to the stage which is attained in controls at the time of meta- morphosis (Smith, ’39, p. 942). In the chick embryo after hypophysectomy at 33 to 38 hours primary morphological sex differentia- tion proceeds in a normal manner until about the thirteenth day, when reductions in inter- tubular tissue of the testis and in ovarian cortex are first noted; however, the gonoducts in both sexes develop normally both in time and manner (Fugo, ’40). On the contrary, Wolff and Stoll (37) report that the entire course of sex-differentiation of gonads and eonoducts is normal up to the time of hatch- ing in the cyclocephalic chick embryo with- out hypophysis produced by local x-irradia- tion at the 12- to 15-somite stage. By the injection of estradiol into such pituitaryless chick embryos intersexual males are produced as readily as in normal male embryos (Wolff. °37). It seems, therefore, that the functional processes concerned in sex-inversion are wholly independent of hypophyseal action. Whether the gonad during early stages of 596 formation could be stimulated by hypo- physeal implants or injection of gonado- trophins is conjectural. Does the production of gonadotrophins in- crease quantitatively with advance in devel- opmental age? As regards the _ relative gonadotrophic potency at various embryonic ages the available evidence is not easily assessed, inasmuch as different species of animals and different indices of sexual maturity (1.e., precocious ovulation or open- ONTOGENY OF ENDOCRINE CORRELATION elicited in the genital system of the mouse even though the dosage is doubled. Further bio-assay studies are needed at successively later stages in order to ascertain whether the quantity of gonadotrophin continues to increase with fetal age and at what rate of increase. After birth a gradual increase in gonado- trophin content is clearly evident. Using, for the bio-assay test, the immature female albino mouse at a stage (20 to 22 days of age; 6 rot i Me ose, Fig. 212. The prostatic region of two male rabbit fetuses of the same age (28 days). A, Decapitated at 19+ days; B, normal litter male control. Note the arrested development of the two anterior prostatic buds in A as contrasted with their ramified stage in B. (Courtesy of Jost, ’48.) ing of vagina, increase in gonad weight, etc.) were used for the bio-assay test and, moreover, the unit quantity of implants of anterior pituitary tissue either varied greatly or was not considered. Although the data are inadequate for measuring relative potency, they in general do indicate the presence of gonadotrophins in the anterior pituitary of mammals (including the human) at the more advanced stages prior to birth. The most convincing evidence of their quantitative increase during fetal life has come from a study of the pig fetus. Using the immature mouse as the test animal, Smith and Dortz- bach (29) found that gonad-stimulating hormone is present in readily detectable amounts in anterior pituitaries of pigs of 25 cm., in lesser amounts at the 20- to 21-cm. stage, and none earlier than the 17- to 18-cm. stage. Prior to the last stage no response is to 8 days prior to the probable time of onset of sexual maturity in the normal) when the genital system is sensitive to small amounts of anterior pituitary, Clark (’35) determined the amount of gonad-stimulating hormone in the anterior lobe of the rat from the day of birth up to and including sexual maturity. Using increase in ovarian weight as the index of activity, the changes in content with increasing age were determined sepa- rately for each sex. The activity of the female pituitary increases rather gradually during the first week after birth and then rises sharply between the thirteenth and twentieth days (the probable period of most extensive follicular growth in the normal rat), reach- ing at that time a level of activity per unit weight of anterior pituitary which is not excelled at subsequent ages even after full sexual maturity is attained. On the other ONTOGENY OF ENDOCRINE CORRELATION hand, the activity of the male pituitary from the time of birth rises gradually and con- tinuously up to the age of puberty, attain- ing at that time a high level which is more or less uniformly maintained thereafter. Similarly, in the rabbit the activity of ante- rior pituitary tissue, as revealed by the rabbit ovulation test, is relatively low at 4 weeks of age but at 3 months is as high as in adult animals (Wolfe and Cleveland, ’31). In sum, so far as the evidence for mammals permits generalization, the gonadotrophic hormones apparently increase quantitatively with advance in developmental age, begin- ning during the later fetal stages, often attaining a relatively high level at birth, and usually the highest level prior to the onset of sexual maturity. Possible exceptions to this generalization may depend, even when the unit dosage is constant, upon individual and species variables in gonadotrophin con- tent and in the degree of sensitivity of the responding test organ of the recipient animal. A related problem for consideration is the period in the course of development in which the gonad, as a receptor gland, acquires the capacity to respond to administered hypo- physeal hormones. Is gonad responsivity gradually or suddenly attained? In the classic investigations of Smith (27) and Smith and Engle (’27) it was found that the number of implants of anterior pituitary required to induce precocious sexual maturity in male and female rats and mice is higher in im- mature than in older animals, the number required being roughly inversely proportional to the postnatal age of the host. Thus, in general, as the recipients approach puberty not only fewer implants but less time is required to obtain the response. These ob- servations clearly suggest that the gonads undergo developmental changes (maturation) which enable them to respond gradually more rapidly to hypophyseal stimulation. Subsequent investigations have fully es- tablished the concept of a gradual increase in responsivity of the gonads after birth to gonadotrophic stimulation. The time course of development of gonad reactivity has been particularly well worked out in the postnatal rat by Price and Ortiz (44). In the adminis- tration of equine gonadotrophin particular care was taken to keep the dosage and length of treatment constant, the main significant variable being the age of the recipient, which ranged from the day of birth to maturity. The degree of response at a given age was ascer- tained by comparing with normal controls the induced changes in weight and histology 597 of the gonads and accessory reproductive organs, the latter being end organs which reflect changes in amount of sex hormones liberated. Using these criteria, the ovary was found to be only slightly responsive between the day of birth and the sixth day of age; between the fourth and tenth days a definite though small response was indicated; and at succeeding ages the response gradually increased until a maximum was attained at 26 days of age. The response of the testis is definite though relatively small between birth and the sixth day. At successively later ages the reactivity of the testis increases rapidly and reaches a peak at 14 days. En- hanced production of sex hormones by the gonads is indicated by the response of at least some of the accessory organs as early as the sixth day in males and the tenth day in females. In general, the capacity of the gonads and accessory sex structures for precocious structural differentiation develops gradually in response to gonadotrophic stimu- lation. Thus, in placental mammals (rat and mouse) subsequent to birth a correlation exists between the developmental age of the gonads and their degree of responsivity to gonadotrophins. Since within a few days after birth of the rat the response is relatively weak, especially in females, it might be expected that during fetal stages the gonads would exhibit little or no response to administered gonado- trophin. Such appears to be the case, since subcutaneous injection of equine gonado- trophin into fetal rats near term fails, during the short time interval allowed for action, to bring about any significant increase in growth of either the ovary or the testis; however, the testis is otherwise stimulated as is shown by a significant increase in the size and number of the interstitial cells. No increase in androgen secretion is indicated, since enhanced development of the prostate and other accessory sex structures is not apparent (Wells, ’46). The effects of treating young opossum in the pouch at different ages (beginning on the eighth day after birth) for short periods of time with equine gonadotrophin reveals the time at which the gonads are responsive to the dosages used (Moore and Morgan, ’43). Capacity to respond occurs earlier in males than in females. In males precocious develop- ment of the prostate is readily produced sub- sequent to but usually not earlier than the seventieth day after birth. Since the response of the prostate is a reliable indicator of androgenic activity it is evident that the 598 production and release of male sex hormones by the testis are enhanced by gonadotrophin. In older treated males interstitial cell hyper- trophy occurs but spermatogenesis is not hastened. Using the marked response of the uterus as a sign of increased liberation of estrogens by the ovary, there is no apparent stimulation of the ovary by gonadotrophin prior to the hundredth day after birth. How- ever, at the age of 125 days the ovary is def- initely stimulated in follicular development and in luteinization and, as is signalized by the accelerated development of the uterus, in the production and release of ovarian hormone. The gonad of the chick likewise shows an increase in Capacity to respond to gonado- trophic hormones with advance in develop- mental age. According to Domm (737), in chick embryos examined at 18 days following daily injections of pituitary extracts (““He- bin”), beginning at a time just prior to and continuing through the period of morpho- logical sex differentiation (5 to 9 days of incubation), the gonads in both sexes are hypertrophied, apparently as a result of a marked increase in intertubular tissue of the testis and of medullary tissue in right and left ovaries. There is, however, no significant advance in the differentiation of either the seminiferous cords or the ovarian cortex. Although structural changes are produced, little or no enhancement of sex-hormone activity is indicated, since neither the comb nor the gonoducts in either sex show any significant change from the normal. The responsiveness of the gonads in newly hatched chicks to daily pituitary injections is still further increased with certain striking sex differences. In contrast to the embryo, a precocious spermatogenesis sets in and sex hormone is actively released by the testis, as is indicated by a precocious onset of male sex behavior, an accelerated growth of the head furnishings, and a hypertrophy of the ductus deferens. In the female, although no increased follicle growth in the ovarian cor- tex is apparent, the oviduct is hypertrophied, an indication of enhanced production of female sex hormone. Curiously, whereas the right ovary is unmodified, the medulla of the left ovary shows a pronounced hypertrophy. Since medullary hypertrophy is regularly associated with an accelerated growth and masculinization of the head furnishings, it may be assumed that the left ovarian me- dulla is the source of male hormone under the conditions of the experiment. Although the time of initial response to ONTOGENY OF ENDOCRINE CORRELATION administered gonadotrophins has not been ascertained, it is clearly evident from the above data that the gonads, which are al- ready highly reactive by the eighteenth day of incubation, become still more so in newly hatched chicks and probably reach a Maximum in still older chicks. Apparently, then, as the gonads undergo developmental changes, their reactivity to hypophyseal stimulation progressively increases. In the salamander (Amblystoma tigrinum) the time sequence in the development of gonad reactivity has been worked out fairly completely by Burns (734) and Burns and Buyse (731, °33, ’34). The administration of hypophyseal substances to young larvae (30 to 35 mm. long) at the stage of sex differen- tiation readily stimulates precocious growth and maturation. Gametogenesis sets in pre- cociously in both sexes but to a much greater degree in males than in females, indicating a sex difference in time of attainment of the responsive state. However, soon after meta- morphosis at a time when the gonads are still very immature (sexual maturity in the normal is attained many months later), the growth of the testis and of the ovary as well is greatly enhanced by hypophyseal extracts. In males spermatogenesis reaches a peak in- cluding the formation of mature spermatozoa; in females many of the ova are practically mature, although falling short of nuclear maturation and ovulation. It is thus apparent that the gonads of post-metamorphic stages have acquired a marked capacity for pre- cocious growth stimulation in sharp contrast to a much more limited yet definite capacity for induced growth during the larval period. The question next arises as to whether the progressive increase in gonad reactivity is dependent upon the liberation of gonado- trophic hormones during the course of nor- mal development. As many studies show (for background evidence and citations see Smith, °39), in sexually immature stages (from a few days after birth to onset of sexual maturity) of mammals (rat and mouse), ablation of the anterior lobe during the period when the gonad is shown to be responsive to administered gonadotrophin results in either arrested development or a retrogression of the gonad to a more im- mature state. Structural maturation of the gonad is arrested and functional activity ceases, as is reflected in the atrophic state of sex-hormone dependent structures such as the uterus, prostate, and seminal vesicle. Generally speaking, the extent of retrogressive change appears to be roughly proportional ONTOGENY OF ENDOCRINE CORRELATION to the state of reactivity of the gonad, being higher in older than in younger stages. Since the structural and functional changes that result from hypophysectomy may be pre- vented or restored to normal by the adminis- tration of gonadotrophins (in the form of either implants or extracts), it is clearly evident that the processes of maturation of the gonad are dependent upon the presence of specific hypophyseal secretions. Such se- cretions are clearly essential to the main- tenance of structural development and func- tion of the gonads, beginning early in post- natal life of the rat and mouse, and even prior to birth in the rabbit, as noted above (p. 595). The correlation of the time of onset and subsequent increase of gonadotrophic activity with histogenesis of cell types of the ante- rior pituitary is a difficult problem to re- solve. Only a few studies have been oriented toward its solution. The time course of differentiation of the chromophilic cells as related to the initiation of gonadotrophic activity has been worked out most completely for the anterior lobe of the pig fetus (Smith and Dortzbach, ’29; Nelson, ’33). The baso- philes differentiate first (50- to 60-mm. stage), then increase markedly in number, and be- come the predominate type between the 70- to 100-mm. stages. The acidophiles arise first at the 70- to 80-mm. stage, gradually in- crease in number during subsequent stages, and at the 160- to 170-mm. stage show a marked rise. This decided increase in number of acidophiles coincides approximately with the time (170- to 180-mm. stage) that the gonad-stimulating hormone is first indicated by functional tests. The activity, although slight at first, increases at the 200- to 221-mm. stage and still more at the 250-mm. stage. Since a significant increase in the acidophile population precedes by a short interval of time the initiation of gonadotrophic activity, it would appear that an age relationship exists between acidophile differentiation and the appearance of the gonad-stimulating principle. Although such a corresponding relation probably has significance in understanding the ontogeny of functional activity of the hypophysis as an integrated glandular organ, it has little value in assigning a specific physiological secretion to a particular cell type. The shift in proportional numbers of the three kinds of cells and their changes in cell structure under experimental condi- tions comprise a much more reliable indi- cator of the cell strain involved. By disturbing 59 experimentally the interfunctional relation of anterior pituitary and gonad (including sex structures dependent upon gonad secre- tions) during embryonic life, significant new clues may be obtained as to the possible relationship of the time order of differentia- tion and relative frequency of the cell types at a given stage with the onset and increase of gonadotrophic activity. Specially indicated is a study of the effects, either separately or in combination, of embryonic castration and the injection of sex hormones, each of which during postnatal (or equivalent) stages is known to alter the anterior pituitary both in structure and in gonadotrophic activity (either acceleration or inhibition). Also, in ascertaining the possible secretory signif- icance of the granules of the chromophiles histochemical methods may be useful. For instance, the periodic acid—Schiff reaction indicates that the granules of the basophiles (adult human pituitary) are mucoprotein in nature (Herlant, ’49) and therefore allied chemically to gonadotrophic hormones which contain a proportion of polysaccharides and glucosamine (Evans et al., ’39). Such studies as are proposed above might equally well be suitable for ascertaining the time in the course of ontogeny that gonad secretions have a reciprocal action on the anterior pituitary. Although the effects of castration, and especially of the injection of sex hormones, on the differentiation of sex structures of the embryo have been widely explored, little or no attention has been directed in such altered physiological conditions to the possible occurrence of changes in the anterior lobe. As functional tests indicate, both the gonad and anterior pituitary have hormonal activity beginning at least during the later stages of embryonic life and usually attaining gradually a higher value after birth or an equivalent time. So far as can be judged from the available evi- dence, the interlocking of functional activity is first clearly indicated at about the time the gonads are approaching sexual maturity (Engle, °31). At approximately that time the gonad secretions apparently stimulate the liberation of the gonadotrophic hormones. If correctly determined, does this signify that reciprocal functional relations are established when the gonad attains a level of secretory activity sufficient to cause the release of the previously accumulated gonadotrophins? Or is the relation unidirectional at first, during which time the gonadotrophins are released only in sufficient quantities to maintain gonad growth, maturation, and function? If 600 true, what regulates the gradual release? Answers to these questions await further in- vestigation. ANTERIOR PITUITARY AND GROWTH In addition to the production of specific trophic hormones which act selectively on the growth and function of certain “target” endocrine glands, the anterior pituitary pro- duces a growth hormone which plays an essential role in governing the general body growth of the organism. The concept had its origin from observations on human dis- orders, gigantism and acromegaly, in which overgrowth of the body or its parts is a characteristic feature. Since the overgrowth was usually associated with an adenomatous enlargement of the anterior lobe, the condi- tion was attributed to an overabundance of some growth-promoting principle. Conversely, it gradually became apparent that certain types of arrested growth, i.e., pituitary dwarfism, might be associated with an underactivity of the anterior lobe. Such pre- sumptive evidence initiated, over 40 years ago (1912), an era of experimentation which has continued up to the present time. Three high lights in the history of the concept may be recognized. (1) The ante- rior pituitary is essential for growth, since (a) complete hypophysectomy of an im- mature animal arrests growth and results in dwarfism, and (b) implants and crude ex- tracts can restore growth in hypophysecto- mized animals and augment growth in normal animals (first clearly established in the rat by Smith, ’30). (2) Experimental dwarfism is insufficient proof for the view that a specific hormone exclusively concerned with growth is secreted, since removal of the anterior lobe likewise causes growth regres- sion and diminished function of the gonad, thyroid, and adrenal cortex, secretions of the last two being clearly essential for nor- mal growth. Cessation of growth after hypo- physectomy thus appeared to be partly, at least, the result of multiple gland deficiency. As a consequence the principle came to be recognized that the hypophyseal hormones, through an interplay with other hormones and nutritional factors as well as genetic constitution, determine the ultimate size and form of the body and its component organs or parts. (3) Isolation of the growth hormone in chemically purified form by Li, Evans, and Simpson (745). In broad terms the objective here is, so far as the data permit, to trace during the course ONTOGENY OF ENDOCRINE CORRELATION of ontogeny the succession of events in the establishment of the functions of the ante- rior pituitary as related to the over-all growth of the body. Attention will be centered on such problems as (a) the time and degree to which the anterior pituitary is essential to the promotion of bodily growth, (b) the interplay of hypophyseal growth hormones and hormones of “target” glands in main- taining the metabolic processes of organismic growth, and (c) concomitant dependence up- on the genotypic constitution, nutritional factors, etc. of the organism. That the growth of the body in the anuran tadpole is dependent upon growth-promoting substances of the anterior pituitary is clearly indicated (see Table 27). Ablation of the ingrowing primordium of the anterior lobe from the early anuran embryo results in a diminished growth rate of the body as a whole. At first growth retardation is scarcely noticeable, but at about the mid-larval period (30 to 32 mm.) it is pronounced, at which time “an abrupt change in the direction of the growth curve ensues” (Smith, ’20, p. 46). This marked drop in velocity of growth, designated as the “critical point,” is followed by a later period of continuous but slow rate of growth. Moreover, the fat-body becomes unusually large, indicating that the utiliza- tion of fat is disturbed and contrasting sharply with the minute size of the fat-body of the normal tadpole at the time of com- pletion of metamorphosis. Growth in size of the hypophysectomized tadpole is restored essentially to normal in three ways: by feed- ing fresh anterior lobe (bovine) or dried residues thereof (Smith, *18, ’20), by trans- plantation of anterior lobe from adult frogs (Allen, ’28), and by intraperitoneal injection of extracts of bovine pars anterior (Smith and Smith, ’23). By the latter treatment the volume of the hypophysectomized tadpole in a period of three months may be increased to three times that of uninjected normal con- trols, and the growth of normal tadpoles may be accelerated by an increase of twice the volume of uninjected normal controls. The ultimate attainment of a size notably in excess of the normal in the treated hypo- physectomized tadpoles may be attributed in part to the persistence of the larval period and consequent extension of the larval growth span (Smith, ’20). Although the hypophysis is undoubtedly essential to the growth of the tadpole, it is by no means certain that a single growth- promoting principle is concerned. The situa- tion is complicated by the fact that cessation ONTOGENY OF ENDOCRINE CORRELATION of growth after hypophysectomy is associated with concomitant arrested growth and di- minished function of the adrenal cortex and thyroid glands. Likewise, restoration of growth to normal by hypophyseal adminis- tration is accompanied by concomitant res- toration of the growth and function of the adrenal cortex and thyroid. This thus poses the problem whether tadpole growth is the result of multiple gland activity. Does thyroid removal affect the growth of the anuran tadpole? If the anterior pituitary alone is essential it might be expected that the thyroidectomized tadpole would continue to grow although metamorphosis would not take place. In contrast to a distinct re- tardation of growth rate in the pituitaryless tadpole, the thyroidless tadpole, produced by removal of the thyroid primordium shortly after its formation, continues to grow, ul- timately attaining a body size much in excess of the normal control (Allen, ’18; Hoskins and Hoskins, *19). Concomitantly with the increased growth of the body, the vertebral skeleton continues to increase in size. Calci- fication proceeds extensively, but little or no ossification takes place. On the contrary, in the hind limb growth of the skeleton prac- tically ceases, and although calcification of the cartilage proceeds the process of ossifica- tion is much retarded (Terry, 18). More- over, growth in body size may be augmented by four times that of thyroidectomized con- trols by intraperitoneal injections of bovine anterior lobe extracts (Smith and Smith, 23). The degree of growth is sufficiently striking to suggest that the apparent growth- promoting action of the anterior pituitary is independent of thyroid activity. However, the extent to which the increase in body size may be attributed to excess growth stimuli from the hypertrophied anterior pituitary of the thyroidless tadpole or to a supplementary injection of pituitary prin- ciples, or to still other factors, is difficult to assess for two main reasons: (1) The rate of growth of thyroidless, pituitaryless, and normal frog larvae on a similar dietary regime may be nearly identical. (2) In thyro- hypophysectomized tadpoles growth of the body, although limited, nevertheless con- tinues, body and hind limb growth keeping pace with one another, indicating that fac- tors other than hypophyseal or thyroid hor- mones are concerned. Since, however, either thyroid preparations or iodine administered to tadpoles deprived of both glands or of either one alone promptly arrests growth with ensuing metamorphosis, it is clear that 601 the thyroid or its active principle has an antagonistic action to anterior pituitary. In summary, it appears that the anterior pituitary has a growth-promoting action, whereas the thyroid has an inhibitory action on the processes of growth of the anuran tadpole. This apparent contrasting action on tadpole growth is difficult to interpret, since as yet the evidence is neither adequate nor critical. Moreover, the elucidation of their action on growth is still further complicated by the commonly overlooked fact that hypo- physectomy affects not only the thyroid but also the adrenal cortex of the tadpole. It is now well established—for mammals, at least —that the hormones of the adrenal cortex participate in carbohydrate metabolism (in- fluence on the rate of gluconeogenesis from protein). The possibility of a linkage between the anterior pituitary, the thyroid, and the adrenal cortex in regulating growth and differentiation of the anuran tadpole must, therefore, be recognized. Such a functional interaction may in part be resolved by analyzing the effects of purified growth hormone both separately and in all possible combinations with thyroxine, thyrotrophin, ACTH, and adrenocortical hormones on the growth of hypophysectomized, thyrodecto- mized, and thyro-hypophysectomized tad- poles. Also, the extent to which dietary nutrients play a role in tadpole growth needs to be more precisely ascertained than here- tofore. The role of the anterior pituitary in regu- lating body growth has been most clearly demonstrated in mammals, especially in the actively growing postnatal rat, for which the evidence is most complete and critical. Complete removal of the anterior lobe from young rats during the period of active growth (usually 26 to 36 days after birth) invariably results in cessation of over-all body growth and leads to permanent dwarf- ism as a consequence. Save for slight fluctua- tions among individuals the size, weight, and length remain stationary. Skeletal growth is arrested. The growth stasis thus displayed in pituitaryless rats is readily broken at any time by daily homeo-implants of anterior lobe or injections of growth hormone preparations. Such treatment quickly causes a resumption of growth, which continues more or less normally. Undergrowth (cretinoid dwarfism) is like- wise produced by the removal of the thyroid. In rats dwarfed by thyroidectomy, growth is induced by the administration of growth hor- mone extracts. On the contrary, thyroid ad- 602 ministration to pituitaryless rats fails to stimulate growth. Although these results in- dicated that the anterior pituitary alone is essential for body growth, it was early shown by Smith (33) that the hormones of both glands have to some extent a synergistic action in promoting growth. This was brought out by the discovery that the simul- taneous administration of growth-hormone extracts and of thyroid (the latter in dosages insufficient to prevent increase in weight) to rats dwarfec by removal of both anterior pituitary and thyroid during active postnatal growth causes a significantly greater skeletal growth and increase in body length and weight than when growth hormone extracts alone are given. This original concept of synergistic action has been confirmed and broadened by the use of hormones in chemically purified form as they became available. Purified growth hormone restores growth in hypophysecto- mized rats (also stimulates overgrowth in normal animals), whereas thyroxine has only an insignificant effect on increase in body weight (Marx et al., 42). In combina- tion these hormones are more effective in promoting growth. Thyroxine augments the purified growth hormone in increasing body weight and in reactivating skeletal growth in rats dwarfed by hypophysectomy at 26 to 30 days after birth. The increase in body weight and length of animals, even after a postoperative interval of one year or longer, is significantly greater when both hormones in optimal proportions to each other are administered than when the growth hormone alone is given (Evans, Simpson, and Pen- charz, ’39; Becks et al., °46). Similarly, the capacity of the growth hormone to stimulate growth in pituitaryless rats is markedly en- hanced in combination with thyrotrophic hormone preparations. The thyrotrophic hormone seems to parallel thyroxine in its synergistic effect with growth hormone in promoting the growth of tissues. It seems probable that the thyrotrophic hormone ex- erts its effect indirectly by increasing the activity of the thyroid gland, but this point has apparently not been clarified. Beyond the apparent stimulation of the synthesis and retention of protein nothing is known concerning the mechanism of the synergistic action of these hormones. In contrast to thyroxine, partially purified ACTH has a counteracting effect on the ac- tion of the growth hormone. In female rats hypophysectomized at 26 to 28 days of age, ACTH in combination with the growth ONTOGENY OF ENDOCRINE CORRELATION hormone almost completely nullifies the growth-promoting action of the latter. There is little or no gain in body weight or increase in width of the epiphysis (tibia), as opposed to a marked increase in these features ob- tained with growth hormone alone. ACTH alone is ineffective in eliciting an increase in growth of the body or the epiphysis of the tibia. However, the adrenal glands exhibit a marked increase in weight under the in- fluence of ACTH alone or in combination with growth hormones (Evans et al., ’43). The fact that the adrenals are hypertrophied (stimulated by ACTH) furnishes presump- tive evidence that adrenal cortical hormones are involved, especially since somatic growth is apparently not inhibited by ACTH in rats deprived of the adrenals and is retarded in normal growing rats by the administration of adrenal cortical hormones. How can these observations be interpreted in terms of hormonal action on growth proc- esses? On the basis of considerable evidence, the most plausible interpretation is that both the growth and the adrenal cortical hormones alter metabolic processes but in opposite directions. The growth hormone seems to have the property of promoting pro- tein synthesis and retention, and as a con- sequence promotes bodily growth. The adre- nal cortical hormones, on the other hand, owing to their capacity to accelerate the catabolic phases of protein metabolism, have a retarding effect on rats. (For supporting evidence and theory see Long, ’42, ’43, ’49.) Thus, in rats at the stages of growth under consideration there exists an interlocking of function of the hormones of the anterior pituitary and at least of the thyroid and adrenal cortex in regulating the growth processes. A disturbance in the quantity or rate of hormonal output of any one of these endocrine glands seems to produce a_ hor- monal imbalance, the consequence of which is an arrest, retardation, or distortion of the growth processes. In the normal animal it seems probable that the quantity or rate of output of each of these hormones is regu- lated by a delicate balance in hormonal level in the blood stream, which in turn is adjusted to the growth potential of the or- ganism. Although this picture is relatively simple and general, it must be recognized that other hormones such as insulin are in- volved in the unusual complexity of factors that influence growth (see pp. 607-611). The question next arises as to whether the interlocking of hormonal function operating at 26 to 30 days in regulating growth is ONTOGENY OF ENDOCRINE CORRELATION likewise found at successively earlier ages, i.e., whether a gradual development of this linkage takes place with increase in age. In contrast to a somewhat abrupt cessation of increase in body weight and length and to a reduction in rate of skeletal growth in rats deprived of the anterior pituitary at 28 days, rats hypophysectomized at 6 days of age continue to gain in body weight and length at a rate of about 50 per cent of nor- mal for approximately three weeks, at which time increase in weight ceases. Increase in skeletal length and differentiation likewise continues at a similar rate (60 per cent) for the same period of time, whereupon it de- clines abruptly to a still lower rate (20 per cent), at which rate it continues for some time (Walker et al., 50). Similar effects are obtained by removal of the hypophysis from rats at intermediate ages (13 and 21 days). It would appear, therefore, that, irrespective of the age of removal, the pituitaryless rat continues to gain in body weight and length and to advance in skeletal growth and differ- entiation but only at a subnormal rate and for a limited period. Eventually these proc- esses are arrested. The time of arrest and the developmental age ultimately attained seem to vary in accord with the age of rat at hypophysectomy. In general, the younger the rat the greater is the amount of increase in body weight and length subsequent to hypophysectomy. At a glance these effects would seem to indicate (a) a less complete regulation of growth by the anterior pituitary in the rat at earlier than at later ages subsequent to birth and, as a corollary, (b) a rough index of the amount of intrinsic growth capacity that may be independent of the anterior pituitary. However, such interpretations are inadequate, since pituitary removal at early ages disturbs the function of receptor endo- crine glands, the secretions of which, as noted above, likewise play a role in regulating metabolic processes of growth. A functional linkage with the thyroid at early postnatal ages is indicated. Rats deprived of the thyroid on the day of birth exhibit for a long period (up to at least 111 days of age) an exceed- ingly slow rate of increase in body weight and length as well as in size of the skeleton. Skeletal differentiation continues at a greatly retarded rate. The administration of purified growth hormone to such thyroidless rats (be- ginning 30 days after thyroidectomy and continuing for 30 days) increases body weight (double that of untreated athyroid controls of equivalent ages) and length; skeletal di- 603 mensions increase, but skeletal differentiation is no more advanced than in untreated thy- roidless rats of equivalent ages. Other features such as the ears, pelage, and genitalia re- main immature. Thyroxine alone or in combination with growth hormone stimulates a marked increase in body weight (tripled over that of untreated athyroid rats of the same age) and in skeletal dimensions and differentiation. The rate of gain in body weight and skeletal dimensions appears to be higher with thyroxine alone than with growth hormone alone. Depending upon dosage, thyroxine alone may accelerate dif- ferentiation (skeleton) more than growth increment. In combination with growth hormone skeletal differentiation tends to keep pace with growth increment (Ray et al., SOE It is clear from the above considerations that the growth processes in the postnatal rat from the day of birth onward are greatly altered by ablation of either the anterior pituitary or the thyroid gland, indicating that both glands are involved as essential regulators. However, that normal growth after birth is dependent upon secretions of both glands is best brought out by treatment of the thyroidless rat with thyroxine and purified growth hormone, which in combina- tion have a synergistic action in promoting body growth. However, since the purified hormones were not administered until the athyroid rats had attained an age of 30 days (i.e., after thyroid removal on the day of birth), no indication is furnished as to whether prior to this age these hormones are equally effective in promoting or regulat- ing growth, i.e., whether from the day of birth onward an interlocking of function exists or whether a certain degree of develop- ment takes place before the rat is responsive to the hormones. The fact that a significant amount of growth takes place in the absence of either gland poses problems relative to the intrinsic growth potential (independent of hormones) and the degree to which the hormonal regulating mechanisms are es- tablished at a given age. The solution to these problems awaits further investigation. Whether ACTH, which as noted above has a counteracting effect on the action of the growth hormone on somatic growth of older rats, likewise has a similar action on normal rats or on those deprived of the anterior lobe at early postnatal ages is a problem as yet unresolved. The possibility of a linkage of the hormones of the anterior pituitary and the adrenal cortex in regulat- 604 ing growth processes shortly after birth must be recognized, since (1) even prior to birth the adrenal cortex becomes subnormal fol- lowing hypophysectomy (decapitation) and (2) in normal rats of 4 to 6 days post partum the adrenal cortex is functionally responsive to the administration of ACTH, as indicated by hypertrophy (Moon, ’40) and a decline in ascorbic acid content (Jailer, °50). How- ever, whether body growth can be modified in opposite directions by administration of growth hormone and ACTH or ACH to in- fant rats remains to be analyzed. This early period may be a critical one in which the interlocking of hormonal function is still undergoing progressive changes with age, i.e., not yet fully integrated. This suggestion is partly in keeping with the findings of Jailer, who noted that (1) the administration of adrenaline does not cause an increase in ascorbic acid in the adrenal cortex until the rat has attained an age of 8 days and (2) upon exposure to low temperature (5°C.) there is no decline in adrenal ascorbic acid until the sixteenth day after birth. It is of interest in this connection to note in the dwarf mouse that growth (increase in weight) is as rapid as in unaffected sibs up until the fourteenth day after birth (time of weaning), after which time growth practically ceases, dwarfism becoming manifest thereafter (see pp. 586, 590 and 605). As to the time of onset of the growth- promoting activity of the anterior pituitary in the course of ontogeny in mammals, in- formation of only a suggestive nature is available (see Table 27). The preceding analysis indicates that, as early as the sixth day after birth of the rat, the anterior pituitary seems to be essential for maintain- ing a normal growth rate. In a similar man- ner the thyroid is essential from the day of birth. The question naturally arises next as to the effects of removal of anterior pitui- tary or thyroid from the fetus. Complete removal of the pituitary by decapitation of a 19- to 22-day fetal rabbit (Jost, *51a) or by destruction by x-irradiation in a 13-day fetal mouse (Raynaud and Frilley, ’43, ’47) has yielded inadequate but nevertheless sug- gestive results with respect to effects on body growth. In the case of the rabbit fetus general body development (similar in trunk proportions, size, and fetal movements to normal litter mates) seems to proceed in a more or less normal manner in the absence of the head until nearly term (28 days). Sim- ilarly, body growth of a mouse fetus seems to take place in a normal fashion when the ONTOGENY OF ENDOCRINE CORRELATION thyroid is reduced to an epithelial nodule (without follicles or colloid) by x-irradiation of the buccal-pharyngeal region of a 13-day embryo. These findings may be tentatively interpreted as indicating that these glands are not yet effective as regulators of body growth during the late period in fetal life of a mammal. In contrast to the mammalian fetus, according to Fugo (’40) and Case (’52) the chick embryo when hypophysectomized (removal of all of the forebrain region) at 33 to 38 hours of incubation develops some- what normally in body proportions and shape, but the size of the body is consider- ably smaller than in normal controls of the same age (latter third of incubation). Whether such contrasting effects can be attributed to species differences or to a more precocious onset of the growth-promoting ac- tivity of the anterior pituitary in birds than in mammals remains problematical. The hypothesis that the anterior pituitary may be an ineffective regulator of body growth of the mammalian fetus during the later stages of development naturally raises a number of questions. During these stages is growth-promoting activity absent? Or is activity present without being released? If unreleased, when is it first detectable, and does it increase quantitatively with advance in developmental age? So far as is known to the writer, the only experiments designed to answer some of these questions are those of Smith and Dortzbach (’29). Using the hypo- physectomized rat as a test animal, these investigators found that daily implants of the anterior pituitary from pig fetuses from about 110 to 260 mm. elicit a definite body- weight increase and skeletal growth of the recipient. Moreover, the growth-promoting effect appears to be specific, since implants of other fetal tissues (muscle, brain, and blood) fail to stimulate growth. Anterior pituitary implants from smaller fetuses (70 to 90 mm.), even though the amount of tissue used may be equivalent to that which is effective from larger fetuses, do not pro- duce a significant change in weight and length. On the basis of the quantity of tissue required it appears that, after activity is first detectable, anterior pituitaries from suc- cessively older pig fetuses become progres- sively more effective in stimulating general body growth of a juvenile rat dwarfed by hypophysectomy. A gradual increase in growth-promoting activity of the pituitary with advance in fetal age is thus indicated. Is this initial appearance and increase in growth-promoting activity of the anterior ONTOGENY OF ENDOCRINE CORRELATION pituitary correlated in time with the cor- responding development of thyroid activity? In the fetal pig, insofar as the bio-assay methods are reliable sensitive indicators of activity, the thyroid hormone is first detect- able by the metamorphosis test in anuran tadpoles (see p. 583) in a fetus of 90 mm. in length, which is approximately the stage when growth-promoting activity of the an- terior lobe is first detectable by the rat test. Such a coincidence in the time of onset of hormonal activity may not indicate a causal connection, but merely the initial or pre- paratory phases in the development of a functional synergism between anterior pi- tuitary and thyroid, a functional relationship of established importance in regulating gen- eral body growth of young mammals after birth. Whether in the fetal pig the increase in growth-promoting activity of the pituitary is paralleled by increase in thyroid activity is not known. Such an increase might be expected, however, since in the bovine fetus thyroid iodine increases quantitatively with advance in fetal age. Since hypophysectomy of the mammalian fetus disturbs the development of the adrenal cortex, the gonad, and probably the thyroid, presumptive evidence is furnished for the view that specific trophic hormones con- cerned with these receptor glands are re- leased at least during the later stages of fetal development. The apparent absence of changes in over-all body growth of the fetus after hypophysectomy, however, suggests that specific growth-promoting activity is not essential for the maintenance of fetal body growth, a suggestion in keeping with the observation on dwarf mice, where body growth continues normally during fetal and early postnatal life. The validity of these suggestions may be tested by investigating the effects of chemically purified hormones either separately or in combination on the normal fetus as well as on fetuses from which one or more endocrine glands have been removed. But few attempts have been made _ to correlate the onset of growth-promoting ac- tivity with the temporal course of differentia- tion of the cell types of the anterior pitui- tary (see Table 27). In the pig fetus the stage (about 110 mm.) at which growth- promoting activity is first detectable by the rat test is preceded by a pronounced chro- matophilic differentiation in which the population of basophiles greatly exceeds that of the acidophiles. In successively older fetuses, however, the acidophile population 605 gradually rises, becoming marked at 160- to 170-mm. stages. This rise seems to parallel an apparent increase in growth-promoting activity of the anterior lobe as noted above. It is obviously difficult to assign to either type of cell a functional role in the elaboration of growth-promoting activity. It might be suspected, however, that the acidophiles are at least associated with the development of such activity, since these cells form adenom- atous tumors in such human disorders as acromegaly and related gigantism (Sever- inghaus, ’36) and also since in the dwarf mouse, according to Francis (44), (a) the number of typical acidophiles (accompanied by an increase of chromophobes) becomes markedly reduced at the time (twelfth day after birth) when the young dwarfs can be singled out from normal mice by length measurement and weight, and (b) the typ- ical acidophiles (only a few chromophobes remain) are completely absent in the adult dwarf. At corresponding ages in normal mice (including those heterozygous for the dwarf gene) the acidophiles and chromo- phobes comprise the predominating types of cell in the anterior lobe. The basophiles are apparently not involved, since this type of cell does not generally occur in the an- terior lobe of normal mice at any age except in those of advanced age or in pregnant females. The problem of relating the func- tional differentiation of cell types to the production of growth-promoting activity is particularly complex, since, in addition to the growth hormone, trophic hormones may likewise be produced, the latter acting di- rectly on such receptor glands as the thyroid and adrenal cortex, stimulating them to release secretions which play a role in regu- lating the growth of an organism. By way of general interpretation, it may be stated that during ontogeny there appears to be a gradual unfolding with time of anterior pituitary and receptor-gland_ hor- mones that act in harmony in regulating the ultimate size and form of the body and its component parts or organs. The functional interlocking of these hormones, although initiated in part during the later phases of embryonic or fetal life, does not appear to attain full expression until postembryonic stages of development. The period of elabora- tion of growth-regulating hormones may be regarded as a second phase in the develop- mental realization of the specific constitu- tional growth potential as provided by the genotype of the fertilized egg. During this phase, in contrast to an earlier one in which 606 the growth process is independent of cir- culating hormones, the growth-regulating hormones gradually take over via the sys- temic circulation the function of coordi- nating and regulating in final size and form the expression of the intrinsic growth pattern of the developing organism and its parts. To paraphrase a passage of D’Arcy Thompson (42, p. 264), hormonal regulation of growth processes is neither simple nor specific, but implies a far-reaching and complicated in- fluence on metabolism of already established growth patterns of the developing body and its parts. The actual growth expression is also dependent upon environmental factors, since nutrients, oxygen supply, temperature, etc. have an effect on the physicochemical back- ground in which the synthetic processes of growth take place. ISLETS OF LANGERHANS Concerning the problem of the develop- ment of functional activity of the endocrine portion of the pancreas, the islets of Langer- hans, little or no precise information is available. Nevertheless, an attempt will be made in this section to direct attention to and to analyze so far as the evidence permits such major problems as (a) the initiation of insulin production and (b) the unfolding of mechanisms of regulation of insulin secretion into the blood circulation. (See Table 27.) MORPHOGENESIS, CYTOGENESIS, AND INITIATION OF INSULIN PRODUCTION As a basis for a proper understanding of the functional development of the islets, it is first of all essential to examine the normal course of morphogenesis of the islet tissue and the cytogenesis of the cell types. The sequential steps in these processes have been especially well worked out for the fetal and early postnatal stages of the albino rat by Hard (’44). Generally speaking, the islets originate as cellular outgrowths from the epithelium of the dorsal and ventral pan- creatic lobes during their union and trans- formation into the branching system of ducts and terminal acini, which comprise the exocrine glandular portion of the pancreas. A very few islets are first identified as offsets from the solid dorsal and ventral lobes, respectively, on the thirteenth and fourteenth days of gestation. The majority of embryonic islets, however, are set off from the epithelium of the primitive pancreatic ONTOGENY OF ENDOCRINE CORRELATION tubules (formed by a rearrangement of cells of the coalescing pancreatic lobes), which are the progenitors of series of repeated units of exocrine ducts, ductules, and acini. Such islets, which are first apparent on the fif- teenth day, increase rapidly in number and in size with advance in fetal age, becoming most marked in an 18-day fetus. During the first week of postnatal life a very active formation of new islets occurs. The majority of these arise from the terminal portions of the ducts at the bases of the acini.* It is of interest to note that the potency for islet formation, although present during the en- tire course of morphogenesis of the exocrine gland, tends to shift progressively to the more terminal portions of the ducts as they arise. With respect to the sequence of differen- tiation of the islet cell types, the beta cell, generally accepted as the source of insulin in adult animals, is the only cell type to arise during fetal life of the rat. The alpha cell does not appear until the second day after birth. The so-called delta cell (inter- preted by some as a stem cell) has not been identified during the developmental stages under consideration. The beta cell is first identified in the * The apparent sequential development of two main groups of islets in the rat seems to be in gen- eral agreement with the older findings on several mammalian species (for review, see Bargmann, 39). According to the older investigators the first group of islets, the so-called islets of Laguesse, de- generate and disappear, while the second group is retained as the definitive islets of Langerhans. The validity of this view may be questioned on the basis of the difficulty in distinguishing the islets as to their exact source of origin and in following their subsequent fate. Moreover, in the rat, although there is a sequence in generation of islet tissue, the set of embryonic islets persists into the postnatal period and apparently differs in no essential partic- ular from the second group except as to position of origin. Similarly, in the development of the chick a sequence in the generation of islets has been de- scribed by Potvin and Aron (’27), the transitory islets appearing at 8 days of incubation, the defini- tive islets two days later. A more recent investiga- tion by Villamil (’42), however, indicates that the pattern of islet formation in the chick is actually more complex, in that two types of islets—the so- called “light” and “dark” islets—are simultane- ously present on the eighth day. These islets exhibit a differential behavior from the beginning. The dark islets give rise to alpha and delta cells begin- ning, respectively, on the eighth and fourteenth days, whereas the light islets begin on the twelfth day either to degenerate or to give rise to cells with beta granules. The ultimate fates of these two kinds of islets need to be worked out. ONTOGENY OF ENDOCRINE CORRELATION 1814-day fetus within a few hours after the vascular capillary network has developed within the larger islets, a relationship of probable functional significance. Only a few beta cells with beginning granule formation occur at this stage. During the succeeding four days two significant changes take place. (1) The number of beta cells in process of differentiation rapidly increases, reaching a peak in 20-day-old fetuses. (2) A concom- itant increase in the accumulation of beta granules occurs on the side of the cell toward the capillary, becoming well marked in many cells on the twenty-first day, an orien- tation suggestive of an actively secreting beta cell. The number of fully mature beta cells packed with secretory granules reaches a peak on the twenty-second day (birth). On the day of birth the prospective alpha cells may be identified as cords or groups of nongranular cells at the periphery of the islet, i.e., about a core of beta cells. On the second day after birth a few of these outer- most cells begin to form secretion granules in the area of cytoplasm adjacent to a cap- illary. The number of such cells gradually increases, so that by the fifth day the ma- jority of them may be identified as alpha cells, a few of them having attained a fully mature state through a gradual increase in number and size of the secretion granules. The cytological picture is such as to indicate actively secreting alpha cells by the fifth day of postnatal life. The question next arises as to whether the appearance of beta cells with secretory gran- ules coincides in time with initial imsulin production and secretion into the blood of the fetus. As an index of beginning func- tional activity it is significant to note that in the rat fetus near or at term (1) the islets with differentiating beta cells are highly vascularized, (2) the secretory granules of the beta cells are oriented toward a capillary, and (3) the number of fully mature beta cells packed with secretory granules reaches a peak. The total picture is such as to indi- cate insulin secretory activity. However, information concerning the first appearance of insulin in the pancreas and the quantitive changes in content during embryogenesis is meager, owing chiefly to the lack of suitable microchemical or cyto- chemical methods for its detection. Although the earliest time of appearance has not been determined, insulin is already present in the pancreas of a 5-month fetal calf in propor- tionally greater quantities than in pancreases of older fetuses, of young calves, and even 607 of the adult (Banting and Best, ’22; Fischer and Scott, ’34). (See Table 27.) A progres- sive decrease in quantity of insulin per gram of pancreatic tissue seems to take place with advance in ontogenetic age of the calf. Such a decline with age may possibly represent a change in balance between rates of produc- tion and liberation of insulin, which in turn may be related to the rate of utilization of the hormone in the tissues of the growing organ- ism. It would be of considerable significance in this connection to determine changes not only in the cytological picture of the islets but also in the ratio of islet tissue to acinar tissue with increasing age and body weight (cf. Hess and Root, ’38). Certain indirect evidence presented by Carlson et al. (11, 714) and by Aron et al. (23) and Aron (724) indicates that insulin in both dog and cat fetuses is produced and is liberated during the latter half of the gestation period. If pregnant dogs or cats are pancreatectomized early in the gestation period, the mother develops hyperglycemia resulting in death; if similar operations are performed later in the gestation period, i.e., after the seventh to ninth week of gestation, no signs of hyperglycemia develop until after parturition. Fetal insulin is, therefore, seem- ingly protecting the mother after pancre- atectomy. Whether this can be attributed to (1) passage of fetal insulin into the maternal circulation, (2) oxidation or utilization of excess blood sugar by the fetus, or (3) com- pensatory adjustment of hormonal mecha- nisms concerned in regulating carbohydrate metabolism in the mother and placenta re- mains unsettled. (For a comprehensive re- view of the subject of carbohydrate and other types of metabolism in the placenta and fetus see Huggett, ’41.) UNFOLDING OF MECHANISMS OF REGULATION OF INSULIN SECRETION Although in the postnatal mammal a num- ber of regulating mechanisms are probably involved either directly or indirectly, the principal regulator of insulin secretion ap- pears to be the blood sugar level itself. Moreover, the amount of sugar in the blood is governed in the main by an interlocking relationship between liver function (glyco- genesis and glycogenolysis) and the rate of secretion of insulin by the islets. It would appear possible, therefore, that a clue as to the period of onset of insulin regulation of sugar in the blood stream might be obtained 608 by an examination of the developmental changes in blood sugar and liver glycogen with the time that insulin secretory activity of the beta cells is cytologically indicated. Although the causal relations between the accumulation of carbohydrates in the blood and liver and the histogenesis of the cell types of the islets have not been explored on a scale commensurate with the importance of this problem, some suggestive information is available for examination. In the normal chick embryo, Konigsberg (’54) finds by quantitative microchemical methods that the blood sugar rises from 87.1+2.8 mg. per cent on the eighth day to a level of 112+5.5 per cent on the tenth day, remaining some- what constant at this level until about the fourteenth day. Following this plateau, the blood sugar again rises to an average level of 151.9+6.5 mg. per cent on the sixteenth day. These findings are in general agreement with those of Leibson and Leibson (743) with respect to the periods of increase and the plateau in blood sugar level, save for the additional point that the blood sugar level apparently reaches a still higher peak after the eighteenth day. Similar parallel changes likewise take place in the amount of liver glycogen. Initially present as traces on the sixth day, the amount of liver glyco- gen tends to increase gradually until the ninth day, then declines, becoming most marked on the twelfth day (Dalton, ’37)— according to Konigsberg changing from 3.90 +0.62 per cent (dry weight) on the tenth day to 2.32+0.35 per cent on the twelfth day, and to a value of 7.17+0.63 per cent on the fourteenth day. Beginning on the thirteenth day the glycogen content of the liver rises rapidly to its highest value on the nineteenth day and then appears to drop precipitously at about the time of hatching (Dalton, ’37; Leibson, *50). It is significant to note that immediately prior to the time of onset of the second rise in carbohydrate accumulation in the blood and liver a few beta cells are first identified on the twelfth day according to Villamil (42). Whether subsequently the number of beta cells in- creases concomitantly with the rapid rise in carbohydrates has not been worked out. In the albino rat fetus at relatively late stages (ca. 16 days to term), according to Corey (32), the blood sugar increases quan- titatively with advance in fetal age, more or less gradually at first, followed by a period of very rapid increase (20+ day to term.) This period of rapid increase coincides in time not only with the initiation of beta cell ONTOGENY OF ENDOCRINE CORRELATION differentiation and a most rapid increase in their number, but also with a prominent development of the capillary network within the larger islets. A similar pattern of quan- titative increase in liver glycogen is exhib- ited in fetuses between the 16+ day and term (Corey, ’35). By way of generalization, both blood sugar and liver glycogen are definitely present in measurable quantities at a time before beta cells can be distinguished cytologically by secretory granules and in the case of blood sugar (probably present from the beginning of vascular circulation) even before the islets are formed. It is clear, therefore, that the early accumulation of carbohydrate in the blood and liver is independent of insulin secretory activity of the islets. At later stages in development, at about the time that the rate of increase in blood sugar and liver glycogen somewhat abruptly become mark- edly accelerated, beta cell differentiation is initiated. However, whether the striking cor- respondence in time between these events is a true index of onset of insulin regulation remains problematical. On the problem of the time in the course of development that hormones of the anterior pituitary and other endocrine glands _ be- come interrelated with the functional ac- tivity of the islets, the evidence is so scanty as scarcely to permit consideration. Never- theless, on the basis of considerable evidence bearing on such interrelations in adult mam- mals, it seems worth while to outline at least briefly some of the specific kinds of problems that challenge the embryologist. The current view that certain anterior pituitary hormones act directly and/or in- directly in exerting physiological effects which are opposite to that of insulin poses the problem as to when such opposing effects arise in the course of development. In pitui- taryless chick embryos (produced by removal of the forebrain region at 33 to 38 hours) profound changes from the normal occur in (1) accumulation of carbohydrates in the blood and liver, (2) the adrenal cortex, and (3) the thyroid. The amount of blood sugar is increased over normal values (see Fig. 213) from the eighth to the thirteenth day, with a drop on the fourteenth day; the amount of liver glycogen increases from the tenth to the fourteenth day (Konigsberg, ’54).* Furthermore, pituitary removal leads to sub- normal activity of the thyroid and adrenal * Cf. Jost (51a) for evidence that the quantity of liver glycogen is markedly reduced after pituitary removal (decapitation) in the rabbit fetus. ONTOGENY OF ENDOCRINE CORRELATION cortex, as is reflected respectively in the arrest of the synthesis and storage of thyro- globulin and in the accumulation of adrenal cortical lipids and ascorbic acid. Since these glandular deficiencies are first apparent dur- ing the period of initiation of active histo- genesis of the cell types in the normal anterior pituitary (10 to 13 days) it may be inferred that the anterior pituitary is func- tionally active in the production and release of thyrotrophic and _adrenocorticotrophic hormones. It is perhaps noteworthy that in the pitui- taryless chick embryos the period of onset of the apparent subnormal hormonal actiy- ity of the adrenal cortex and thyroid coin- cides more or less closely with that of the most marked increase in concentration of blood sugar and liver glycogen. Such a correspondence in time seems to suggest that the abnormal accumulation of carbohydrate is somehow brought about by a hormonal imbalance. Whether, however, such imbal- ance can be attributed to (1) the absence of a possible direct-acting or an indirect- acting pituitary principle, (2) a diminished hormonal output of the adrenal cortex and/ or thyroid, or (3) their combined absence remains equivocal. Another problem is presented by the ob- servation that almost immediately following this period of correlation of apparent hor- * In this study of the mechanisms involved in the production of hypoglycemia in the embryo by in- sulin treatment, Zwilling made the significant dis- covery that the glycogen content of the yolk-sac membrane is greatly increased by the action of in- sulin. This observation, along with others, led to the interpretation that insulin produces its hypogly- cemic effect in the embryo by the inhibition of gly- cogenolysis in the yolk-sac membrane, which as a consequence fails to release glucose in normal quan- tities into the vitelline veins. These experimental studies substantially confirm the concept of Claude Bernard that the yolk-sac membrane (“les parois du sac vitellin’’) of birds functions as a temporary liver (“foie transitoire’) until the embryonic liver is morphologically prepared to assume it [for a his- torical treatment and discussion see Needham (31)]. Three related questions remain to be an- swered by future research, namely: (1) At what period in development of the chick is the glycogenic function shifted from the yolk-sac membrane to the liver? (2) Is the shift correlated with onset of hor- monal activity? (3) Is there a special period in liver development when the enzymes, such as phos- phorylase and phosphoglucomutase, essential for the formation of glycogen from glucose become active? [According to O’Connor (53) the initial appearance of glycogen in the liver is associated with changes in enzymatic activity. | 609 monal imbalance and high levels of the carbohydrates, i.e., between the twelfth and fourteenth days, the blood sugar level de- clines, whereas liver glycogen continues to increase. Does this hypoglycemic effect in- dicate the onset of some disturbance in the mechanism of conversion of liver glycogen into glucose? If so, is the effect connected with beta cell differentiation which is seem- ingly in progress at this time? Although the specific problems raised by these questions have not been analyzed, the following ob- servations on normal and hormone-treated embryos appear pertinent. (1) The time that the higher rate of increase of blood sugar and liver glycogen begins in the normal chick embryo agrees fairly closely with onset of the hypoglycemic effect in the pituitary- less embryo. (2) The injection of insulin (5 units) into the yolk sac of normal chick embryos on the fifth day produces during succeeding days a diminution in all carbo- hydrate fractions in the embryo proper; how- ever, by the twelfth day, the amount of free sugar and glycogen returns to normal levels (Zwilling, °51)* (3) The injection of cortin (0.02 cc. of adrenal cortical extract) into the yolk sac on the fifth day produces a subsequent rise in blood sugar, which reaches a peak at 10 to 12 days, returning abruptly to normal values by the fourteenth day (Zwilling, 48). These three lines of evidence furnish additional support for the hypothesis that the period between the twelfth and fourteenth days is one of a transitory nature in the development of mechanisms for regulating the quantity of blood sugar by hormonal activity. Whether functional activity of the islets plays a role at this time remains problematical. A final problem for brief mention here is posed by the current theory that adrenaline has an accelerating effect, whereas insulin has an inhibiting effect, on sugar formation in the liver. Through such an opposing action these hormones play an important contribu- tory role in the homeostatic regulation of blood sugar. Although the time in ontogeny that such a counter action sets in is un- known, it is of interest to note that adren- aline is present in the adrenal medulla very early in development, and at stages before any sign of functional activity of the islets is manifested. Furthermore, with advance in developmental age adrenaline increases quantitatively, reaching an unusually high concentration in late fetal life (see pp. 590— 593). In view of these observations it would be of considerable interest, as a first step in 610 the analysis of the problem, to ascertain at what stage in ontogeny (during either the later prenatal or the early postnatal stages) a temporary decrease in liver glycogen and an increase in blood sugar can be produced by adrenaline injection and whether such changes can be prevented by insulin injection. ORGAN OR TISSUE ACIDOPHILES +t ANTERIOR PITUITARY BASOPHILES IMMATURE FOLLICLES + COLLOID (N) THYROID FOLLICLES + COLLOID (-P) ASCORBIC ACID (N) ASCORBIC ACID (-P) ADRENAL CORTEX LIPIDS (N) va) [WO | -— ~ + + LIPIDS (-P) e _ OVARIAN CORTEX (WN) TESTIS (lett) OVARIAN CORTEX (estradiol) OviDUCTS (N) vipucTSs 9 oS OVIDUCTS (estradiol) (9) OVIDUCTS (low temperature ) ISLETS BETA CELL (N) CARBONIC ANHYDRASE (WN) FIBRINOGEN + PROTHROMBIN (N) BLOOD GAMMA GLOBULIN TRACE GLYCOGEN (N) LIVER GLYCOGEN (-P) I NOEPENDENT FUNCTIONAL DEVELOPMENT a @ o eee — Eee ——————— ; +++ + n LITTLE OR NO APPARENT INCREASE = ACTH RESTORES TO NORMAL ONSET OF CHEMICAL INTEGRATION STATE OF CHEMICAL COORDINATION ONTOGENY OF ENDOCRINE CORRELATION of enzymes involved in glucose metabolism (for a comprehensive treatment of the subject see Cori, 46, and Young, °48). To sum up this account, it is at once apparent that the problem of unravelling the sequence of events leading to the establish- ment of hormonal mechanisms which con- esate were 201 COMMENTS AUTHORITY PAYNE, ‘46 RAHN, ‘39. WILSON, ‘52 18 20 m.t<: | ee DEFINITIVE AFTER HATCHING INCREASE IN NUMBER, SIZE. aa AMOUNT OF COLLOID RELATIVELY FEW, SMALL SIZE; LITTLE OR NO COLLOID FUGO, ‘40 HOPKINS, ‘35 MARTINDALE, ‘41 VALUES IN pg/mg VALUES IN pg /mg TENDS TO REMAIN CONSTANT AT i2- DAY LEVEL 14 16 CASE, ‘52 ++ + 110 + COMPLETE DISAPPEARANCE BY iith DAY INJECTION PRIOR TO 8th DAY NO EFFECT AFTER Iith DAY COMPLETE DISAPPEARANCE BY iith DAY INJECTION PRIOR TO 8th DAY NO EFFECT AFTER tith DAY EXPOSURE ON 8th DAY — 70% PERSISTING OVIDUCTS VILLAMIL, ‘42 INCREASES RAPIDLY UNTIL CLARK, ‘SI 18th DAY (adult level) VANGOOR, ‘40 ALWAYS PRESENT BY I4th DAY RICKER NORAND GLADSTONE, ‘25. VALUES IN mg.% KONIGSBERG, 54 INCREASING TO PEAK AFTER . CAT oHING ZWILLING, ‘48 VALUES IN mg % ‘ HYPERGLYCEMIA —*HYPOGLYCEMIA KONIGSBERG, 54 VALUES IN % DRY WT KONIGSBERG, 94 RISES TO HIGHEST VALUE ON 9th LEIBSON, ‘50 DAY: DROPS AT HATCHING. ZWILLING, ‘SI 1 VALUES IN % DRY WT KONIGSBERG, 54 INCREASING FUNCTIONAL INTERDEPENDENCE WILLIER ef af, ‘37. ‘39 WILLIER, Fig. 213. Chart depicting manifold correlations in development of humoral activity in the chick embryo (after Willier, 54). In any comprehensive attack on the over- all problem consideration should also be given to the time in the course of development when the many enzymes concerned with carbohydrate metabolism become active, es- pecially in the liver and skeletal muscle, and also to whether the activity of such enzymes, once established, can be influenced by hor- mones. A pertinent clue may be taken from the work of Cori and collaborators (see Colowick et al., 47), who in an investigation of the effects of hormones on enzymatic re- action in vitro showed that insulin, and hor- mones of the anterior pituitary (A.P.E.) and adrenal cortex, have an influence on the activity of hexokinase, the first of a chain tribute to the homeostatic regulation of blood sugar is unusually complex. As may be seen from the accompanying chart (Fig. 213), three main periods or phases in the time scale of the development of humoral activity in the chick embryo may be recognized (Willier, 54). The phase of independent functional ac- tivity, extending roughly from the sixth to the tenth day, is characterized by a display of initial functional activity and/or reactiv- ity by a variety of organs and tissues, to wit, glycogen begins to accumulate in the liver on the sixth day, gradually increasing on subsequent days; blood sugar, although pres- ent earlier, increases in concentration; and ONTOGENY OF ENDOCRINE CORRELATION the thyroid barely begins to store thyroglob- ulin in follicles on the ninth day. Seem- ingly, during this phase the processes concerned with the accumulation of carbo- hydrate in the liver and blood are to a high degree, if not entirely, independent of hormonal activity. These processes, although of obscure nature, are in some manner partly linked with the glycogenic and glycogeno- lytic functions of the yolk-sac membrane (in this organ glycogen attains maximum concentration at 8 days, whence it diminishes to a lower level by the twelfth day according to Zwilling, *51). The period from about the tenth through the thirteenth day is characterized by a concatenation of events of a widely diverse nature. Such an impressive array of events as are listed in the chart (many are not listed) clearly suggests the onset of chemical integration. Apparently the glands and or- gans have become correlated functionally with each other in a chainlike manner through hormones and other humoral sub- stances transported in the blood vascular circulation. Since significant changes in car- bohydrate accumulation take place during this period in normal and in experimentally treated embryos, it may be inferred that the hormonal mechanisms which contribute to the homeostatic regulation of blood sugar are beginning to unfold. From approximately the fourteenth day to the time of hatching is a period of appar- ent maturation of functional activities pre- viously set in operation. A strengthening and extension of functional interlocking of the glands via the hormones and other humoral substances is postulated on the basis of the progressive increases in specific glandular tissue, in quantity of glandular secretions, and in quantities of sugar and certain other constituents of the blood. In conclusion, the over-all problem of the development of the functional activity of the islets and the mechanisms of regulation of insulin secretion into the blood is only beginning to take shape. In perspective the foundations have been laid for a compre- hensive analytical attack on the problem which should be pressed at an increasing tempo. PARATHYROID GLANDS Although the functional role of the para- thyroids of the adult mammal has been reasonably well worked out (Greep, ’48), relatively little attention has been directed 611 to the problem of the development of their functional activity. In the past practically all studies have been directed toward an elucidation of the site of origin and morpho- genesis of the parathyroids. Since these glands arise typically in four different loci, are extremely small, and are intimately associated with the thymus and thyroid, studies on the effects of their experimental removal in the embryo are fraught with un- usual difficulty. As an indispensable background for under- standing the functional development of the parathyroids, the main sequential steps in their morphogenesis, histogenesis, and growth will be examined. These processes have been most carefully studied in the developing human embryo by Norris (737). It is now generally agreed that these glands, commonly four in number, arise early in development (8 mm. embryo of man) as four localized proliferations of entodermal cells of the dorsal-lateral wall of the third and fourth pharyngeal pouches. Each local- ized proliferation takes on the form of a solid bud, which gradually becomes detached from the pouch, whereupon it assumes a globular shape, only later in fetal develop- ment attaining its definitive, more or less elliptical, form. Although arising from sep- arate pouches, there are no essential differ- ences in morphogenesis, in histogenesis, or in rate of growth of the four parathyroid primordia. The chief or principal cell (or variants thereof) appears to be the only cell type to arise during embryonic life and, moreover, is the only type of cell found in postnatal life in most animals; the so-called oxyphile cell (possibly representing a transitional or regenerative phase of the chief cell) is ‘an inconstant component appearing in the hu- man gland only after childhood’ (Greep, 48). Singularly, there is no clear cytological evidence of secretory activity in the chief cells of the adult glands, even though the physiological evidence for endocrine secre- tion is irrefutable. During the course of morphogenesis, histo- genesis, and growth of the parathyroids of man three main points stand out as possible indices of functional activity: (1) clear chief cells, the essential elements of the glands in postnatal life, are present in the early pri- mordia of the parathyroids (9-mm. embryo) and at all subsequent stages in their develop- ment during fetal life; (2) the growth in- crement of the glands (mainly by increase in number of chief cells) per unit length 612 of the embryonic body is relatively slow up to the 80-mm. stage, from which time onward to the mid-fetal period (ca. 160 mm.) it becomes much greater; and (3) the period from 100 to 150 mm. is one in which the most active development of sinusoidal cir- culation takes place. Each gland is trans- formed from nearly a solid ball of parenchym- atous cells into a body of anastomosing cords with intervening sinusoids. At the 160 mm. stage, almost immediately after sinu- soidal circulation is well established, a marked increase in the number of paren- chymatous cells and concomitant increase in the size of the glands occur. Thereupon cell types possibly representing different phases in the secretory cycle of the chief cell first appear. It becomes apparent from the lines of evidence just given that, although the essen- tial cellular elements are present from very early stages onward in the morphogenesis of the parathyroids, it is not until the mid- fetal period that the glands present a struc- tural pattern indicative of endocrine activity. Whether such a pattern, however, marks the initial time of active production and release of the parathyroid hormone remains uncertain of decision. A more reliable index to the onset of functional activity is the relation of struc- tural pattern of the parathyroids and the mobilization of calcium in the blood of the fetus. According to current theory, which is based upon a considerable body of experi- mental evidence on postnatal mammals, the principal regulator of the secretion of the parathyroids is the calcium level in the blood. Moreover, a homeostatic relationship appears to exist between the levels of calcium (and phosphate) in the blood and the se- cretory activity of the glands. From the standpoint of theory only a few studies have a bearing on the problem of the relationship of calcium concentration to parathyroid ac- tivity in the fetus. In both human (Nicholas et al., °34) and dog (Hoskins and Snyder, *33) the fetal serum calcium level is 1 to 2 mg. per cent higher than that of the mother, suggesting that the fetal glands may be functioning independently of the mother in regulating calcium level. In keeping with this view are the observations on the rat (Sinclair, 42) that the fetal glands at term, although constituting no more than 5 per cent of the total weight of the adult glands, are twice as large in proportion to fetal body weight as are those of the adult. Moreover, the activity of the fetal glands as judged ONTOGENY OF ENDOCRINE CORRELATION by weight changes is suppressed by high and stimulated by low maternal serum calcium levels (high serum phosphorus likewise has a stimulating effect). Although the fetus is obviously solely de- pendent upon the mother for its source of calcium and phosphorus, the question arises as to whether the parathyroid hormone can cross the placenta in either direction. That the hormone may not pass the placenta in either direction under normal conditions is indicated in the dog by injecting parathyroid extract (Collip) separately into the mother and fetus (Hoskins and Snyder, ’33). When injected into the fetus the serum calcium is raised in fetal but not in maternal blood; when injected into the mother the serum calcium is raised in the maternal but only to a limited extent in the fetal blood (tends to parallel the elevation of serum calcium produced in the mother). In the rat the fetal parathyroids, although greatly hypertrophied (weight may be doubled over that of normal controls) in parathyroidectomized mothers, have no ameliorative effect, since such mothers develop tetany 2 to 4 days before term (Bodansky and Duff, *41a; Sinclair, *42). On the basis of these results it may be inferred that in the normal animal the fetal hormone (assuming its presence in the cir- culation) is not transmitted across the pla- centa to the maternal circulation. However, whether the maternal hormone is trans- mitted to the fetus remains uncertain. As a tentative hypothesis it is suggested that the transmission of calcium itself rather than of the maternal hormone can better account for the changes in the fetus which result from disturbances in calcium concentration in the maternal blood. To a high degree the fetus appears to be autonomous in regulating the calcium (and phosphate) levels in its blood, tending to maintain uniform levels even under adverse conditions in the mother. For example, the fetus can mobilize calcium through the pla- centa and apparently maintain normal cal- cium levels even when only very low concentrations of calcium are available in the maternal circulation. If, however, the maternal calcium concentration falls below a certain critical level, the fetal serum calcium level becomes subnormal and there- by calls forth compensatory activity, as is reflected in the hypertrophy of the fetal para- thyroids (Bodansky and Duff, ’41b; Sinclair, 42). As in the adult, regulation in the fetus (at the latter stages of fetal life) seemingly involves the integrated cooperation of a va- ONTOGENY OF ENDOCRINE CORRELATION riety of conditions, chief of which are the function of the parathyroids, the availability of mobilizable calcium, and calcium utili- zation. It might be expected on_ theoretical grounds that the maintenance of calcium and phosphate in the fetal blood in concentrations above those in the mother is in some way related to active deposition of those calcium salts characteristic of developing bone. In the albino rat active bone formation is in progress during the latter third of the ges- tation period (Strong, ’25). Ossification is initiated in a limited number of skeletal elements (first in the clavicle) on the 17+ day. During subsequent days the number of such elements with beginning ossification rapidly increases, reaching a peak near term (21 days), when most of the bones of the fetus exhibit ossification centers. Obviously calcium and phosphorus are being utilized in increasing amounts as the number of ossifying bones increases. Although the con- centrations of calcium and phosphorus in the blood are known to be high at term, the amounts present at earlier stages have not yet been determined. Whether the concentra- tion of these substances in the blood increases concomitantly with the progressive increases in the calcification of cartilage and of periosteal bone would be of interest. Although calcium and phosphate occur in fetal blood in concentrations above those actually essential to the process of ossifica- tion, the rate of their utilization is appar- ently determined by the developing skeletal elements. Two main lines of evidence may be cited to show that the individual elements of the vertebrate skeleton differ among them- selves in morphogenetic pattern, which among other activities sets the specific site and amount of ossification in a given element. (1) In the rat fetus the time of appearance of bone salts and the specific locus of their deposition varies from one skeletal element to another (Bloom and Bloom, 740). (2) Prospective (‘undifferentiated’) osteogenic tissue of skeletal elements, such as the jaw (Meckel’s cartilage and membrane bone), femur, and palatoquadrate bar, when iso- lated from the chick embryo (5%- to 6-day) and grown separately as explants in vitro shows a remarkable capacity for independent development of shape and histology. Differ- ences in physiological properties are likewise expressed, since phosphatase activity is found in those elements that in normal develop- ment ossify (e.g., femur and palatoquadrate) but not in those that fail to ossify (e.g., 613 Meckel’s cartilage). Thus, in the absence of both blood and nerve supply each element develops in its own characteristic fashion almost exactly as it does in its appropriate position in the body of the embryo (Fell, Bi); From these considerations it is apparent that the pattern of the skeleton is already mapped out in the early embryo long before the onset of ossification or even chondrifica- tion. Each element has acquired its own intrinsic growth pattern and special physi- ological properties for selecting and utilizing calcium and phosphate from the common blood pool of the fetus. The quantity of up- take would appear to vary from one indi- vidual element to another in accordance with the amount of osteogenic tissue (calcifiable tissue) formed or capable of being formed. Furthermore, as the number of skeletal ele- ments with osteogenic tissue increases, the quantity of uptake would be expected to in- crease progressively until the osteogenic po- tentialities have reached a peak in their expression. Although the pattern of osteogenic potency is set early in the development of a given skeletal element, its full expression in normal bone growth is apparently dependent upon a multiplicity of interconnected conditions, chief of which are (1) availability of ade- quate concentrations of calcium and _ phos- phate in the blood; (2) presence of the para- thyroid hormone, which plays an important role in the mobilization and metabolism of calcium; (3) the presence of vitamin D, which apparently facilitates the utilization of calcium and phosphate in ossification; (4) internal secretions of the pituitary (growth hormone), thyroid, and sex glands, which, in influencing the growth process in general, tend as a rule to have a nonspecific regula- tory effect on the growth of bone; and (5) mechanical factors (i.e., tension of muscle and ligaments, etc.) that influence the final surface modelling of the bones during the later stages of their development. (For an excellent and comprehensive treatment of the subject see Clark, °52.) In conclusion, it will be apparent from the foregoing condensed account that the develop- ment of parathyroid activity presents prob- lems of an intricate and complex nature. The onset of secretory activity cannot be considered separately from other physiolog- ical activities, since all activities are co- ordinated in maintaining steady states within the fetal body. Future progress in elucidating the physiological role of the fetal parathy- 614 roids is dependent mainly upon advances in the understanding of (1) the nature of the properties of the developing skeletal elements that enable them selectively to remove and utilize calcium and phosphate from the blood stream, (2) the mechanisms of deposition of calcium salts in bone, and (3) the time when the parathyroid hormone is produced and released into the fetal blood. A promising approach to the problem of the time of onset of parathyroid activity is that of testing the glands at progressively older developmental stages for capacity to bring about bone re- sorption when grafted beneath the periosteum of bone in young postnatal host animals. The value of this proposal is obvious from the studies of Chang (’51), who showed for young mice and rats that parathyroid tissue when grafted beneath the periosteum of the parietal bone causes a marked local resorp- tion of bone, indicating the emanation of parathyroid hormone from the graft. The effect is highly specific, since other tissues so grafted cause little or no bone resorption. REFERENCES Adams, A. E. 1946 Variations in the potency of thyrotropic hormone of the pituitary in animals. Quart. Rev. Biol., 27:1-32. Adler, L. 1914 Metamorphosestudien an Ba- trachierlarven. I. Exstirpation endokriner Drii- sen. A. Exstirpation der Hypophyse. Roux’ Arch. Entw.-mech., 39:21-45. Allen, B. M. 1918 The results of thyroid removal in the larvae of Rana pipiens. J. Exp. Zool., 24: 499-519. 1927 Influence of the hypophysis upon the thyroid gland in amphibian larvae. Univ. Calif. Pub. Zool., 37:53-78. 1928 The influence of different parts of the hypophysis upon size growth of Rana tad- poles. Physiol. Zool., 7:153-171. 1929 The influence of the thyroid gland and hypophysis upon growth and development of amphibian larvae. Quart. Rev. Biol., 4:325-352. 1938 The endocrine control of amphib- ian metamorphosis. Biol. Rey., 73:1-19. Anderson, D., Billingham, R. E., Lampkin, G. H., and Medawar, P. B. 1951 The use of skin grafting to distinguish between monozygotic and dizygotic twins in cattle. Heredity, 5:379-397. Aron, M. 1924 Le fonctionnement du pancréas et la régulation glycémique chez l’embryon des mammiferes. Indications fournies par leur étude au point de vue du fonctionnement du pancréas et de la régulation glycémique chez ladulte. Arch. Internat. Physiol., 22:273-298. , Stulz.) Es, and) Simon, R. 1923.) Hone-= tionnement du pancréas foetal aprés ablation du pancréas maternel. Compt. Rend. Soc. Biol., 89: 571-573. ONTOGENY OF ENDOCRINE CORRELATION Banting, F. G., and Best, C.H. 1922 Pancreatic extracts. J. Lab. & Clin. Med., 7:464 472. Bargmann, W. 1939 Die Langerhansschen In- seln des Pankreas. Handbuch der Microskop- ischen Anatomie des Menschen, 6 (Part 2): 197- 288. Springer, Berlin. Bascom, K. F. 1923 The interstitial cells of the gonads of cattle, with especial reference to their embryonic development and significance. Am. J. Anat., 31:223-259. Becks, H. M., Simpson, M. E., Evans, H. M., Ray, R. D., Li, C. H., and Asling, C. W. 1946 Re- sponse to pituitary growth hormone and thyroxin of the tibias of hypophysectomized rats after post- operative intervals. Anat. Rec., 94:631-655. Bloom, W., and Bloom, Margaret A. 1940 Calci- fication and ossification. Calcification of develop- ing bones in embryonic and newborn rats. Anat. Rec., 78:497-523. Bodansky, M., and Duff, V. B. 1941a Effects of parathyroid deficiency and calcium and _ phos- phorus of the diet on pregnant rats. J. Nutrition, 21:179-192. , and Duff, V. B. 1941b Dependence of fetal growth and storage of calcium and phos- phorus of the diet in pregnant rats. J. Nutrition, 22:25-41. Burns, R. K., Jr. 1934 The transplantation of the adult hypophysis into young salamander larvae. Anat. Rec., 58:415-429. 1949 Hormones and the differentiation of sex; in Survey of Biological Progress, Vol. 1, pp. 233-266. Academic Press, New York. , and Buyse, A. 1931 The effects of ex- tracts of the mammalian hypophysis upon im- mature salamanders. Anat. Rec., 57:155-185. , and Buyse, A. 1933 The induction of precocious maturity in the reproductive tract of recently metamorphosed female salamanders by an extract of the mammalian hypophysis. Anat. Rec. 58:37-53. , and Buyse, A. 1934 The effect of an ex- tract of the mammalian hypophysis upon the re- productive system of immature male salamanders after metamorphosis. J. Exp. Zool., 67:115- 135. Carlson, A. J., and Drennan, F. M. 1911 The control of pancreatic diabetes in pregnancy by passage of the internal secretion of the pancreas of the fetus to the blood of the mother. Am. J. Physiol., 28:391-395. , Orr, J. S., and Jones, W. S. 1914 The absence of sugar in the urine after pancreatec- tomy in pregnant bitches near term. J. Biol. Chem., 17:19-22. Case, J. F. 1952 Adrenal cortical-anterior pitui- tary relationships during embryonic life. Ann. N. Y. Acad. Sci., 55:147-158. Chang, H. 1951 Grafts of parathyroid and other tissues to bone. Anat. Rec., 71717:23-47. Clark, A. M. 1951 Carbonic anhydrase activity during embryonic development. J. Exp. Biol., 28: 332-343. Clark, Helen M. 1935 A prepubertal reversal of the sex difference in the gonadotropic hormone ONTOGENY OF ENDOCRINE CORRELATION content of the pituitary gland of the rat. Anat. Rec., 67:175-192. Clark, W. E. LeGros 1952 The Tissues of the Body: An Introduction to the Study of Anatomy, 3d ed. Oxford, at the Clarendon Press. Colowick, S. P., Cori, G. T., and Slein, M. W. 1947 Effect of adrenal cortex and anterior pituitary extracts and insulin on the hexokinase reaction. J. Biol. Chem., 168:583-596. Corey, E. L. 1932 Placental permeability to in- sulin in the albino rat. Physiol. Zool., 5:36-48. 1935 Growth and glycogen content of the fetal liver and placenta. Am. J. Physiol., 112: 263-267. Cori, C. F. 1946 Enzymatic reactions in carbo- hydrate metabolism. Harvey Lectures, 47;253- Dif. Courrier, R. 1951 Contribution a l’endocrinolo- gie de la thyroide. Acta endocrinologica, 7:54-59. Dalton, A. J. 1937 The functional differentia- tion of the hepatic cells of the chick embryo. Anat. Rec., 68:393-409. Dawson, A. B. 1953 Histochemical evidence of early differentiation of the suprarenal gland of the chick. J. Morph., 92:579-595. Deanesley, R. 1938 Adrenal cortex differences in male and female mice. Nature, 147:79. Domm, L. V. 1937 Observations concerning an- terior pituitary-gonad interrelations in the fowl. Cold Spring Harbor, Symp. Quant. Biol., 5:241- RW Donaldson, J. C. 1919 The relative volumes of the cortex and medulla of the adrenal gland in the albino rat. Am. J. Anat., 25:291-298. Edmunds, H. W. 1950 Pituitary, adrenal and thyroid in cyclopia. Arch. Path., 50:727-735. Elliot, T. R., and Armour, R.G. 1911 The devel- opment of the cortex in the human suprarenal gland and its condition in hemicephaly. J. Path. & Bact., 75:481-488. Engle, E. T. 1931 The pituitary-gonadal rela- tionship and problem of precocious sexual matur- ity. Endocrinology, 15:405—420. 1939 Gonadotropic substances of blood, urine and other body fluids; in Sex and Internal Secretions, edited by Allen, Danforth, and Doisy, chapter XVIII, pp. 1003-1044. Williams & Wil- kins Co., Baltimore. Etkin, W. 1935 Effect of multiple pituitary pri- mordia in the tadpole. Proc. Soc. Exp. Biol. & Med., 32:1653-1655. 1938 The development of thyrotropic function in pituitary grafts in the tadpole. J. Exp. Zool., 77:347-377. 1939 . Lh Ly! NE ahd er. I MOLT AND METAMORPHOS/S HEMIMETABOLOUS DEVELOPMENT 4 TH. INSTAR ———>** B\s METAMORPHOSIS dence showing that the thyroid is activated at the time of metamorphosis. Unfortunately, the histological and cytological study of the pituitary in relation to metamorphosis does not present unequivocal evidence of any change in activity at this period (Clements, °32; Schliefer, °35). In fact, the considerable variation in the degree of cytological differ- EGGS © * Gk Gr tas Oe SS eee ON ‘6 peas ot eNO UE, i ah Sea 5 TH. INSTAR he e —, q 4 HOLOMETABOLOUS DEVELOPMENT Fig. 220. The life history of a hemimetabolous insect, the grasshopper Melanoplus differentialis, is shown at the left. Holometabolous development in the giant silkworm moth, Platysamia cecropia, is shown at the right. (From C. D. Turner, General Endocrinology.) METAMORPHOSIS entiation among different species of anurans as reported by these workers suggests that little confidence can be placed in the cyto- logical correlations that have been adduced in specific instances. Unpublished studies by the writer of the growth and histology of the anterior pituitary during the develop- ment of the tadpole in R. pipiens likewise failed to reveal any change in growth rate or histological picture indicative of a marked change in activity related to metamorphosis. It may be concluded that, although the activation of the thyroid at the beginning of metamorphosis is clearly dependent upon a thyrotrophic hormone from the pituitary, it cannot be stated whether the initiation of thyroid activity results from an increase in thyrotrophic hormone production or from other factors. CONTROLLING FACTORS IN INSECT METAMORPHOSIS The metamorphosis of hemimetabolous in- sects is commonly described as taking place in stepwise fashion, a small step with each larval* molt. In holometabolous forms, on the other hand, transformation from larva to adult is said to take place in one stage, the pupal stage, at the end of larval develop- ment (Fig. 220). However, it must be real- ized that in all insects there is a process of remodelling taking place at all stages. In hemimetabolous insects noticeable progress toward the adult condition is made with each molt. Nevertheless the last molt, some- times called the metamorphic molt, usually presents a considerably greater amount of change than any of the earlier ones. In that respect this last molt is similar to the pupal stage of holometabola. Also, although the holometabolous animal does not seem to change toward the adult condition in any obvious way during larval molts, yet the imaginal discs may undergo considerable microscopic change. It is to be expected, therefore, that the basic mechanisms of meta- morphosis would be similar in both types of insects. The earlier experimental work on insect metamorphosis gave only small promise of showing a common basic pattern throughout insects, since many contradictory conclusions were drawn from work on different species. *The term “larva” will be here used for the young of hemimetabolous insects (nymphs) as well as for those of holometabola in order to emphasize the equivalence, as revealed by experimental analy- sis, of the early instars in both types of insects. 637 However, in recent years largely through the comprehensive analyses of Wigglesworth on the bug Rhodnius (Hemiptera, hemimetab- ola) and Williams on the _ silkworm Platysamia (Lepidoptera, holometabola) a clarification and simplification of our under- standing of the metamorphic process in in- sects has taken place. It might be best for the sake of clarity to present in outline the interpretation of these experiments and then to proceed to consider the extent to which the work on other forms can be fitted into ~~ ———— Med. NS.C ae a A bb . ; Fig. 221. Relations of head endocrine glands in an insect, generalized type. The brain contains two groups of neurosecretory cells on each side (Med. N.S.C. and Lat. N.S.C.). These are connected by nerves to the corpora cardiaca (C.C.) and corpora allata (C.A.). (After Cazal, ’48.) the pattern seen in Rhodnius and Platy- samia. Before doing so, however, it may be well to describe briefly the basic morphology of the organs with which we shall be concerned. Each side of the brain of the insect commonly possesses two groups of large neurons, a medial and a lateral group (Fig. 221). These neurons contain stainable droplets which ap- pear to be secretions. Presumably these se- cretions are released as hormones, either directly into the blood stream or first being conducted down the axons as in other neuro- secretory cells (Cazal, 48; Scharrer, 52a; Thomsen, ’52). Lying more posteriorly in the head, on either side of the anterior end of the aorta, are two pairs of glands. The more anterior pair of these are the corpora cardiaca (C. C.) and behind them the corpora allata (C.A.) (Fig. 221). These are joined to the brain and to each other by nerves. In some 638 insects the C.A. of the two sides are fused to form an unpaired median gland. In the thorax of the insect strands of large cells are found. These are commonly designated as “prothoracic glands.” In the larval and early adult dipteran the arrangement of some of these structures is different (Fig. 222). The cells equivalent iM Fig. 222. Head endocrine glands in a larval dipteran, Eristalis tenax (after Cazal, ’48). to the C.A. are located dorsal to the aorta, those corresponding to the C.C. are ventral and joined to the C.A. region by lateral lobes presumably equivalent to the protho- racic glands (Possompés, ’50a). The entire complex forms the so-called ring gland of Weismann. RHODNIUS Rhodnius is a blood-sucking bug which molts at a definite interval (varying from 15 to 28 days depending on the stage) after a single meal of blood. Wigglesworth (’34— 51) in a series of investigations found evi- dence that the molt is brought on by a hormonal stimulus from the brain. For ex- ample, he found that decapitation before a certain stage (the critical period) after the meal prevents the subsequent molt, but decapitation later does not. Brains from post—critical-period animals induce molt in non-activated abdomens. When active ani- mals or parts thereof are joined in parabiosis to non-activated tissues the stimulus to molt- ing is transmitted to the latter, even when the parts are joined by capillary tubing (Fig. 223). By transplantation the active region of the brain was found to be the part of the protocerebrum containing the large neuro- secretory cells. The brain hormone in Rhodnius, acting alone, induces a molt in which larval organs undergo complete transformation to the adult METAMORPHOSIS condition. However, when the corpora allata are active they change the effect of the brain hormone, with the result that the larval tissues lay down structures appropriate for the next nymphal stage, not for the adult condition. During all larval instars except the last in the normal development of the ani- mal both brain and C.A. are active. The molt which ensues is therefore a larval molt from which the animal emerges as the next larval instar. At the last larval stage the C.A. fails to be active in secreting its hormone. There- fore under the unmodified influence of the brain hormone the larval structures undergo complete transformation to the adult con- dition. The evidence upon which this interpreta- tion of C.A. function is based is varied and complex. Much of it derives from the re- markable way in which whole insects or body regions from different insects can be joined together in parabiosis (Fig. 223). By joining insect parts, activated and inacti- vated, with and without brain and corpora allata, the effectiveness of the stimulus in each such part can be tested. In addition to this technique, the brain and corpora allata can be dissected out of larvae and adults and implanted into larvae of various stages. Also, by cutting the head immediately be- hind the brain, preparations can be secured which have C.A. intact but no brain. The evidence derived from these varied proced- ures shows that an insect body-part molting under the influence of fifth (last) instar body fluids assumes the adult condition ir- respective of its original instar level. Im- planted C.A. from younger instars than the fifth change the molt so that nymphal struc- tures are produced. Conversely, fifth instar animals metamorphosing under the influence of stimuli derived parabiotically from earlier instars do not form adults but instead form giant sixth instar nymphs. Curiously, even adults which, of course, do not normally molt at all, can be made to molt when in para- biosis with active nymphal instars. When the metamorphic stimulus comes from an early instar the molt may lead to at least partial reversion to nymphal characters. Wigglesworth (48) found that the fifth in- star larva was more, rather than less, effec- tive in inducing complete transformation in its parabiont when its C.A. was intact than when it was removed. This rather unex- pected result was interpreted as showing that not only does the C.A. not secrete any of its hormone in the fifth instar, but it absorbs or otherwise counteracts any such hormone METAMORPHOSIS persisting in the tissues from previous stages. Recently Wigglesworth (751), following the lead of Williams and other workers on lepidopterans (see below), found that the brain hormone does not act directly upon the tissues but rather activates the protho- racic glands to the production of their hor- mone. His evidence consists of the fact that transplantation of the strands of gland cells "> HORMONE ~ AFTER CRITICAL PERIOD A CAPILLARY TUBE AFTER CRITICAL PER/OD 8 CAPILLARY TUBE 639 the diapause when implanted into a non- activated pupa (Fig. 224). Removal or im- plantation of the corpus allatum, on the other hand, has no effect, showing that activity of this gland plays no role in the adult transformation. That the brain hormone acts by stimulating the prothoracic glands and not directly on the tissues was shown (Wil- liams, ’47) by the fact that implantation of an FIRST NYMPH BECOMES DWARFED ADULT BEFORE CRITICAL PERIOD (O—F/FTH NYMPH BECOMES 6 TH-STAGE CSUPERNUMERARY) NYMPH BEFORE CRITICAL PER/OD Fig. 223. The union of nymphal instars of Rhodnius by capillary tubes. A, Fifth nymph, decapitated after the critical period and thus with active glands, is coupled with a first nymph, decapitated before the critical period. Both individuals undergo imaginal molts, the first nymph becoming a diminutive adult. The fifth nymph of Rhodnius, serving as the donor in this instance, furnishes prothoracic hormone with little or no C.A. hormone. B, Second-stage nymph, decapitated after the critical period, is combined with a fifth nymph that had been decapitated before the critical period. The second nymph, acting as the donor, delivers both C.A. and prothoracic hormones into the body of the fifth nymph. The fifth nymph molts and, under the influence of both hormones, becomes a sixth-stage (supernumerary) nymph. (From C. D. Turner, Gen- eral Endocrinology.) from the thorax of activated individuals is sufficient in itself to induce molt in non- activated abdomens. Thus the picture of the mechanism of metamorphosis in Rhodnius may be diagrammed as in Figure 225. PLATYSAMIA The work of Williams on Platysamia began with an analysis of the mechanism of diapause in this lepidopteran. The animal normally spends the winter as a pupa, and experiment has shown that the freshly col- lected pupa will not develop into an adult unless it is first exposed to a period of at least two weeks of cold. Williams (’46) was able to show that the influence of cold was exerted through activation of hormone pro- duction in the animal’s brain. An activated animal induces imaginal development in a non-activated parabiont. The brain from a cold-treated animal is effective in breaking activated brain into a diapausing abdomen is not effective unless a prothoracic gland is implanted along with it. At first Williams (49) was of the opinion that the mechanism of post-diapause development involved a dif- ferent brain hormone from that found by others (see below) to be concerned with lar- val and pupal molts. Later, however, Wil- liams (752) was able to show by ligation experiments and transplants of the protho- racic organs and brains from various stages that the same hormonal mechanism is in- volved in the pupal and larval molt as in the breaking of diapause. Since the role of the C.A. in larval molts was established by others in Lepidoptera (see below), the mech- anism of hormonal control of development in Platysamia may be diagrammed as in Figure 225. The correspondence of the mech- anism controlling metamorphosis in Rhod- nius and Platysamia is seen to be quite complete. In this interpretation diapause 640 METAMORPHOSIS Fig. 224. Parabiotic union in moth pupae. A,B, Brainless pupa of Platysamia cecropia grafted to a chilled pupa of the same species. The brain of the chilled parabiont provides hormone necessary for the development of both animals. C,D, Brainless diapausing pupa of Telea polyphemus grafted to a chilled pupa of P. cecropia. The successful development of both pupae shows that the brain factor is not species specific. This fact also shown by E, which is adult of P. cecropia developed from a brainless pupa which had received the brain from a chilled pupa of T. polyphemus. (From C. D. Turner, General Endo- crinology.) results from a cessation of brain activity in the pupa and the consequent necessity of reactivation of the brain by exposure to cold. INSECTS IN GENERAL We may now consider whether the endo- crine mechanism discussed above for Rhod- nius and Platysamia is applicable to insects in general. Before doing so, however, a note on terminology is necessary. As experiment- ers have uncovered various factors in insect metamorphosis, names descriptive of their actions have usually been given to these factors. Such names are generally rather unsatisfactory, often being vague, too gen- eral, too restricted or sometimes even imply- ing a mode of action later found incorrect. When the source of a factor has been traced to a particular organ it is generally more satisfactory to name it in terms of that organ, and such terminology is favored in this discussion. The term “hormone” will be used for these factors since in general they seem to be hormone-like in their activity. How- ever, as will be seen below, the exact cor- respondence to a hormone as understood in vertebrate physiology is open to some ques- tion. The following terms, their definitions and synonomy are important. Brain Hormone. An agent produced by the protocerebrum, presumably by the neuro- secretory cells contained therein. It acts by METAMORPHOSIS stimulating the prothoracic glands to secrete their hormone. The brain hormone corre- sponds to the factor variously designated as “molting hormone,” “metamorphosis hor- mone,” “pupation hormone,” and “growth and differentiation hormone.” Corpus Allatum Hormone. An agent pro- duced by the corpus allatum. When acting with prothoracic hormone it modifies the tissue response in such a way as to lead to a larval molt. This factor was formerly desig- RHODNIUS Brain Bpotieeate Gland Larval CiAe Molts ~sTses Larval Molt bas Prothoracic Gland Pupal CAs. \ . Developments ‘Tissues Imaginal Molt PLATYSAMIA Prothoracic Gland CA. 1 oS Tissues Larval Prothoracic Gland Pupal Development AANAUSE ProtherecieGland Ci ¢ Imaginal Development 641 three aspects of insect development: (1) the molting process as it occurs in the larva, (2) pupation or the metamorphic molt, and (3) the differentiation of adult structures from imaginal discs or other larval sources. Re- sults on the Diptera will be considered separately after the general discussion, since these results at first did not seem to fit in with results in the other insects. For reviews of various aspects of the problem published during the last ten years see Needham (42), DIPTERA 5 ere Prothoracic Region GVA: Region Brain i Tissues Molt Brain Larval Molt Tissues cold Ge Brain Pupal & Imaginal Devel. Tissues y Fig. 225. Schemata of endocrine relations in insect metamorphosis according to the Wigglesworth- Williams concept. nated as “inhibitory hormone” and “juvenile hormone.” Prothoracic Gland Hormone. An agent pro- duced by certain glandular cells of the in- sect’s thorax (in some extending also into the head region). It seems to act directly upon the tissues, stimulating them to assume the adult condition except when acting in con- junction with C.A. factor as noted above. The prothoracic factor was and still is often designated as the “growth and differen- tiation hormone,’ a term appropriate in indicating the wide action of the hormone but objectionable in its extreme vagueness and its previous use for the brain hormone. The prothoracic gland is presumed to corre- spond to the glandular structures of different insects variously called “intersegmental gland,” “ventral gland” and “peritracheal gland.” It will be profitable to consider separately Bodenstein (42), Joly (45), Wigglesworth (45), Mendes (’47), Scharrer (’52a), Turner (48), and Williams (’49). Larval Molting. The clearest evidence that blood-borne factors control the molting proc- ess in pre-metamorphic stages comes from implantation experiments. Limbs or pieces of skin of caterpillars of various species trans- planted from one larva to another molt with the host, not at the same time as the donor animal. This is true even if transplant is made from one species to another, or from pupa or imago to larva (Bodenstein, °33; Furukawa, ’35; Piepho, ’39). Since such results follow when the implant is in the fat body or haemocoele and not in contact with the host hypodermis the molting stimu- lus would appear to be transmitted by the blood. Direct attempts by injections to demon- strate the presence of the molting factor in 642 the blood have not, in the opinion of the writer, proved conclusive. Buddenbrock (731) reported an acceleration of ecdysis in various caterpillars after the injection of blood from molting larvae as compared to controls in- jected with blood from non-molting indi- viduals. His data, however, are not convinc- ing and further doubt is thrown on the conclusion by the careful work of Schirfeld (35), who found control procedures such as mere sticking with a needle as effective as the injection of blood. Pupation, or the Metamorphic Molt. Much evidence has accumulated to show that these metamorphic molts are, like larval molts, hormonally controlled. The first experi- mental results in this field, however, were complicated and obscured by the secondary effects upon pupation of interference with nutritution and oxygen supply. Kopec (22) reported that the removal of brain before a definite “critical” period pre- vents metamorphosis in Lymantria (Lep- idoptera). Initially he interpreted his results as indicative of the activity of a metamor- phosis hormone derived from the brain (this appears to have been the first time evidence for hormonal control of metamorphosis in insects was adduced). The subsequent study of Kopec (’24) led him to conclude that the mechanism of brain action lay in the in- hibition of feeding which follows its removal, rather than in hormone production. Starvation had the same inhibitory effect on metamor- phosis as did brain removal, and the critical period for starvation and brain removal coincided. Bounhiol (’38), in experiments on yarious lepidopterans, obtained extensive confirmation of this viewpoint. The conclu- sion of these authors that brain removal before critical period inhibits feeding, and that it is the consequent starvation which prevents metamorphosis, is acceptable. How- ever, this interpretation does not preclude the possibility that the brain also produces a hormone necessary for metamorphosis. The discovery of the importance of the oxygen supply, and tracheal growth for pu- pation, is another factor which complicates the interpretation of experiments on isolated insect parts. Bodenstein (739) found that, whereas the abdomen of Drosophila could not pupate when isolated from the thoracic region before a certain critical period, it could do so if kept in circulating oxygen. This he interpreted as showing that the “thoracic” factor in pupation to which his previous experiments on isolated abdomens had pointed was the downgrowth of tracheal METAMORPHOSIS connections. This interpretation was sup- ported by histological studies on the source and time of formation of the tracheal supply of the abdomen. The necessity of an adequate respiratory mechanism for the development of insect structures had also been indicated in the work of Fraenkel (735) on Calliphora (Diptera) and Fukuda (739, ’40a,b) on the silkworm. The uncertainties that arise in the inter- pretation of experiments on the ablation and isolation of parts because of the interference by the non-specific factors of nutrition and respiration make it difficult to interpret much of the early evidence purporting to show the existence of specific hormonal factors in pu- pation. Fraenkel (’?35), working on the fly Calliphora, found that a posterior part of a grub separated from the anterior portion by a ligature would pupate only if the ligature were made within 16 hours of presumptive pupation time. The anterior part could al- ways pupate. Blood from pupating larvae in- duces pupation in non-pupating posterior halves. That this blood-borne factor is not itself a sufficient stimulus for pupation is shown by the fact that it is effective in the abdomen only if it is ligatured within 24 hours of normal pupation. It therefore seems possible that the action of the injected blood is nutritive rather than being hormonal as Fraenkel inferred. Caspari and Plagge (735), working on two lepidopterans, obtained evi- dence that removal of the brain prevents pupation. Brain implants seemed to restore the pupation capacity though the evidence presented on this point is not adequate. Kiihn and Piepho (’38) on Ephestia (Lepidoptera) and Bodenstein (’38a) on Drosophila (Dip- tera) by ligature experiments showed the importance of some factor in the anterior end of the larva for pupation. Piepho (’38) showed that pieces of larval epidermis im- planted into the fat body of host larvae pu- pate synchronously with the host. Plagge (38) found brain removal before a critical period in the caterpillar prevented pupation. Reimplantation of the brain or injection of blood from post-—critical-period larvae seemed effective in inducing pupation, though the evidence here is not clear-cut. Similar results on the hymenopteran Trypoxylon were re- ported by Schmieder (’42). In silkworms, Fukuda (’40a,b) found that decapitation pre- vented pupation when done early enough. Later, i.e., after the critical period for decapi- tation, the prothorax is still necessary for pupation. At this time implants of protho- racic glands are capable of bringing on pupa- METAMORPHOSIS tion in isolated abdomens. This report pre- sented the first evidence on the importance of the prothoracic glands. A number of work- ers have found correlations between molting activity and cytological evidences of secre- tion in the prothoracic glands (Kaiser, ’49; Rahm, °52). In contrast to the mass of evi- dence linking the prothoracic glands with pupation factor is the finding of Piepho (48) that these glands can be removed at any time in the last instar of the moth Galleria with- out preventing pupation. The nature of the hormonal factors in molting and metamorphosis could not be clarified until the role of the corpora allata had been worked out. Following the lead of Wigglesworth a number of investigators were successful in demonstrating that the C.A. played much the same role in many other insects as in Rhodnius. Bounhiol (737, ’38) extirpated the corpus allatum (leaving the brain intact) from the caterpillars of a num- ber of lepidopterans and showed that preco- cious pupation ensued. Conversely he found that grafts of the corpus allatum inhibit pu- pation and permit supernumerary molts. Bounhiol concluded that the corpus allatum produces a hormone inhibiting pupation but permitting molting. Pflugfelder (37) like- wise found premature metamorphosis result- ing from corpus allatum removal in the stick insect (Dixippus morosus—Orthoptera ). Pfeif- fer (’39, 45a) reported that transplants of corpus allatum from young to older larvae in the grasshopper suppressed metamorpho- sis, thereby permitting extra molts and the growth of nymphs to twice normal length. Evidence for the existence of an inhibitory factor in the corpus allatum was also found by Fukuda (’40a) in Bombyx (Lepidoptera), by Piepho (’40, 50) in Galleria (Lepidoptera) and by Scharrer (’46) in Leucophaea (Or- thoptera). In her cytological study of the corpora allata of the grasshopper Melanoplus, Mendes (48) found evidence of secretory activity throughout the intermolt period in all stages except the last. In this stage secre- tion apparently stopped in the middle of the intermolt. Comparable results on other in- sects were obtained by Pflugfelder (38) and Rahm (’52), corresponding to those expected from the experimental evidence. The inhibi- tory action (in the sense of Wigglesworth, °34) of the corpus allatum on insect meta- morphosis may thus be said to be well estab- lished for a variety of insects. Differentiation of Imaginal Primordia. A number of early reports had indicated the dependence of imaginal discs upon some 643 humoral factors for their development. In Vanessa (Lepidoptera) Bodenstein (735) found young imaginal discs capable of re- sponding to a differentiating factor when transplanted to pupae. Fukuda (739) reported the accelerated development of larval ovaries of the silkworm when transplanted into pupae. This was independent of the sex of the host and without effect on it. These re- sults indicate that some condition, pre- sumably the presence of hormonal factors in the tissue fluids of pupae, stimulates the dif- ferentiation of the primordia of the adult structures. In the moth Phryganida, Bodenstein (’38b) reported that abdominal ectoderm is unable to differentiate without a factor from the thorax which is normally produced in pupae 1 to 2 days old. However, he thought that this effect was not transmitted through the blood, since it did not pass through a glass tube connecting two insect regions. As men- tioned above his later work (’39) on Drosoph- ila led him to conclude that the thorax factor was tracheal development. However, in view of the work of Fukuda and Williams discussed below this interpretation does not appear to apply to Lepidoptera. In the silk- worm Fukuda (’41b) showed that the pro- thoracic glands are necessary for the differ- entiation of the imaginal discs of the pupae. It is apparent from the above discussion of non-dipterous insects that, except for the report of Piepho (48) on the waxmoth, all the data permit interpretation in terms of the Wigglesworth-Williams concept, namely, that the brain stimulates the prothoracic glands to secrete a hormone which induces metamorphic transformation in the tissues. At larval molts the activity of the C.A. so changes the response of the tissues that a larval molt rather than metamorphosis en- sues. The study of metamorphosis in Diptera has centered around the functioning of the ring gland of Weismann. In the larvae of Drosophila this gland has been identified as the source of a molting hormone by Boden- stein (’44), who found that a larval head deprived of its nervous system and ring gland would not molt when transplanted into the abdomen of the adult fly. But if two lar- val ring glands were included with the trans- planted head, one or two molts followed. In these, as in Wigglesworth’s experiments with Rhodnius, the molting stimulus is found to be non-specific for the species. A pupation factor was localized in the anterior part of the body of Calliphora by the work of 644 Fraenkel (735). Burtt (738) identified the active region as the ring gland by showing that the removal of this gland or cutting the nerves to it prevents pupation. In Drosophila there is a giant strain which fails to pupate at normal size. Implantation of ring glands (but not of various other tis- sues) from normal larvae into individuals of this strain leads to pupation at normal size (Hadorn, °37). Precocious pupation in the normal larvae is attained by ring gland im- plantation (Hadorn and Neel, ’38). Vogt (42) found that ring gland implants induce the next normal molt in Drosophila tissues, that is, third (last larval) instar abdomens are induced to pupate by ring gland im- plants whereas second instar eye-antenna primordia merely molt to the third instar stage when implanted into these same ab- domens along with the ring gland. This in- terpretation that ring glands induce the next normal molt agrees with the results of Boden- stein (’43, °44), who found that imaginal discs from older larvae do not grow when implanted into adult abdomens unless ring glands are implanted along with them. It is therefore possible to suppose that a hormone or hormones produced by the ring gland reg- ulate molting, pupation and imaginal disc development in Drosophila, the effect in each case depending on the degree of development which has been attained by the reacting system. The possibility that this effect of the ring gland is hormonal is supported by the work of Frew (28), who found that imaginal discs of mature blowfly larvae in- vaginate when cultured in tissue fluids of pupae but not of larvae. It was difficult at first to reconcile the work on the ring gland of flies with the re- sults on other insects, since the ring gland by itself seemed capable of such varied func- tions and apparently was not dependent upon stimulation from the brain. But Possompés (’50a,b), working with Calliphora, has re- cently shown that the lateral lobes of the gland, which histologically seem comparable to the prothoracic glands, are the sources of the hormonal stimulant to the tissues and that these lobes are dependent for their acti- vation upon stimulation by the brain, prob- ably by the neurosecretory cells in that or- gan. Like the prothoracic glands in other insects, these lateral lobes degenerate after metamorphosis (Vogt, ’42). As mentioned above, the dorsal cells of the ring gland ap- pear to be homologous to the C.A. of other orders of insects (Cazal, ’48). The function- ing of this region was also brought into line METAMORPHOSIS with the C.A. of other insects by the finding of Vogt (46) that this region causes at least local inhibition of metamorphic transforma- tion. There is reason to suppose, therefore, that the apparent divergence of the Diptera from other insects rests primarily upon the complex nature of the ring gland. Wiggles- worth (’48) further suggests that the prom- inence of the pupation-inducing activity among ring gland effects may rest on a great capacity of the C.A. region in the late larva for neutralizing any of the C.A. “inhibitory” hormone left in the tissues from earlier stages. According to this interpretation the relationships of the hormones in Diptera would differ only in minor respects from those in Rhodnius as indicated in Figure 225. MODE OF ACTION OF THE METAMORPHIC STIMULUS The striking similarity in pattern of the mechanisms by which metamorphosis is reg- ulated in insects and amphibians as discussed above encourages us to think that a further inquiry and detailed comparison may well enable the information in one of these fields to throw light upon the other. In that spirit let us proceed to consider a number of fun- damental questions in morphogenesis and te compare insects and amphibians with respect to them. HORMONAL NATURE OF REGULATORY MECHANISM Are the basic regulating mechanisms hormonal in both groups? The formal dem- onstration of the action of hormonal sub- stances of thyroid and anterior pituitary seem clear and fairly complete in regard to amphibian metamorphosis. Evidence that these glands act through hormones is varied, but most important is the fact that pure or relatively pure extracts are effective in re- pairing the deficiencies created by removal of these glands. The evidence is still incom- plete in that changes in the level of hormone concentration in the blood have not been demonstrated. In contrast, the evidence from insects is far less complete. Notably there is a lack of purified extracts. The report of Plagge and Becker (’38) appears to be the only one in which positive results were ob- tained with even crude extracts. Of course the effectiveness of body fluids which has been reported a number of times (see above and also Schmidt and Williams, ’49) is pre- sumptive evidence of the presence of a hor- METAMORPHOSIS 645 pK.co.osren_| = H.K.GOLDSTEIN | Fig. 226. Above, A chain of six brainless diapausing pupae had been established in parabiosis and a single chilled pupal brain implanted under the facial window of the anterior-most individual. When pupae 7 and 2 showed the initiation of development, about 17 days later, pupae 7 and 6 were detached as seen above. Below, The same preparation seven weeks later. After detachment of the brain- containing pupa, the activation has continued to spread down the chain of interconnected pupae. The isolated pupa, however, failed to develop. (After Williams, ’52.) 646 mone. However, since the general nutritive values or other positive or negative factors carried in these fluids may play a role in the reaction, they are not of themselves conclu- sive. In spite of these lacks the concept of a brain hormone and of a prothoracic hormone seem fairly well justified. With respect to the corpus allatum, the evidence for hormone action is less direct and the possibility of some action of the gland upon the metamor- phosis hormone other than by secretion of a separate principle has not been excluded. As noted above, Wigglesworth has postulated elimination of the C.A. hormone by this gland in the pre-metamorphic instar. In any case it seems important to realize that the properties of the metamorphic principles in insects may be somewhat different from those we are accustomed to see in vertebrate hormones. A notable instance is the slowness with which the principles may spread from one parabiont to another, as seen in the re- ports of Wigglesworth (734) and Willams (52; Fig. 226). Again the prominence of local effects from C.A., effects that persist through several molts (Piepho, *50) seems hardly consistent with the idea that the viable graft is secreting a hormone. The per- sistence of C.A. effects for one or two molts after removal of the corpus (Scharrer, °46; Pflugfelder, ’37) is also difficult to reconcile with the usual prompt disappearance of hor- mones from circulation, at least as hormone action is known in vertebrates. METABOLISM IN RELATION TO METAMORPHOSIS The well-established effect of the thyroid hormone in raising the rate of metabolism in mammals (Means, ’48) raises the prob- lems of whether this hormone has a similar effect in the amphibian larva and whether, if it has, this effect is responsible for meta- morphic changes. The possibility that a meta- bolic effect underlies metamorphosis has pro- vided the basis for numerous speculations on the nature of metamorphic change (e.g., Huxley, 29). Yet the weight of the experi- mental evidence is against the concept of a role of metabolic rate in metamorphosis. In the first place, despite the evidence of earlier studies it is doubtful whether the thyroid does exert a metabolism-increasing effect upon amphibian larvae. The evidence up to 1934 was examined by Etkin (’34), who con- cluded that the increase in oxygen consump- tion per unit weight that has been reported during thyroid-induced amphibian meta- METAMORPHOSIS morphosis did not signify an increase in metabolic rate, since it resulted from the loss of body water rather than from an increase in oxygen consumption per individual or per unit dry weight. In studies on the oxygen consumption of tadpoles during normal meta- morphosis in the bullfrog, Etkin found no increased oxygen consumption irrespective of the basis of calculation. The rate per indi- vidual animal decreased during the meta- morphic climax. Furthermore, the analysis of the literature on cold-blooded vertebrates shows that there is no adequate evidence for ascribing a metabolic rate effect to the thy- roid hormone in cold-blooded vertebrates generally (Etkin, Root and Mofshin, ’40). Though the evidence in this field continues to be conflicting, subsequent reports have in general confirmed the above interpretation (Smith and Everett, ’43). Even the workers who ascribe to thy- roid a stimulating effect on metabolism dur- ing amphibian metamorphosis do not regard this effect as of causal significance in the process of metamorphosis itself (Helff, ’26; Needham, ’42). Dinitrophenols, which raise oxygen consumption in mammals and fish (Means, *48; Root and Etkin, *37), do not influence amphibian metamorphosis (Cut- ting and Tainter, ’33). Cyanides have been reported to inhibit the effect of thyroid on metamorphosis, and this has been taken as supporting the metabolism theory in metamorphosis (Demuth, ’33; Hoff- man, 35). Borland (’43), however, found a marked accelerating effect of cyanides on metamorphosis. In any case the evidence from cyanides, like that from dinitrophenol, is too indirect to be very helpful on the question of the role of metabolic increase in metamor- phosis. Nevertheless, taken as a whole the weight of the evidence negates the idea of a causal role of a metabolism-raising factor in amphibian metamorphosis. An influence of the corpus allatum on metabolic rate in insects has been reported by Thomsen (749), who found a 20 per cent decrease after allectomy with almost com- plete restoration of the rate by replacement therapy. She regarded the activity of the hor- mone as being primarily on metabolic rate. The numerous influences exerted by the C.A. on the physiology of the adult insect (Pfeiffer, ’45b) may also be cited as evidence for a fundamental role of C.A. on basic meta- bolic processes. Basal metabolism, being the summation of numerous and varied metabolic pathways, is too general a concept to yield significant data METAMORPHOSIS for the analysis of morphogenic factors. Work on individual tissues or enzyme sys- tems holds much greater promise. Enzyme analysis has been most successfully applied to insects. Williams (751) and his collabora- tors, who have studied the cytochrome sys- tem, find a close correspondence between cytochrome c and the activity of prothoracic gland hormone. The emergence from dia- pause is correlated with a shift from a cy- anide-insensitive flavoprotein oxidation sys- tem to a cytochrome system sensitive to cy- anide. By inactivating cytochrome oxidase with carbon monoxide and other compounds Williams was able to stop the post-diapause growth processes in the pupa. The cyto- chrome system in Drosophila (Bodenstein and Sacktor, 52) also parallels the growth activity, although perhaps not the hormone level. Although only preliminary results are available at present it would appear that en- zymatic analysis holds great promise of an understanding of hormone activity in mor- phogenesis. This approach should be ex- tended to the amphibians, where the varia- tions among the tissues in responsiveness to thyroid would seem to provide favorable material for such analysis. DIRECT AND INDIRECT TISSUE RESPONSE Amphibians. The demonstration of a hor- monal stimulant in the blood raises the question as to whether the responses of the individual tissues are all made directly to this stimulus or whether some at least are not indirect responses induced by neighbor- ing tissues. This question is most conven- iently explored by transplanting tissues be- tween animals metamorphosing at different times. If the graft undergoes metamorphosis synchronously with the host tissue, then it may be concluded that it is responding di- rectly to the metamorphic stimulant rather than indirectly through some neighboring organ or by self-differentiation. Such evi- dence of direct response in amphibians exists for gills (Kornfeld, ’14), skin (Uhlenhuth, 17; Helff, ’31a), intestine (Sembrat, ’24), tail muscle (Helff and Clausen, ’29; Fukai, 34), tongue (Helff, 29), and eye (Vrtelowna, 25; Schwind, ’33). In some cases the tissue response is one of growth, in others one of histolysis or a combination of the two proc- esses. A particularly striking type of evidence of direct tissue response was secured by Koll- ros and collaborators (Kollros, 42; Kalten- 647 bach, ’49). For example, Kollros implanted bits of thyroid directly into the tadpole hind- brain and thereby brought about a localized maturation of the lid-closure reflex mecha- nisms. By irradiating parts of the tadpole with x-rays Puckett (37) was able to inhibit the erowth-promoting effects of thyroid and thereby separate the growth from the resorp- tion actions of the thyroid hormone, showing that these responses are independent of each other. On the other hand, Helff and his associates have been able to show that some tissue changes are not direct responses to hormones but are dependent upon stimulation from neighboring structures. The formation of the tympanum in the frog occurs during late stages in the metamorphic climax. The pro- duction of this structure involves a modifi- cation of the integument and also the forma- tion of a connective tissue sheet, the lamina propria, containing peculiar yellow fibers. Helff (28) showed that skin from any region transplanted to the ear region was induced to form tympanum. Conversely skin from the ear region transplanted elsewhere failed to form tympanum. By transplanting the an- nular tympanic cartilage he was able to show that this cartilage was the source of an inductive influence leading to the forma- tion of tympanic membrane in skin overly- ing it; this inductive effect was present, though weaker, in killed cartilage (Helff, 40). Helff (34) also found some capacity for tympanum induction in the quadrate and suprascapular cartilages which, because of their positions, can have no influence in the formation of the tympanum in normal metamorphosis. The inductive capacity of the annular cartilage, moreover, persisted for some time after the completion of the mem- brane in normal metamorphosis. However, the yellow fibers of the lamina propria were shown to form only under influence of con- tact with the columella (Helff, ’"31c). Pre- sumably the tympanic and columella carti- lages develop under direct thyroid stimula- tion since on transplantation each continues at least partial development (this point was not directly explored by Helff). A more complex case in which inductive phenomena play a part is that of the forma- tion of the skin window or opercular per- foration, for the foreleg in Anura. This is a definitely circumscribed area of skin of the operculum which undergoes autolysis in the late prometamorphic period (Fig. 217). The forelegs which have meanwhile developed in the opercular cavity are thrust through the 648 weak spot or opening formed in the auto- lyzed area. A number of mechanisms were suggested to explain the formation of the window. Weber (’25), for example, main- tained it is produced by the action of the poison glands at the base of the leg. The pres- sure of the developing leg is an obvious pos- sibility. But no experimental analysis was offered until Helff ((26) showed that skin from other areas transplanted to the appro- priate place would undergo autolysis during metamorphosis and, conversely, opercular skin transplanted to the back would not form the opening. This was interpreted as showing that the skin is not a self-differentiating structure nor does it respond directly to the thyroid hormone but, rather, it is induced to form the skin window in response to some local factor. The identification of any local factors con- cerned proved quite difficult. Helff was able to show that it was not the forelimbs, as their early removal did not prevent the histolysis in the region of the skin window from taking place. By pressure the legs accelerate the actual breaking through of the skin window but this is secondary to the histolysis. De- generating gill tissue, when transplanted to the back, induced histolysis in the overlying skin. Helff concluded that the autolyzing gill tissue was the inducing agent responsible for the formation of the skin window. This in- terpretation was supported by Van der Jagt (29), who found that gill tissue increased in potency for the induction of autolysis in the overlying skin as metamorphosis pro- gressed. However, this conclusion of Helff is not satisfactory for several reasons. In normal metamorphosis the opercular histolysis be- comes evident before (in Rana catesbeiana as much as a month before) any perceptible gill reduction occurs. Nor does the gill tissue make the close contact with the area of his- tolysis that Helff’s experiment indicates to be necessary for its action to be effective. Fur- thermore, the capacity for inducing histolysis in overlying integument seems to be wide- spread in degenerating tissue, muscle, for ex- ample, also showing this effect. Later investigations by European workers showed that, at least in the species investi- gated, different areas of skin have different potencies for skin window development. Opercular skin transplanted to the tail under- went histolysis when the tail resorbed, whereas skin from the back did not. In one species (R. ridibunda) opercular skin formed a perforation when transplanted to the back METAMORPHOSIS where no degenerating muscle underlay it (Blacher, Liosner and Woronzowa, ’34; Lios- ner and Woronzowa, ’35). In a reinvestiga- tion of this problem Helff (’39) likewise found that the skin area was not entirely without determination. He concluded that induction by degenerating gill tissue, by de- generating skin glands and by self-differen- tiation (i.e., direct response to hormone) each plays some part in the formation of the skin window. However, it would seem that until the tissue whose histolysis is active in inducing skin window formation is definitely identified (as explained above it does not seem possible that it is gill tissue) the exact role of self-differentiation cannot be deter- mined because the experimenter cannot be sure that the skin has not previously been exposed to some inductive influence when it is taken for transplantation. Recent work by Kaltenbach (’49) has shown that a fairly complete response can be elicited by implan- tation of thyroxine pellets to the neishbor- hood of the prospective skin window area, thereby greatly strengthening the case for self-differentiation. Other suggestions of dependent differentia- tion of tadpole tissues have been made from time to time, for example, that cutting off of circulation to the tail by the growth of the urostyle leads to tail resorption. This crude concept was disproved by Helff (30), who showed that the tail resorbed normally after extirpation of the urostylar primordium. Insects. That the hormonal factors in in- sect metamorphosis operate in Many Cases directly on the tissues seems evident. For example, implanted bits of skin lying in the haemocoel or in the fat body molt (Piepho, ’38; Bodenstein, 44) and imaginal discs continue their development (Boden- stein ,.43). Most striking is the fact reported by Williams (747) that isolated abdomens which have been largely cleaned out of in- ternal organs except the heart, still are able to respond to prothoracic gland stimula- tion. Although the few insect tissues that have been studied appear to respond directly to the metamorphic stimulus, indirect action can- not be excluded. It must be realized that in amphibians it is with regard to intra-organ details that Helff found indirect action of the metamorphic stimuli. Such intra-organ details have not yet been explored in insects with respect to metamorphic changes. It is worth noting, however, that local effects of one part upon another are well recognized in the sex organs in this group (Stern, ’41). METAMORPHOSIS TISSUE SPECIFICITY AND SENSITIVITY Amphibians. The responses of the tissues to the thyroid hormone are highly specific. Some tissues (i.e., leg and eye) show growth, others (i.e., tail tissues) show histolysis and resorption, still others (i.e., digestive organs) a combination of the two processes. Further- more, the response of a given organ varies among the species; the tail is resorbed and the legs are stimulated to growth in anurans but not in urodeles. Even among anurans, the degree of dependence of a given structure upon hormones appears to vary from species to species. Allen (725) found that in the ab- sence of the thyroid gland the legs of Bufo proceeded relatively further in their develop- ment than did those of various Ranidae. Gonadal development is generally independ- ent of the thyroid and metamorphosis (Swingle, °18; Hoskins and Hoskins, 719). But even this varies with the species; Krichel (31) found accelerated ovogenesis in Bufo viridis in normal metamorphosis and as a result of thyroid feeding. Continuance of the development of the brain in the absence of the thyroid from the larval to the adult con- dition was reported by Hoskins and Hoskins (19a), but subsequent investigation by Allen (24) showed that only the superficial form changes but no internal maturation occurs in the absence of the thyroid. The type of response made by a given group of cells is independentt of its histologi- cal character. In the tadpole, tail muscle and skin respond by histolysis whereas back muscle or skin do not. In fact, the respon- siveness of muscle varies with the region of the tail from which it is taken, anterior muscle degenerating more quickly than pos- terior and axial more readily than peripheral (Clausen, *30). Histological examination of metamorphos- ing tissues shows striking examples of speci- ficity of response that is independent of his- tological differentiation. Champy (’22) re- ported a sharp line of separation between the epithelium of the limb and that of the body with respect to their mitotic response to thy- roid treatment. Transplantation of the skin shows that the capacity for forming the glands of the dorsal plicae is strictly local- ized (Helff, ’31a), as is the capacity for pro- ducing pigment (Lindeman, ’29). Another striking case of cell specificity is that of the Mauthner cells, which degenerate under the influence of thyroid whereas neighboring cells may be stimulated to proliferate (Weiss 649 and Rossetti, 51; Kollros and Pepernik, ’52). The highly specific reactivity of larval structures described above is acquired at a definite and very brief period. The embry- onic tissues are not sensitive to thyroid, but they acquire their individual peculiarities of reactivity soon after the larval organs are definitely differentiated (Champy, ’22; Etkin, 60). Some experimenters report a relatively slight variation in the time at which differ- ent tissues in the same animal become sensi- tive to the thyroid hormone. The anuran tail is reported to develop sensitivity before the limb buds; the skin, particularly its pigment pattern, is last to become responsive (AI- phonse and Bauman, ’34; Kuhn, ’33; Moser, HOE A number of workers report progressive sensitization of the tissues to thyroid hor- mone during development (Allen, °38; Geigy, °41). The writer does not regard the evidence on this point as satisfactory because of the difficulty of controlling the quantita- tive evaluation of response and the influence of environmental factors. The basic property of thyroid hormone sensitivity is acquired in an all-or-none fashion at the time of oper- culum formation in Rana pipiens larvae (Etkin, 50). Changes in sensitivity, if any, are relatively slight. As compared to the role of tissue specificity, this factor can be of only secondary significance in determining the time and pattern of metamorphic changes. The problem of the acquisition of sensitivity of the tissues deserves a more sustained ex- perimental analysis, particularly as to its biochemical basis, than has as yet been given COP ite The problems of the evolution of the mech- anisms of metamorphosis do not fall within the scope of this chapter, but some compara- tive observations relating to tissue sensitivity may be made. In permanently neotenous urodeles the tissues are insensitive to the thy- roid hormone. Yet the thyroids and pituitary glands of these animals are effective when implanted into organisms capable of showing the metamorphic response (Swingle, 722; Charipper and Corey, ’30). Whether the hor- mone contained in these glands is ever re- leased during the lifetime of these animals is unknown. In the axolotl, a facultatively neotenous salamander, there is no loss of tissue sensi- tivity. This is shown most clearly by the fact that axolotl tissues transplanted to Triton hosts at embryonic stages metamorphose with the host (Geigy, ’38). Similarly, skin fragments transferred from axolotl to Amblys- 650 toma metamorphose with the latter, but in the reverse transplant, i.e., from Amblystoma to axolotl, the transplant remains indefinitely larval (Shtern, ’33). The well known arti- ficial induction of metamorphosis in the axolotl by thyroid and pituitary treatments likewise points to the existence of tissue sen- sitivity, although not indicating the level of sensitivity. It may be inferred that the neot- eny of the axolotl results from lack of glan- dular activity, that of perennibranchs from loss of tissue sensitivity. Insects. In insects as in amphibians each structure of an animal responds in its own specific way to the common metamorphic stimulus. Also with regard to the early time of acquisition of tissue specificity there is a striking similarity between amphibian tis- sues and those in insects (except possibly for Diptera, as noted below). Piepho (’38), for example, finds the capacity to pupate appears in caterpillars upon emergence from the egg. Similarly in Rhodnius the ability of early instars to respond to the metamorphic stimu- lus seems substantially complete (Wiggles- worth, *40). Bodenstein (735) found that the stage from which the thoracic leg transplant was taken made no difference in its respon- siveness to host stimuli in Vanessa. AI- though no quantitative evaluation of the sen- sitivity of the tissues seems to have been at- tempted, it can be seen from the facts stated that in these forms, at least, changes in tis- sue sensitivity do not play a significant role in determining the time or character of molt or metamorphosis. Scharrer (52a) has inter- preted the production of incomplete adults after C.A. removal at early instars in Leuco- phaea as evidence that the tissues in these animals are not yet fully competent to give the complete hormone response. However, such effects may merely result from the per- sistence of C.A. hormone as currently under- stood. In Drosophila, on the other hand, there is evidence that a progressive increase in re- sponsiveness of the tissues takes place during larval development. Hadorn and Neel (’38) found older larval tissues more sensitive than younger ones to the pupation stimulus from ring gland implants. Bodenstein (’43) re- ported greater response to transplanted ring glands from older imaginal discs, and similar results were reported by Vogt (742). It would appear from these results that changes in tissue responsiveness may play a significant role in insect development. On the other hand, the greater response of older discs may come about because they start from a more METAMORPHOSIS advanced base line rather than because of any increase in sensitivity. This interpreta- tion seems more consistent with the view, developed above, that larval development in Diptera as in other insects is controlled by a succession of hormonal pulses. Sensitivity in insects, as in amphibians, may be acquired fully at an early stage. METAMORPHIC PATTERN Amphibians. The importance in metamor- phosis of the time relationships of the changes which take place is illustrated by the abnormality of the metamorphic pattern induced by strong doses of thyroid (Fig. 227). The animals produced by such treatment usually die in the process of metamorphosis in a condition characterized by protruding, overdeveloped lower jaws, with tails largely resorbed but hind legs little more than un- differentiated nubbins. Large open skin win- dows for the forelegs may form but the legs are not sufficiently developed to protrude. Such distortions were early recognized and descriptively analyzed a number of times (Schreiber, 34a). Many authors speculated on the possibility that the normal pattern of metamorphic change results from differ- ences in the threshold of response of different tissues to thyroid hormone (see, for example, J. Huxley, ’22; Schreiber, ’34b). To test such theories a number of workers have studied the effects of immersion in different concen- trations of thyroid substances upon the progress of metamorphosis in tadpoles (Allen, °32b; Etkin, ’35a). Since the results of these studies, consistent among themselves, have repeatedly been misinterpreted as supporting the idea that differential thresholds of re- sponse account for the spacing of metamor- phic events (for example, see Needham, 42), it is necessary here to examine the idea of thresholds with some care. The technique used by Blacher and Allen was to observe the tadpoles after a definite period of immersion in solutions of different concentrations of thyroid substance and note which metamorphic changes had occurred. This procedure showed that events charac- teristic of the early stages of metamorphosis, such as leg growth and intestinal reduction, occurred within the specified period at lower concentrations of the thyroid hormone than did the later metamorphic events, such as mouth changes and forelimb emergence. It is clear, however, that this procedure does not reveal true thresholds, for it is possible that by observing the animals for longer METAMORPHOSIS periods even the lowest concentrations might be seen to lead to late metamorphic events. This was clearly recognized by Allen, who stated, “this ‘threshold of response’ is really a question of time of response.” By following the changes in individual animals continuously for long periods, Etkin found that any effective concentration of thy- 651 malities in coordination observed by all earlier workers who gave massive thyroid doses. No single concentration of thyroxine was found effective in inducing a normal ‘“‘time- table” of metamorphic events. Etkin was partly, though not entirely, successful in in- ducing a normal timetable in thyroidec- e S rs a 4s 8 x = Fig. 227. Abnormal tadpoles metamorphosing under stimulation from strong doses of thyroid (after Schreiber, roxine was capable of inducing both early and late metamorphic changes if allowed to operate for a long enough period. The speed of metamorphosis increased with the concen- tration of thyroxine. In low concentrations the first effect observed was an activation of leg growth, and only after the lapse of con- siderable time, during which the legs grew to full size, were late metamorphic events initiated. These later events proceeded in an extremely slow and protracted manner. With low concentrations, therefore, early meta- morphosis is normal in its time relations but the late events of the metamorphic cli- max are unduly protracted. In high concen- trations the events of the metamorphic cli- max were precipitated at their normal rates but not enough time was permitted for early events to undergo their appropriate develop- ment. This is the explanation of the abnor- 34a). tomized tadpoles by treating them first with low concentrations of thyroxine and subse- quently raising the concentration (Fig. 228). It would appear from the above results that the responses of the tissues show the ordinary stoichiometric relations common to many chemical reactions, i.e., the higher the con- centration the faster the reaction proceeds. No true thresholds of response are shown by the tissues. In urodeles the velocity of tissue response is also a function of the concentra- tion of active principle when small doses are used (Zavadovsky, ’26). According to this interpretation the tem- poral spacing of metamorphic events (Fig. 229) depends upon two factors: (1) inher- ent differences in the tissues with respect to rate of response, and (2) the patterns of activation of the thyroid gland. The histo- logical evidence discussed above indicates 652 that in Anura the thyroid gland at the begin- ning of metamorphosis is only slightly active but that through the prometamorphic period becomes more and more active until in the metamorphic climax it reaches a peak of activity accompanied by release of some of Normal Variable Daus METAMORPHOSIS Insects. No direct analysis of the problem of the correlation of metamorphic events is available in insects. Wigglesworth’s interpre- tation is that the same hormone is responsible for the metamorphic molt and for the growth of imaginal structures in Rhodnius. This is Fig. 228. Metamorphic pattern and thyroxine concentration in the tadpole. Graphs show hind leg growth. The times of occurrence of ten metamorphic events are shown by numbers. The top curve is that of a normal tadpole. The other curves show metamorphic pattern in thyroidectomized tadpoles treated with various thyroxine solutions. The number with each curve expresses the dilution of the solution exponentially, i.e., 6 equals one part thyroxine in 10® parts of water. Bottom graph of animal treated with increasing thyroxine concentrations starting from 9 and going to 6-2/3. (After Etkin, ’34.) its stored hormone. The relatively low thy- roid hormone concentration in the blood at the beginning of metamorphosis serves to activate the very sensitive leg primordia; the higher concentrations as they are built up in the blood speed the metamorphic proc- ess until in the metamorphic climax the flooding of the body with thyroid hormone produces the rapid reaction even in the least sensitive tissues, such as tail and tympanic ring. also the current interpretation of Williams (52). Hence it is possible that the temporal relationships of events in pupation and meta- morphosis are a function of the interplay of differential tissue sensitivity and changing hormone level as in amphibian metamorpho- sis. In Platysamia, for example, a pulse of low hormone activity leads to pupation. Then, after diapause a stronger pulse leads to com- pletion of imaginal development. On the other hand, the difference between METAMORPHOSIS Degree of Development Tail length Mouth widening a DAYS SWELL. i Fig. 229. Diagrammatic representation of rate of tissue changes during normal metamorphosis in tadpole. Note that each structure has its own time for beginning and completion of its metamorphic change. A.C.P., Resorption of anal canal piece; S.W.F.L., skin window for the forelegs; E, emergence of forelegs, arbitrarily designated as day number 30. metamorphic and larval molts in insects would appear to depend upon a qualitative difference in the stimulus, the presence or absence of the C.A. factor. In the interpre- tation of Wigglesworth (40, ’48) this factor serves to activate a different cellular mech- anism than does the metamorphic hormone. Because of the more rapid response of the tissues to C.A. hormone the activation of the metamorphic cellular mechanism is fore- stalled. Hence Wigglesworth’s term “juvenile hormone,” indicating a positive mode of ac- tion, replacing his original “inhibitory hor- mone” for the C.A. factor. It would appear, however, that the evidence for this mode of ac- tion is not clear-cut and an action of the C.A. hormone on the prothoracic gland hormone rather than directly on the tissues has not been excluded. The possibility exists that, at least for some organs, quantitative differ- ences in the degree of stimulation of the same cellular mechanism may account for the different developmental changes in molt- ing, pupation and metamorphosis. THE ACTIVATION OF THE METAMORPHIC MECHANISM In amphibians, metamorphosis depends upon a chain of events precipitated by the action of the thyrotrophic hormone of the anterior pituitary. In insects there is an analogous situation in the action of the brain in stimulating the prothoracic glands. In this section it is proposed to examine what is known of the factors determining the ac- tivation of the initiating stimulus from the pituitary or brain, 654 CONTROL OF PITUITARY ACTIVITY IN AMPHIBIANS The evidence indicates that the thyroid gland lies ready, so to speak, to respond to pituitary stimulation at all times from the pri- mordial stage onward. One experimental re- sult leading to this conclusion is the fact that when thyroid and pituitary primordia are brought close together in the tail-bud stage of the frog the thyroid is activated within a few days and an extremely precocious metamorphosis is induced (Etkin, ’36b, *39). The act of transplantation itself has no in- fluence, since transplants to other localities lead to no precocity (Hoskins and Hoskins, 19a; Etkin, °39). The thyroid activation must therefore be a result of locally concen- trated thyrotrophic pituitary hormone. An- other indication of the dependence of thyroid activation upon the time of arrival of the pituitary stimulus rather than upon any change in the thyroid itself is shown by the fact that the introduction of additional thy- roid primordia (not into the vicinity of the pituitary) does not influence the time of metamorphosis of the host (Allen, ’30; Choi, 32) In contrast to the thyroid the pituitary is independent of any other organ of the body for its activation. This conclusion is based on experiments in the transplantation of the pituitary primordium (Etkin, *38). When the primordium is removed from its normal site and transplanted to another part of the body, the animal undergoes metamorphosis. But this metamorphosis is generally delayed in its onset and protected in its execution. This is interpreted as indicating that the grafted pituitary generally develops its func- tion more slowly than normal. Unlike the thyroid, its function depends upon its devel- opmental rate, which is somewhat inhibited by the experimental interference, rather than upon a hormonal stimulus which would, of course, reach it at the same time wherever it is located. The precocious activation of the thyroid primordium when placed close to the pitui- tary primordium, as mentioned above, indi- cates that the production of thyrotrophic hormone begins very early. That the thyroid- stimulating field existing around the pitui- tary primordium is due to the same hormone as that which later normally stimulates thyroid development is indicated by the find- ing that histologically differentiated thyroids respond to the thyrotrophic field. Presumably the amount of hormone produced by the pri- METAMORPHOSIS mordium is so small as to be ineffectual when carried by the circulation. Before cir- culation is established, however, it appears to accumulate around the pituitary primor- dium in concentration sufficient to produce an effective thyrotrophic field. As to the pattern of post-primordial devel- opment of function in the pituitary, little is known. Using implantation techniques In- gram ('29a) found that it took 15 to 20 pituitaries from tadpoles at the beginning of metamorphosis to equal the effect of three adult frog pituitaries. This indicates, con- sidering the difference in size of the glands concerned (for which exact figures are not available), that the hormone is present in fair quantity in the tadpole pituitary. It must be supposed, therefore, that hormone production begins in the primordium and increases at an unknown rate to reach a high level at the beginning of metamorphosis. Whether the subsequent increase in thyroid activity to the climax of metamorphosis and the eventual cessation of thyroid activity at the close of metamorphosis are paralleled by corresponding changes in the activity of the pituitary is not known. As seen above, the failure of the axolotl to metamorphose is not due to insensitivity of the tissues to thyroid hormone or of the thyroid to pituitary stimulation. It must therefore rest somehow on the failure of the pituitary to become active. Bytinsky-Salz (35) explored this problem by exchanging hypophyseal rudiments between the axolotl (Amblystoma mexicanum) and a variety of amblystoma (A. tigrinum) which metamor- phoses normally. He found that the meta- morphic response of the host was that char- acteristic of the host species, not of the pituitary-donor type. This he interpreted (contrary to the conclusion stated above) as indicating that the pituitary is not function- ally self-differentiating but is dependent upon some controlling factor elsewhere in the body for its own activation. However, the writer does not find the evidence pre- sented to be convincing. The grafts were made in the orthotopic position, being made to replace the animal’s own _ prospective pituitary area which had been removed. It is therefore impossible to be sure that the pituitary which developed in the animal came from the graft rather than by regen- eration of the host’s own tissues. This criti- cism is especially pertinent since hypophy- sectomies similarly carried out by Blount (32) yielded a high proportion of regener- ated glands. METAMORPHOSIS In this connection it is interesting to note that in the neotenous Necturus maculosus, pituitaries of the adult were found potent in thyrotrophic principle when tested by implantation (Charipper and Corey, °30). Presumably in this form the pituitary-thy- roid relation remains normal in spite of the loss of the metamorphic response of the tissues to thyroid hormone. EFFECTS OF ENVIRONMENTAL FACTORS IN AMPHIBIANS Considerable attention has been given to the influence of environmental factors on amphibian metamorphosis. From the vantage point of our present knowledge of the hor- monal mechanisms of metamorphosis the effects of environmental factors must be considered in terms of their possible influ- ences on these mechanisms, particularly on the activation of the pituitary. They may therefore be appropriately considered at this point. An external factor may influence the sensitivity or capacity for response of the tissues to thyroid and thus affect metamor- phosis. It is clear that Roentgen rays act in this way since Puckett (737) found it pos- sible to prevent growth processes by x-ray dosage while resorption changes were per- mitted to continue. Environmental factors may, on the other hand, influence metamorphosis by their ef- fects on the thyrotrophic activity of the pituitary. Starvation inhibits the onset of metamorphosis in tadpoles when it comes before the beginning of metamorphosis, but accelerates it thereafter (D’Angelo, Gordon and Charipper, 41). The tissues of starved animals retain their sensitivity to thyroxine and their thyroids respond to anterior pitui- tary injections. It is therefore on the thyroid- activating mechanism of the pituitary that starvation before metamorphosis must act. The functional self-differentiation of the pituitary may be supposed to be brought to a halt in the absence of nourishment. Once ‘he pituitary has been completely activated, +n the other hand, the reduced food supply <2ems to make the released thyroid hormone miore effective, perhaps because the cessation of tissue growth permits more active tissue differentiation. Crowding inhibits growth more than it does metamorphosis (Adolph, °31). This would also appear to be due to a differential effect, greater on the growth and development of the pituitary than on that of the body as a whole. 655 The effects of temperature on metamor- phosis appear to operate both on tissue sensi- tivity and on the endocrine mechanism. A number of workers report that the failure of metamorphosis which occurs in the cold (below 10°C.) is accompanied by loss of tissue sensitivity to thyroid (Huxley, 29; Hartwig, ’36). Yet an effect of low temper- ature also on the endocrine mechanism is indicated by the failure of hibernating tad- poles to continue their metamorphosis when restored to warm conditions (Fosi, ’35), and by the fact that manipulation of the tem- perature can differentially affect metamor- phosis and growth (Adler, °16). At 16°C. tadpoles grow larger before metamorphosis than at 24°C. (Etkin, unpublished data). ACTIVATION OF INSECT MECHANISMS The release of the hormonal mechanism of molt and metamorphosis in Rhodnius_ is clearly dependent upon the feeding stimulus, since metamorphosis always follows feeding by a definite number of days, 15 to 20 according to stage (Wigglesworth, °34). If the ventral nerve cord is cut molting is prevented. Presumably feeding provides a stimulus that reaches the brain through the nerve cord. Wigglesworth maintains that the stimulus arises from the stretching of the body wall. Clearly the mechanism of activation of the brain in Rhodnius cannot apply to in- sects which do not feed in single huge meals. But we have very little indication of the mechanism governing hormone production or release in other insects. The effects of starvation on insect metamorphosis are sim- ilar to those on amphibians. Before a critical period which is near the time of metamor- phosis, starvation delays metamorphosis; after the critical period it may have no effect or may even accelerate the process (Bounhiol, ’38). As explained above, Wil- liams found that chilling was necessary for the activation of the brain in the diapausing larva of Platysamia cecropia. A comparison of the insect and amphibian reveals a highly suggestive similarity in the relation of the primary endocrine apparatus (i.e., neurosecretory brain cells in insects and pituitary in amphibia) to environmental factors. The vertebrate anterior pituitary is anatomically closely related to the brain both in position and by way of the infundib- ulum. There appear, however, to be no secretory nerves from the brain to the ante- rior lobe. However, in recent years the blood 656 supply of the anterior lobe has been shown to consist in part of a portal venous system which brings blood from the hypothalamus to the pituitary (Green, 751). It is believed that through this venous system the neuro- secretory apparatus in the hypothalamus is able to exert some controlling influences on the activity of the anterior lobe. It would thus appear that the varied instances where environmental factors influence the meta- morphic process in insects and amphibians operate through a basically similar mech- anism. The stimuli are received by the nervous system, which transmits them to neurosecretory elements which, in turn, start the endocrine chain mechanism into action. The intimate association of the brain and the primary endocrine apparatus thus per- mits the development of some degree of control of the endocrine system through environmental factors. In this way the life history of an animal can be brought into synchrony with seasonal change. It would be interesting to explore the influence of cold on the pituitary of such amphibians as hibernate in larval stages, along the lines of the work of Williams on Platysamia.* In connection with the neurosecretory pathway in insects Scharrer (’52a,b) has suggested that the corpus cardiacum be con- sidered part of the neuroendocrine complex regulating insect development. She has dem- onstrated that on an anatomical basis the C.C. might be considered as a storage organ for the secretion formed by the neurosecre- tory cells of the brain [perhaps analagous to the sinus gland of crustacea (Passano, 51) ]. Yet experiments with the C.C. have not yielded dramatic results. Pfeiffer (39) asserted that C.C. removal delays molting in the grasshopper, which is consistent with the above theory. Vogt (’46), on the other hand, found in Drosophila that C.C. implants delay puparium formation. Perhaps clear- cut results are not to be expected until both brain and C.C. are treated as a unit. CESSATION OF METAMORPHIC TRANSFORMATION Amphibians. As a final step in the regula- tion of metamorphosis, the processes involved * The posterior lobe of the pituitary (not involved in metamorphosis) is innervated by neurosecretory fibers from the hypothalamus. The extraordinary morphological parallelism between this and the neurosecretory system of insects has been com- mented upon by Scharrer and Scharrer (’44). Per- haps this, too, finds its ultimate significance in a linking up of the endocrine gland with environ- mental factors. METAMORPHOSIS must be brought to a halt. When resorption of the tadpole’s tail is complete, for example, no further change is possible. When the tympanic membrane is fully formed no further development seems possible. Yet it is apparent that such considerations do not apply to all metamorphic events. Do the legs of the tadpole cease their rapid growth because they lose their capacity to respond to thyroid or because the hormone itself ceases? It would appear probable that the sensitivity of the leg tissues changes with development, since the legs slow their growth before the metamorphic climax is complete and while the thyroid gland is still very active (Etkin, 32, ’36a). On the other hand, skin shedding is a characteristic phenomenon of metamorphosis in Amphibia (Etkin, 732). Plainly the capacity of the skin for respond- ing to thyroid by shedding is not lost at the end of metamorphosis, since it is found in the adult. It must be supposed, therefore, that the cessation of skin shedding late in metamorphosis is a consequence of the re- duction in activity of the thyroid which, according to histological studies, takes place at this time (Etkin, ’36a). The inactivation of the thyroid is itself an aspect of metamorphosis that requires explanation. No direct evidence is available to indicate whether this is to be ascribed to a change in sensitivity, as in the case of the tadpole’s legs, or to a cessation of pituitary stimulation. From what is known of the general physiology of the thyroid-pituitary relation we might suppose that the second of the above alternatives holds true, since prolonged stimulation of the thyroid gland by pituitary implants is possible. Further- more, a mechanism for the suppression of thyrotrophic function in the pituitary is suggested by the observation in many ani- mals that thyroid hormone itself, when it reaches a high level in the blood, inhibits the thyrotrophic activity of the pituitary (for a summary of the evidence on this point see Adams, *46). It can be seen from the above discussion that, though several factors appear to con- tribute to the mechanism bringing metamor- phic change to an end, little direct experi- mental knowledge in this field is available. Insects. In insects as in amphibians some metamorphic changes have an inherent end point, as in the resorption of larval organs. This may also apply to constructive changes at metamorphosis when the living cells of the primordium are used up in the metamor- phic change, as in the case of the formation METAMORPHOSIS of the wings. This situation cannot account entirely for the absence of further develop- ment in the adult insect, for experiments reveal that the hypodermis, at least, is po- tentially capable of further molting and can even revert to the production of larval exoskeleton under appropriate stimulation (Fururkawa, ’35; Piepho, ’39). The absence of metamorphic change in the adult insect is at least partly due to failure of the hormonal pulses to continue. The disintegration of the prothoracic glands at the end of metamorpho- sis (Kaiser, °49; Rahm, ’52) accounts for the cessation of the hormonal stimulation there- after. What determines this disintegration or the cessation of C.A. activity at the pre- vious stage in insects is not clear. In any case it is evident that the morpho- genic processes involved in the cessation of metamorphic activity involve both the loss of sensitivity of the tissues and cessation of hormonal activity in both insects and am- phibians. It is also noteworthy that the maintenance of the adult morphological structures brought into existence under hor- monal stimulation does not require the con- tinued presence of that stimulation for their maintenance. In this respect the morpho- genic principles involved in metamorphosis differ from those common, though not uni- versal, in the sex system of vertebrates, where most structures require the continued presence of the hormone under which they developed for their maintenance. This differ- ence correlates, of course, with the cyclic or seasonal nature of sex activity in contrast to the stability of adult characters. A report by Kollros and Pepernik (52), however, indicates that some metamorphic events in amphibians may also depend for their main- tenance upon continued thyroid stimulation. They found that the neurones of the mesence- phalic V nucleus regress if thyroid hormone is withdrawn. SUMMARY The concept of the metamorphic mech- anism operative in amphibian metamorpho- sis may be stated as follows: The anterior pituitary begins its thyro- trophic activity early in embryonic develop- ment. However, it is not until late in larval life that the rate of hormone production reaches a level high enough to stimulate a pulse of growth and secretion in the thyroid gland. With the activation of the thyroid the concentration of its hormone in the blood rises according to a definite pattern. Each 657 of the larval structures responds to thyroid hormone in its own specific fashion and each at its own rate. As a result of the differ- ences in speed (not in thresholds) of re- sponse of the different tissues metamorphic changes occur in a definite sequence of appropriately spaced events. Metamorphic changes are further integrated by inductive effects of one tissue upon others in its im- mediate neighborhood. The pulse of thyroid activity is brought to an end by inhibition of pituitary activity by the high level of thyroid hormone. Subsequent pulses of thy- roid activity during the life of the amphibian produce no metamorphic change because the structures concerned have either com- pleted the possibilities of change or lost their tissue sensitivity. In insects the first known stimulus to metamorphosis originates in the neurose- cretory cells of the brain. These produce a hormone acting upon the prothoracic glands, which in turn produce a hormone acting upon the tissues. Each tissue responds to the hormone of the prothoracic gland by its own characteristic metamorphic transforma- tion. This pulse of hormonal activity is the last of a series of such pulses that occur during larval life. The earlier pulses do not eventuate in metamorphosis because each is accompanied by the secretion of a hormone by the corpus allatum. The response of the tissues to the presence of both prothoracic gland hormone and the hormone of the cor- pus allatum is to undergo a larval rather than a metamorphic molt. In both the amphibian and the insect the metamorphic pattern is the product of an interaction of two basic factors, a pulse of hormonal activity and an inherent pattern of tissue sensitivity. In both organisms the hormonal pulse is produced by two glands. The second gland depends for activation upon the first and the first is intimately related to the brain. The mode of activation of the primary hormonal factor is but little known. Physio- logical self-differentiation may be predomi- nant in most cases. In others environmental factors or the general body metabolism may play a role by way of the brain. The nature of the tissue response presumably rests on an enzymatic basis, as work on insects is beginning to elucidate. The remarkable parallelism between the metamorphic mechanism in insect and am- phibian can have no basis in homology but must be ascribed to the fundamental similar- ity of the life processes in all organisms. 658 There appears to be a limited repertoire of physiological mechanisms by which living things can meet a given problem. Both in- sects and amphibians were presented by a common evolutionary challenge, to evolve a metamorphic stage transitional between a larva with one mode of life and an adult with another. Both organisms have evolved closely analogous physiological mechanisms to meet this challenge. REFERENCES Abderhalden, E. 1915 Studien iiber die von ein- zelnen Organen hervorgebrachten Substanzen mit specifischer Wirkung, I. Pfliigers Archiv., 162:99-129. 1924 Fortgesetzte Studien iiber die Beein- flussung der Entwicklung von Kaulquappen durch Verbindungen mit bekannter Struktur. Pfliigers Archiv., 206:467-472. Adams, A. E. 1946 Variations in the potency of thyrotrophic hormone of the pituitary in animals. Quart. Rev. Biol., 27:1—32. Adler, L. 1914 Metamorphosestudien an Batra- chierlarven. I. Roux’ Arch. Entw.-mech., 39:21- 45. 1916 Untersuchungen iiber die Entste- hung der Amphibienneotenie. Pfliigers Archiv., 164:1-101. Adolph, E. 1931 Body size as a factor in the metamorphosis of tadpoles. Biol. Bull., 67:376- 386. Allen, B. M. 1916 Extirpation experiments in Rana pipiens larvae. Science, 44:755-757. 1917 Effects of the extirpation of the an- terior lobe of the hypophysis of Rana pipiens. Biol. Bull., 32:117-130. 1918 The results of thyroid removal in the larvae of Rana pipiens. J. Exp. Zool., 24:499- 519. 1924 Brain development in anuran larvae after thyroid or pituitary gland removal. Endocrin., 8:639-651. 1925 The effects of extirpation of the thy- roid and pituitary glands upon the limb develop- ment of anurans. J. Exp. Zool. 42, 413-430. 1927 Influence of the hypophysis upon the thyroid gland in amphibian larvae. Univ. Calif. Pub. Zool., 37:53-78. 1929a The influence of the thyroid and hypophysis upon growth and development of am- phibian larvae. Quart. Rev. Biol., 4:325-352. 1929b Transplants of the thyroid anlagen into anuran tadpoles. Anat. Rec., 44:207. 1930 The early development of organ anlagen in amphibians. Contrib. Marine Biol., Stanford University Press. 1931 Role of hypophysis in the initiation of metamorphosis in Bufo. Proc. Soc. Exper. Biol. & Med., 29:74-75. 1932a The dominant role of the pars an- terior of the hypophysis in initiating amphibian metamorphosis. Anat. Rec., 54:65-81. METAMORPHOSIS 1932b_ The response of Bufo larvae to dif- ferent concentrations of thyroxin. Anat. Rec., 54: 45-65. 1938 The endocrine control of amphib- ian metamorphosis. Biol. Rev., 73:1-19. Alphonse, P., and Bauman, G. 1934 Indifférence de la peau de jeunes tetards de Bufo vulgaris vis- a-vis de fortes doses de thyroxine. Compt. Rend. Soc. Biol., 117:567. Atwell, W. J. 1935 Effects of thyreotropic and adrenotropic principles on hypophysectomized Amphibia. Anat. Rec., 62:361-379. 1937 Functional transplants of the pri- mordium of the epithelial hypophysis in am- phibia. Anat. Rec., 68:431-447. , and Taft, J. 1940 Functional transplants of epithelial hypophysis in three species of Am- blystoma. Proc. Soc. Exper. Biol. & Med., 44:53- 55. Baffoni, G. N., and Catte, G. 1950 Il comporta- mento della cellula di Mauthner di raganella nella metamorfose accelerata con sommini stra- zione di tiroide. Atti della Accad. Naz. dei Lincei, 8 Ser., 9:282-287. Balthasart, M. 1931 La metamorphose experi- mentale des amphibiens. Ann. Soc. Roy. Zool. Belg., 62:79-114. Blacher, L. J., Liosner, D., and Woronzowa, M. 1934 Mechanismus der Perforation der oper- cularen Membran der schwanzlosen Amphibien. Bull. Acad. Polonaise Sc. et Let., B, 2:325-— 347. Bliss, D., and Welsh, J. H. 1952 The neurosecre- tory system of brachyuran Crustacea. Biol. Bull., 103:157-169. Blount, R. 1932 Transplantation and extirpation of the pituitary rudiment and the effects upon pig- mentation in the urodele embryo. J. Exp. Zool., 63:113-1441. 1935 Size relationships as influenced by pituitary rudiment implantation and extirpation in the urodele embryo. J. Exp. Zool., 70:131-185. Bodenstein, D. 1933 Beintransplantationen an Lepidopterenraupen. I. Roux’ Arch. Entw.-mech., 128:564-583. 1935 Beintransplantationen an Lepidop- terenraupen. III. Roux’ Arch. Entw.-mech., 133: 156-192. 1938a Untersuchungen zum Metamor- phoseproblem. I. Roux’ Arch. Entw.-mech., 137: 474-505. 1938b Untersuchungen zum Metamor- phoseproblem. II. Roux’ Arch. Entw.-mech., 137: 636-660. 1939 Investigations on the problem of metamorphosis. VI. J. Exp. Zool., 82:329-356. 1942 Hormone controlled processes in insect development. Cold Spring Harbor Symp. Quant. Biol., 70:17-26. 1943 Hormones and tissue competence in the development of Drosophila. Biol. Bull., 84:34 58. 1944. The induction of larval molts in Drosophila. Biol. Bull., 86:113-124. , and Sacktor, B. 1952 Cytochrome c oxi- METAMORPHOSIS dase activity during the metamorphosis of Dro- sophila virilis. Science, n.s. 116:299-300. Bolten, M. 1926 Ein Fall von Thyroid-Insuffi- zienz bei einer Froschlarve. Nederlandsel Tydsch v. Geneesk., 70:1711-1713. Borland, J. 1943 The production of experimental goiter in Rana pipiens tadpoles by cabbage feed- ing and methyl cyanide. J. Exp. Zool., 94:115- 140. Bounhiol, J. 1937 La métamorphose des insectes serait inhibée dans leur jeune age par les corpora allata? Compt. Rend. Soc. Biol., 726:1189-1191. 1938 Recherches expérimentales sur le déterminisme de la métamorphose chez les Lépi- doptéres. Bull. Biol. France et Belg. Suppl., 24: 1-199. Brinck, H. 1939 A histological and cytological investigation of the thyroids of Arthroleptella bi- color villierst and Bufo angustriceps during the normal and experimentally accelerated meta- morphosis. Proc. Linn. Soc. London, 157:120-125. Buddenbrock, W. von 1931 Untersuchungen liber die Hautungshormone der Schmetterlings- raupen, II. Zeit. f. vergl. Physiol., 74:415—-428. Burtt, E. 1938 On the corpora allata of dipterous insects, II. Proc. Roy. Soc. London B, 126:210- O23: Bytinsky-Salz, H. 1935 Heteroplastic transplan- tation of the hypophysis in Amblystoma. J. Exp. Zool., 72:51-73. Caspari, E. 1941 The influence of low tempera- ture on the pupation of Ephestia kuhniella, Zeller. J. Exp. Zool., 86:321-331. , and Plagge, E. 1935 Versuche zur Phys- iologie der Verpuppung von Schmetterlingsrau- pen. Naturwiss., 23:751-752. Cazal, P. 1948 Les glandes endocrines rétro- cérébrales des insectes. Bull. Biol. France et Belg. Suppl., 32:1-227. Champy, C. 1922 L’action de l’extrait thyroidien. Arch. Morph. Gen. Exper., 4:1-56. Charipper, H., and Corey,C. 1930 Studies on am- phibian endocrines, V. Anat. Rec., 45:258. Choi, M. 1932 Homoiotransplantation of the am- phibian thyroid anlage. Folio Anat. Jap., 70:25- 27. Clausen, H. 1930 Rate of histolysis of anuran tail skin and muscle during metamorphosis. Biol. Bull., 59:199-210. Clements, D. 1932 Comparative histological studies of the thyroids and pituitaries in frog tad- poles in normal and accelerated metamorphosis. J. Roy. Microsc. Soc., 52:138-148. Cutting, C., and Tainter, M. 1933 Comparative effects of dinitrophenol and thyroxine on tadpole metamorphosis. Proc. Soc. Exper. Biol. & Med., 31:97-100. D’Angelo, D., Gordon, A., and Charipper, H. 1944 The role of the thyroid and pituitary glands in the anomalous effect of inanition on amphibian metamorphosis. J. Exp. Zool., 87:259-277. Demuth, F. 1933 Uber die Beziehungen des En- ergiestoffwechsels zu Wachstum und Differen- zierung. Roux’ Arch. Entw.-mech., 730:340-352. Etkin, W. 1930 Growth of the thyroid gland of 659 Rana pipiens in relation to metamorphosis. Biol. Bull., 59:285-292. 1932 Growth and resorption phenomena in anuran metamorphosis, I. Physiol. Zool., 5: 275-300. 1934 The phenomena of anuran meta- morphosis, II. Physiol. Zool., 7:129-148. 1935a The mechanisms of anuran meta- morphosis, I. J. Exp. Zool., 77:317-340. 1935b_ Effects of multiple pituitary pri- mordia in the tadpole. Proc. Soc. Exper. Biol. & Med., 32:1653-1655. 1936a The phenomena of anuran meta- morphosis, III. J. Morph., 59:69-90. 1936b A thyrotropic field surrounding the immature pituitary of the tadpole. Proc. Soc. Exper. Biol. & Med., 34:508-512. 1938 The development of thyrotropic function in pituitary grafts in the tadpole. J. Exp. Zool., 77:347-377. 1939 A thyrotropic field effect in the tad- pole, I. J. Exp. Zool., 82:463-496. 1950 The acquisition of thyroxine-sensi- tivity by tadpole tissues. Anat. Rec., 108:541. , Root, R., and Mofshin, B. 1940 The ef- fect of thyroid feeding on oxygen consumption of the goldfish. Physiol. Zool., 73:415—-429. Figge, F. H. 1930 A morphological explanation for failure of Necturus to metamorphose. J. Exp. Zool., 56:241-254. 1934 The effect of ligation of the pul- monary arch on amphibian metamorphosis. Phys- iol. Zool., 7:149-177. , and Uhlenhuth, E. 1933 The morphol- ogy and physiology of the salamander thyroid gland, VIII. Physiol. Zool., 6:450—-465. Fosi, V. 1935 Osservzioni sull’ influenza della temperatura e degli estratte tiroidei sulla neotenia parziale dei girini di Rana esculenta. Monitore Zool. Ital., 46:249-252. Fraenkel, G. 1935 A hormone causing pupation in the blowfly Calliphora erythrocephala. Proc. Roy. Soc. London. B, 778:1-12. Frew, J. 1928 A technique for the cultivation of insect tissues. J. Exp. Biol., 6:1-11. Fukai, T. 1934 On the synchronous degenera- tion of the transplanted tail and the original tail in the metamorphosis of Bufo larva. Folia. Anat. Jap., 12:159-164. Fukuda, S. 1939 Acceleration of development of silkworm ovary by transplantation into young pupa. Proc. Imp. Acad. Japan, 75:19-21. 1940a Induction of pupation in silkworm by transplanting the prothoracic gland. Proc. Imp. Acad. Japan, 16:414416. 1940b Hormonal control of molting and pupation in the silkworm. Proc. Imp. Acad. Ja- pan, 76:417-420. 1941a Induction of metamorphosis in the silkworm by transplanting pupal prothoracic gland. Zool. Mag. (Tokyo), 53:582-584. 1941b Role of the prothoracic gland in differentiation of the imaginal characters in the silkworm pupa. Annot. Zool. Japon., 20:9-13. 1944 The hormonal mechanism of larval 660 molting and metamorphosis in the silkworm. J. Fac. Sci. Imp. Univ. Tokyo, Sect. 4, 6:477-532. Furukawa, H. 1935 Can the skin of imago be made to molt. Proc. Imp. Acad. Japan, 77:158- 160. Garber, S. T. 1930 Metamorphosis of the axolotl following lung extirpation. Physiol. Zool., 3:373- 378. Geigy, R. 1938 Entwicklungsphysiologische Un- tersuchungen iiber die Anuren- und Urodelen- Metamorphose, II. Verh. Schweizer Naturforsch. Ges., 179:1-5. 1941 Thyroxineinwirkung auf verschie- den weit entwickelte Froshlarven. Verhandl. Schweiz. Naturforsch. Ges., 727:161-164. Gorbman, A. 1946 Qualitative variation of the hypophyseal thyrotropic hormone in the verte- brates. Univ. Calif. Pub. Zool., 57:229-244. Gordon, A., Goldsmith, E., and Charipper,H. 1943 Effect of thiourea on the development of the am- phibian. Nature, 752:504-505. Grant, M. 1931 The release of follicular colloid from the thyroid of Amblystoma jeffersonianum following heteroplastic pituitary implants. Anat. Rec., 49:373-395. Green, J. D. 1951 The comparative anatomy of the hypophysis with special reference to its blood supply and innervation. Am. J. Anat., 88:225- Sir Gudernatsch, [J.] F. 1912 Feeding experiments on tadpoles, I. Roux’ Arch. Entw.-mech., 35:457- 483. 1933, Entwicklung und Wachstum; in Handbuch der Inneren Sekretion, edited by M. Hirsch, Vol. 2, pt. 2, pp. 1493-1744. Curt Kabitzsch, Leipzig. Gutman, A. 1926 Metamorphosis in Necturus maculatus by means of thyroxin-adrenalin treat- ment. Anat. Rec., 34:133-134. Hachlow, V. 1931 Zur Entwicklungsmechanik der Schmetterlinge. Roux’ Arch. Entw.-mech., 125:26-49. Hadorn, E. 1937 An accelerating effect of nor- mal “ring-glands’” cn puparium-formation in lethal larvae of Drosophila melanogaster. Proc. Nat. Acad. Sci., 23:478-484. , and Neel, J. 1938 Der hormonale Ein- fluss der Ringdriise (corpus allatum) auf die Pupariumbildung bei Fliegen. Roux’ Arch. Entw.-mech., 738:281-304-. Hartwig, H. 1936 Uber die Beziehungen zwi- schen Schilddriise und Entwicklung bei Sala- manderlarven unter dem Einfluss verschiedener Temperaturen. Roux’ Arch. Entw.-mech., 134: 562-587. Hegner, R. 1922 The effects of prostate substance on the metamorphosis of the intestine of frog tad- poles. Am. J. Physiol., 67:298-299. Helff, O. M. 1926 Studies on amphibian meta- morphosis, I. J. Exp. Zool., 45:1-67. 1928 Studies on amphibian metamor- phosis, III. Physiol. Zool., 7:463-495. 1929 Studies on amphibian metamor- phosis, IV. Physiol. Zool., 2:334-341. METAMORPHOSIS 1930 Studies on amphibian metamor- phosis, VIII. Anat. Rec., 47:177-186. 1931a Studies on amphibian metamor- phosis, IX. Biol. Bull., 60:11-22. 1931b Studies on amphibian metamor- phosis, VI. J. Exp. Zool., 59:167-177. 1931c Studies on amphibian metamor- phosis, VII. J. Exp. Zool., 59:179-196. 1934 Studies on amphibian metamor- phosis, XII. J. Exp. Zool., 68:305-319. 1939 Studies on amphibian metamor- phosis, XVI. J. Exp. Biol., 76:96-117. 1940 Studies on amphibian metamorpho- sis, XVII. J. Exp. Biol., 17:45-60. , and Clausen, H. 1929 Studies on am- phibian metamorphosis, V. Physiol. Zool., 2:575- 586. Hoffman, O. 1935 The antagonistic effect of methyl-cyanide on thyroxin-induced metamor- phosis. J. Pharm. & Exper. Therap., 54:146. Hoskins, E., and Hoskins, M. 1917 On thyroidec- tomy in Amphibia. Anat. Rec., 77:363. , and Hoskins, M. 1919a Growth and de- velopment of Amphibia as affected by thyroidec- tomy. J. Exp. Zool., 29:1-69. , and Hoskins, M. 1919b Experiments with the thyroid, hypophysis and pineal glands of Rana sylvatica. Anat. Rec., 16:151. Huxley, J. 1922 Ductless glands and develop- ment. J. Hered., 13:349-358. 1929 Thyroid and temperature in cold- blooded vertebrates. Nature, 723:712. Ingram, W. 1929a Studies of amphibian neot- eny, II. J. Exp. Zool., 53:387-410. 1929b Studies in amphibian neoteny, I. Physiol. Zool., 2:149-156. Joly, P. 1945 Les correlations humorales chez les insectes. Ann. Biol., 27:1-34. Kaiser, P. 1949 Histologische Untersuchungen uber die Corpora allata und Prothoraxdriisen der Lepidopteren in Bezug auf ihre Funktion. Roux’ Arch. Entw.-mech., 744:99-131. Kaltenbach, J. 1949 Local metamorphosis of Rana pipiens larvae by thyroxin-cholesterol im- plants. Anat. Rec., 103:544. Kollros, J. 1942 Localized maturation of lid-clo- sure reflex mechanism by thyroid implants into tadpole hindbrain. Proc. Soc. Exper. Biol. & Med., 49:204-206. , and Pepernik, V. 1952 Hormonal con- trol of the size of mesencephalic V nucleus cells in Rana pipiens. Anat. Rec., 113:527. , Pepernik, V., Hill, R., and Kaltenbach, J. 1950 The growth of mesencephalic V nucleus cells as a metamorphic event in anurans. Anat. Rec., 708:565. Kopec, S. 1922 Studies on the necessity of the brain for the inception of insect metamorphosis. Biol. Bull., 42:322-342. 1924 Studies on the influence of inani- tion on the development and the duration of life in insects. Biol. Bull., 46:1-21. Kornfeld, W. 1914 Abhangigkeit der metamor- photischen Kiemenriickbildung vom Gesamtor- METAMORPHOSIS ganismus der Salamandra maculosa. Roux’ Arch. Entw.-mech., 40:369—41 4. Krichel, W. 1931 Der Einfluss thyroidaler Sub- stanzen auf Larven von Bufo viridis und die Bedeutung dieser Stoffe fiir die Entwicklung der Keimdriise bis zur Metamorphose. Zool. Jahrb. Abt. f. Allg. Zool. u. Physiol., 48:589— 666. Krichesky, B. 1934 The response of Rana cates- beiana larvae to injections of antuitrin and phy- one. Physiol. Zool., 7:178-191. Kiihn, A., and Piepho, H. 1938 Die Reaktionen der Hypodermis und der Versonschen Driisen auf das Verpuppungshormon bei Ephestia Kiihniella. Biol. Zentralbl., 58:12-51. Kuhn, O. 1933 Uber morphogenetische Schild- driisen-Hormonwirkungen in frithen Entwick- lungsstadien. Nachr. Ges. Wiss. Gottingen, Math.- physik. K1., 7:13-35. Lerman, J., and Salter, W. 1939 The relief of myxedema with proteins of extra-thyroidal ori- gin. Endocrin., 25:712-720. Liebman, E. 1946 On trephocytes and trephocy- tosis; a study on the role of leucocytes in nutrition and growth. Growth, 10:291-329. Lindeman, V. 1929 Integumentary pigmentation in the frog, Rana pipiens, during metamorphosis, with special reference to tail-skin histolysis. Phys- iol. Zool., 2:255-268. Liosner, L., and Woronzowa, M. 1935 Weitere Untersuchungen iiber den Mechanismus der Per- foration der Opercularenmembran im Metamor- phose-Prozess der schwanzlosen Amphibien. Bull. Internat. Acad. Polonaise Sci. et Lettr., Ser. B, Sci. Nat., v.1935, II, pp. 231-250. Lynn, W.G. 1936 A study of the thyroid in em- bryos of Eleutherodactylus nubicola. Anat. Rec., 64:525-539. 1948 The effect of thiourea and phenyl- thiourea upon the development of Eleutherodac- tylus recordit. Biol. Bull., 94:1-15. ——, and Brambel, C. 1935 Effects of lack of iodine upon amphibian growth and metamor- phosis. Anat. Rec., 64: Suppl. 1, p. 46. , and Wachowski, H. 1951 The thyroid gland and its functions in cold-blooded verte- brates. Quart. Rev. Biol., 26:123-168. Marx, L. 1935 Bedingungen fur die Metamor- phose des Axoltls. Ergeb. Biol., 17:244-334. Mazzeschi, A. 1940 Sulle correlazione dell’ap- parato endocrino durante lo sviluppo di “Rana agilis.” Arch. Zool. Ital., 28:297-322. McMullen, E. 1938 The morphology of the aor- tic arches in four genera of plethodontid sala- manders. J. Morph., 62:559-585. Means, J.H. 1948 The Thyroid and Its Diseases. J. B. Lippincott Co., Philadelphia. Mendes, M. V. 1947 Glandulas endocrinas e hormonios nos insetos. Ann. Acad. Bras. Cienc., 19:259-275. 1948 Histology of the corpora allata of Melanoplus differentialis (Orthoptera: Salta- toria). Biol. Bull., 94:194—207. Moser, H. 1950 Beitrage zur Analyse der Thy- 661 roxinwirkung im Kaulquappenversuch. Rey. suisse. Zool., 57 (Suppl. 2): 3-144. Needham, J. 1942 Biochemistry and Morpho- genesis. Cambridge University Press, Cambridge, England. Newth, D. 1949 A contribution to the study of the fore-limb eruption in metamorphosing Anura. Proc. Zool. Soc. Lond., 119:643-657. Noble, G. 1924 The “retrograde metamorphosis” of the Sirenidae; experiments on the functional activity of the thyroid of the perennibranchs. Anat. Rec., 29:100. Novak, V. 1951 New aspects of the metamor- phosis of insects. Nature, 167:132-133. Passano,L. 1951 The X organ-sinus gland neuro- secretory system in crabs. Anat. Rec., 177:502. Pfeiffer, I. 1939 Experimental study of the func- tion of the corpora allata in the grasshopper, Me- lanoplus differentialis. J. Exp. Zool., 82:439-461. 1945a The influence of the corpora allata over the development of nymphal characters in the grasshopper, Melanoplus differentialis. Trans. Conn. Acad. Arts & Sci., 36:489-515. 1945b Effect of the corpora allata on the metabolism of adult female grasshoppers. J. Exp. Zool., 99:183-233. Pflugfelder, O. 1937 Bau, Entwicklung und Funktion der Corpora allata und cardiaca von Dixippus morosus Br. Z. Wiss. Zool., 149:477-512. 1938 Untersuchungen iiber die histolo- gischen Veranderungen und das Kernwachstum der Corpora allata von Termiten. Z. wiss. Zool., 150:451-467. 1939a Wechselwirkungen von Driisen in- nerer Sekretion bei Dixippus morosus Br. Z. wiss. Zool., Abt. A, 152:384408. 1939b Beeinflussung von Regenerations- vorgangen bei Dixippus morosus Br. durch Ex- tirpation und Transplantation der Corpora allata. Z. wiss. Zool., Abt. A, 152:158-184. 1940 Austausch verschieden alter Cor- pora allata bei Dixippus morosus Br. Z. wiss. Zool., Abt. A, 153:108-135. 1941 Tatsachen und Probleme der Hor- monforschung bei Insekten. Biol. Gen., 15:197- 235. Piepho, H. 1938 Wachstum und totale Meta- morphose an Hautimplantaten bei der Wachs- motte, Galleria mellonella L. Biol. Zentralbl., 58: 356-366. 1939 Raupenhautungen bereits verpup- pter Hautstiicke bei der Wachsmotte, Galleria mellonella L. Naturwiss., 27:301-302. 1940 Uber die Hemmung der Verpup- pung durch Corpora allata. Untersuchungen an der Wachsmotte, Galleria mellonella L. Biol. Zentralbl., 60:367-393. , 1948 Zur Frage der Bildungsorgane des Hautungswirkstoffes bei Schmetterlingen. Natur- wiss., 35:94-95, 1950 Uber die Hemmung der Falter- hautung durch Corpora allata. Untersuchungen an der Wachsmotte Galleria mellonella L. Biol. Zentralbl., 69:261-271. 662 Plagge, E. 1938 Weitere Untersuchungen iiber das Verpuppungshormon bei Schmetterlingen. Biol. Zentralbl., 58:1-12. , and Becker, E. 1938 Wirkung arteigner und artfremder Verpuppungshormone in Extrak- ten. Naturwiss., 26:430-431. Possompes, B. 1950a Implantation fractionnée de l’anneau de Weismann chez des larves perma- nentes de Calliphora erythrocephala Meig. Compt. Rend. Acad. Sci. Paris, 230:409-411. 1950b Rédle du cerveau au cours de la metamorphose de Calliphora erythrocephala Meig. Compt. Rend. Acad. Sci. Paris, 231:594— 596. Poulson, D. F. 1945 On the origin and nature of the ring gland (Weismann’s ring) of the higher Diptera. Trans. Conn. Acad. Arts & Sci., 36:449- 469. Puckett, W. 1937 X-radiation and thyroid in- duced metamorphosis in anuran larvae. J. Exp. Zool., 76:303-321. Purves, H., and Griesbach, W. 1946 The effect of thyroid administration on the thyrotropic ac- tivity of the rat pituitary. Endocrinol., 39:274— WHT Rahm, W. 1952 Die imnersecretorische Steuer- rung der postembryonalen Entwicklung von Sialis lutaria L. (Megaloptera). Rev. suisse Zool., 59:173-237. Rehm, M. 1950 Secretionsperioden neurosecre- torischer Zellen im Gehirn von Ephestia kiihniel- la. Z. Naturforsch., 5b, Heft 3, pp. 167-169. Romeis, B. 1922 Quantitative Untersuchungen uber die Wirkung von Thyroxin, Dijodtyrosin, Jodthyrin und Jodthyreoglobulin. Klin. Wschr., 1:1262-1263. Root, R., and Etkin, W. 1937 Effect of thyroxine on oxygen consumption of the toadfish. Proc. Soc. Exper. Biol. & Med., 37:174-175. Scharrer, B. 1946 The role of the corpora allata in the development of Lewcophaea maderae (Or- thoptera). Endocrinol. 38:35—-45. 1952a Hormones in insects; in The Ac- tion of Hormones in Plants and Invertebrates, edited by K. V. Thimann, pp. 125-169. Academic Press, New York. 1952b Neurosecretion, XI. Biol. Bull., 102:261-272. , and Scharrer, E. VI. Biol. Bull., 87:242-251. Schliefer, W. 1935 Die Entwicklung der Hypo- physe bei Larven von Bufo vulgaris bis zur Meta- morphose. Zool. Jahrb., 59:383-454. Schmidt, E., and Williams, C. 1949 Assay for the growth and differentiation hormone of Lepi- doptera by the method of tissue culture. Anat. Rec., 105:487. Schmieder, R. 1942 The control of metamorpho- sis in Hymenoptera. Anat. Rec., 84:514. Schreiber, G. 1934a Le disarmonie della meta- morfosi sperimentale degli anfibi e loro significato nella fisiologia dello svilluppo. Boll. Soc. Ital. Biol. Sper., 9:1211-1212. 1934b L’applicazione delle leggi d’azione 1944. Neurosecretion, METAMORPHOSIS degli ormoni alla metamorfosi degli Anuri. Archiv. Zool. Ital., 20:325-356. Schiirfeld, W. 1935 Die physiologische Bedeu- tung der Versondriisen, untersucht im Zusam- menhang mit ihrem femeren Bau. Roux’ Arch. Entw.-mech., 133:728-759. Schulze, W. 1930 Der Einfluss der inkretori- schen Driisen und des Nervensystems auf Wachstum und Differenzierung; in Handbuch der Normalen und Pathologischen Physiol., edited by A. Bethe, 76(H.1):697-806. Springer, Berlin. Schwartzbach, S., and Uhlenhuth, E. 1928 An- terior lobe substance, the thyroid stimulator, IV. Proc. Soc. Exper. Biol. & Med., 26:153-154. Schwind, J. 1933 ‘Tissue specificity at the time of metamorphosis in frog larvae. J. Exp. Zool., 66: 1-14. Sembrat, K. 1924 Recherches expérimentales sur les facteurs provoquant la métamorphose de Vintestin chez les tétards des Anoures (Pelobates fuscus Laur.). Compt. Rend. Soc. Biol., 90:894— 896. Shtern, N. 1933 Critique of reversibility of meta- morphosis in transplantation of skin from Am- blystoma to axolotl and vice versa. Bull. Acad. Sci. U.S.S.R., pp. 859-881. Singh-Pruthi, H. 1925 Studies on insect meta- morphosis, III. J. Exp. Biol., 3:1-8. Slowikowska, S. 1923 Recherches expérimen- tales sur le réle de la glande thyroide dans la métamorphose des Batraciens. Compt. Rend. Soc. Biol., 89:1396-1398. Smith, D., and Everett, G. 1943 The effect of thyroid hormone on growth rate, time of sexual differentiation and oxygen consumption in the fish, Lebistes reticulatus. J. Exp. Zool., 94:229- 240. , and Mathews, S. 1948 Parrot fish thy- roid extract and its effect upon the oxygen con- sumption in the fish, Bathystoma. Am. J. Physiol., 153:215-221. Smith, P. E. 1916 Experimental ablation of the hypophysis in the frog embryo. Science, 44:280- 282. 1920 The pigmentary growth and endo- crine disturbances induced in the anuran tadpole by the early ablation of the pars buccalis of the hypophysis. Am. Anat. Mem., No. 11. Stern, C. 1941 The growth of testes in Drosoph- ila, I. The relation between vas deferens and testis within various species. J. Exp. Zool., 87: 113-158. Swingle, W. 1918 The acceleration of metamor- phosis in frog larvae by thyroid feeding and the effects upon the alimentary tract and sex glands. J. Exp. Zool., 24:521-544. 1919 Studies on the relation of iodin to thyroid, I. J. Exp. Zool., 27:397-415. 1921 Homoplastic and heteroplastic en- docrine transplants. Anat. Rec., 20:195-196. 1922 Experiments on the metamorphosis of neotenous amphibians. J. Exp. Zool., 36:397- 421. 1923 Thyroid transplantation and anu- METAMORPHOSIS ran metamorphosis. J. Exp. Zool., 37:219-257. Taylor, A. C., and Kollros, J. J. 1946 Stages in the normal development of Rana pipiens larvae. Anat. Rec., 94:7-23. Thomsen, E. 1949 Influence of the corpus al- latum on the oxygen consumption of adult Cal- liphora erythrocephala Meig. J. Exp. Biol., 26: 137-149. 1952 Functional significance of the neu- rosecretory brain cells and the corpus cardiacum in the female blow-fly, Calliphora erythrocephala Meig. J. Exp. Biol., 29:137-172. Turner, C. 1948 General Endocrinology. W. B. Saunders Co., Philadelphia. Tutajew, G., and Philippawa, E. 1931 Uber die Wirkung der Ovarialdriisen und der Milz auf die Metamorphose der Axolotl. Z. Biol., 97:278-286. Uhlenhuth, E. 1917 A further contribution to the metamorphosis of amphibian organs. J. Exp. Zool., 24:237-302. 1919 Nature of the retarding influence of the thymus upon amphibian metamorphosis. J. Gen. Physiol., 7:305-313. 1927 Die Morphologie und Physiologie der Salamanderschilddriise, I. Roux’ Arch. Entw.- mech., 109:611-—749. Van der Jagt, E. 1929 Histolytic influence of atrophying gills of anurans during metamorpho- sis with special reference to resistance of fore- limb integument. J. Exp. Zool., 54:225-247. Vogt, M. 1942 Induktion von Metamorphose- prozessen durch implantierte Ringdriisen bei Drosophila. Roux’ Arch. Entw.-mech., 742:131- 182. 1946 Inhibitory effects of the corpora cardiaca and of the corpus allatum in Drosophila. Nature, 757:512. Vrtelowna, S. 1925 Sur la métamorphose des yeux homotransplantés chez les tétards de Pelo- bates fuscus. Compt. Rend. Soc. Biol., 92:381-383. Weber, A. 1925 Involution de la cavité péri- branchiale aprés la métamorphose des batraciens anoures. Compt. Rend. Soc. Biol., 93:410-411. Weiss, P., and Rossetti, F. 1951 Growth responses of opposite sign among different neuron types ex- posed to thyroid hormone. Proc. Nat. Acad. Sci., 37:540-556. Wigglesworth, V. 1934 The physiology of ec- dysis in Rhodnius prolixus (Hemiptera), II. Quart. J. Micr. Sci., 77:191-222. 663 1936 The function of the corpus allatum in the growth and reproduction of Rhodnius pro- lixus (Hemiptera). Quart. J. Micr. Sci., 79:91- 1292. 1940 The determination of characters at metamorphosis in Rhodnius prolixus (Hemip- tera). J. Exp. Biol., 77:201-222. 1945 Growth and form in an insect; in Essays on Growth and Form, edited by W. E. Le Gros Clark and P. B. Medawar, pp. 23-41. Oxford University Press, London. 1947 The corpus allatum and the control of metamorphosis in insects. Nature, 759:872. 1948 The functions of the corpus allatum in Rhodnius prolixus (Hemiptera). J. Exp. Biol., 25:1-14. 1951 Source of the moulting hormone in Rhodnius. Nature, 768:558. Williams, C. 1946 Physiology of insect diapause: The role of the brain in the production and ter- mination of pupal dormancy in the giant silk- worm, Platysamia cecropia. Biol. Bull., 90:234- 243. 1947 Physiology of insect diapause, II. Biol. Bull., 93:89-98. 1948a Physiology of insect diapause, III. Biol. Bull., 94:60-65. 1948b Extrinsic control of morphogenesis as illustrated in the metamorphosis of insects. Growth (Suppl.), 72:61-74. 1949 The prothoracic glands of insects in retrospect and in prospect. Biol. Bull., 97:111- 114, 1951 Biochemical mechanisms in insect growth and metamorphosis. Federation Proc., 10: 546-552. 1952 Physiology of insect diapause, IV. Biol. Bull., 703:120-138. Wintrebert, P. 1906 La métamorphose de Sala- mandra maculosa (Laur.) en dehors de la moelle et des ganglions spinaux. Compt. Rend. Soc. Biol., 60:73-74. Woitkewitsch, A. 1937 Besitzt die Rinden- und Markschicht der Nebenniere von Saugetieren metamorphogene Eigenschaften? Zool. Jahrb. Abt. Allg. Zool. u. Physiol., 58:11-22. Zavadovsky, B. 1926 Application of the axolotl metamorphosis reaction to the quantitative assay of thyroid gland hormones. Endocrinol., 10:550- 559. Section XIII REGENERATION CHAPTER. «1 Invertebrates L. G. BARTH THE process of regeneration presents many of the problems of embryonic development. In some instances the end product of the two processes is exactly the same although the form changes are vastly different. For ex- ample, a young ascidian may arise by de- velopment of the fertilized egg, through metamorphosis or by a process of regenera- tion from a part of the adult ascidian. Sim- ilarly a sponge may differentiate from a mass of cells that resulted from an aggregation of the dissociated cells of an adult sponge, or from the cleavage and development of a sin- gle egg cell. Certain differences between the two processes are worth emphasis. On the whole, regeneration entails relatively less cell division than embryonic development. Indeed in some species little or no cell division has been reported. Another difference is in the relative amount of growth. The egg increases in size and yolk is converted into protoplasm during embryonic development. In the case of regeneration the protoplasm is already synthesized. The regeneration of a hydranth in Tubularia involves little or no growth, since the regenerate differentiates directly from the cells of the stem. Thus regeneration offers a means of dissociating the processes of cell division and growth from that of dif- ferentiation. Regeneration also invites a study of the reversibility of differentiation. During em- bryonic development a progressive change from undifferentiated to differentiated cells is observed. Accompanying differentiation a restriction in the number of potencies of any particular cell is found. In the adult most if not all cells are in the differentiated state. Therefore when regeneration occurs, some cells must undergo dedifferentiation or else the new parts must come from reserve cells which retain the potencies of embryonic cells. In the highly organized but very labile adult ascidian, cutting experiments have demonstrated that no single organ is neces- sary for the regeneration of a complete in- dividual (Driesch, ’06). And thus we may ask from what cells do the nerve ganglia arise when the abdominal region regen- erates? Whatever the differences, the basic prob- lems of embryology and regeneration remain the same and in this treatment we will use the same organization of subject matter as in the analysis of embryonic development. The problems under discussion are: 1. Polarity of regenerates 2. The stimulus for regeneration 3. Potencies of cells 4. Correlative differentiation and induc- tion 5. External environment 6. Metabolism of regenerates. THE POLARITY PROBLEM IN REGENERATION Just as a developing egg possesses an in- itial polarity, so too a regenerating fragment of an adult animal may exhibit a polarity. The initial polarity of an egg is an animal- vegetal polarity while that of the regenerate may be an anteroposterior polarity or an apicobasal polarity or a distoproximal polar- ity. Fragments of Dugesia regularly form heads at the anterior cut surface of each fragment and tails from the posterior cut surface. In this form the anteroposterior 664 INVERTEBRATES polarity is relatively fixed and may be com- pared in this respect with the polarity of a sea urchin egg. On the other hand the stems of coelenterates and the stolons of ascidians exhibit a very labile polarity as regards re- generation. A fragment isolated from the stem of Tubularia may form a hydranth at the apical end and a stolon at the basal end and thus an apicobasal polarity is indicated. However, other fragments may regenerate a large hydranth at the apical end and a small hydranth at the basal end. Thus the polarity of isolated fragments is not very strong. The stolons of ascidians show a very weak polar- ity and zooids usually differentiate at both ends of a fragment. The weak initial polarity of ascidians and coelenterates has its coun- terpart in the egg of Fucus, in which polarity may be determined by a number of physical and chemical factors. The differences in the kind of polarity, exhibited by fragments of Dugesia on the one hand, and by Tubularia and ascidians on the other, may be simply a matter of intensity since short fragments of Dugesia may develop a head at both the anterior and posterior cut surfaces. Again using the sea urchin egg for comparison with Dugesia regenerates, recall that treatment of the former with potassium thiocyanate and lithium chloride results in bipolar forms (Lindahl, ’°36). The coelenterate stem and ascidian stolon possess weak polarities which may be modi- fied by experiment. Chemical and physical gradients applied along the length of the regenerates impress a strong polarity on them and unipolar forms result (Lund, ’25; Miller, 39). We thus come to regard regenerates as possessing polarities similar to those of eggs and subject to alteration by experimental procedures (Child, *42). The polarity of the regenerate is derived from the polarity of the whole organism. The fact that as many as sixteen fragments of a flatworm will regenerate with antero- posterior polarity makes very unlikely the localization of any morphogenetic substances as accounting for polarity. Therefore the polarity of forms such as Dugesia appears to be maintained chiefly by means of quan- titative differences between anterior cells and posterior cells. On the other hand, polarity in some annelids has a qualitative aspect as well as a quantitative one, since heads do not regenerate from posterior segments. Even in the case of annelids the differences under- lying polarity may be considered as purely quantitative if the additional assumption is 665 made that below a certain quantitative threshold regeneration of a head does not take place. If, for example, we assume that head regeneration requires more energy than tail regeneration, then it may be that the energy sources in the posterior segments are too low for head regeneration but still high enough for tail regeneration. The simplest hypothesis as regards polar- ity assumes the existence of a single gradient possessing a high intensity at the anterior region and a low intensity in posterior re- gions (Child, ’42). This gradient, which is expressed as graded rates of regeneration in Tubularia and as a graded head frequency in Dugesia, is detectable by a variety of methods (Child, ’42). However, the physico- chemical nature of the gradient escapes us as yet. As a physicochemical basis, Child has suggested graded rates of oxidations which would result in graded rates of energy production; Dalcq and Pasteels suggest a yolk gradient in the amphibian egg; Lin- dahl’s studies suggest for the sea urchin egg a gradient in carbohydrate metabolism for the animal gradient and a_ vegetal gradient which may be concerned with pro- tein catabolism. Sulfhydryl gradients have been demonstrated in a variety of forms. New gradients of ribonucleoproteins have been described in the amphibian egg and thus these substances become another possi- ble mediator of the gradients associated with polarity (Brachet, ’50). A somewhat more complicated gradient theory involving two factors in graded con- centration fits some of the facts of regenera- tion. Two-gradient hypotheses have been formulated in some detail by Runnstrom for the sea urchin and by Dalcq and Pasteels for the amphibian egg. As regards regeneration, Liebmann (43) has assumed a two-gradient mechanism for the facts ob- served in regeneration of an annelid. In this instance polarity is assigned to two kinds of cells present in different numbers in anterior as compared with posterior seg- ments. These two kinds of cells determine head regeneration and tail regeneration, re- spectively. Since these cells, “head” eleocytes and “tail” eleocytes, respectively, must mi- grate polarly it is clear that there must be a basic polarity in addition to the eleocytes. The one-gradient theory of polarity carries with it the corollary of dominance by which high regions inhibit low regions in the gradient. The nature of this inhibition will be considered later. The two-gradient theory involves the concept of antagonism and bal- 666 ance. This idea is best illustrated in the experiments dealing with recombination of animal with vegetal blastomeres in the sea urchin egg (Horstadius, ’35). In any case, one of the real problems for future work is that of deriving multiple differentiations from one or even two gra- dients. It is evident that, if the presumptive head of a regenerate contains the same sub- stance or substances as the presumptive tail, some kind of quantitative thresholds for head and tail formation must be assumed. Since the regenerate already contains a va- riety of cell types which may possibly react in the gradient system by differentiating into a variety of structures, the problem of deriving multiplicity from a single gradient may be easier than first appears. THE STIMULUS The egg responds to a number of stimuli by beginning to develop. One theory of stimulation states that detoxification of the unfertilized egg results. The egg is pictured as being inhibited by toxic products which escape or are destroyed at fertilization. What stimulates tissues of the adult to differen- tiate? Cutting and injury serve as the im- mediate stimulus and the problem becomes one of determining just what cutting does to tissues. In this connection, consider the possibility that it is not the cutting or injury per se but rather the isolation of tissues from cor- relative influences of adjacent tissues that is the effective stimulus for differentiation. For example, the normal growth form of the ascidians makes this possibility very probable (Berrill, 51). Perophora reproduces by budding from the stolon. The stolon grows along a surface in a straight line away from the zooid. At fairly regular space intervals along this growing stolon, buds appear and differentiate into zooids. This observation suggests that the tissues of the growing stolon are inhibited from forming buds until the stolon reaches a length at which its tissues are below the threshold for inhibition by the zooid. A bud then forms which develops into a zooid. The stolon continues growing and this new de- veloping zooid exercises an inhibition over the tissues of the stolon until the latter reach a certain distance, when again the inhibition falls below a threshold and a second bud forms. The important fact is that no cutting or external stimulus needs to be applied for the formation of a bud. REGENERATION Therefore when a section of the stolon is cut and removed from a colony the major effect may be the removal of the inhibition exercised by the zooid and not the stimulus of cutting or injury. In this connection, recall Huxley’s experiments (Huxley, ’26) in which the equilibrium, stolon = zooid, was controlled by the addition of external agents such as potassium cyanide and po- tassium chloride. On the other hand the differentiation of zooids from the stolon always occurs at the cut ends and not from the middle of an excised section. This observation shows that the cut end is stimulated in some way. Thus at least two dissociable factors operate in the stimulation of regeneration: (1) correl- ative factors, i.e., the removal of inhibitory regions, and (2) a direct stimulus caused by cutting or injury. The correlative factor will be discussed later and at this point we will proceed to analyze the stimulus of cutting. A number of observations demonstrate conclusively that cutting tissues does not suffice for regeneration unless the wound remains open to the external medium. In Tubularia a hydranth fails to regenerate at a cut end if the perisarc covers the surface of the wound (Barth, *40). Likewise in the ascidian stolon when the cut is made with a blunt scissors the tunic is pinched together and a zooid fails to form. The perisarc and tunic act as barriers to regeneration (Goldin, 48). Furthermore, a cut surface is not necessary for regeneration of a hydranth in Tubularia. If the perisarc is removed from the middle of an isolated piece of stem, regeneration of a hydranth will occur from the intact uncut stem (Zwilling, 739). The general conclusion from these ex- periments and observations is that a layer at the surface of tissues prevents regenera- tion. In Tubularia and ascidians the layer is identified as perisarc and tunic, respec- tively, but in other forms we may assume the presence of some similar limiting layer. In amphibians this layer is a component of the skin (Godlewski, ’28; Rose, *44) which inhibits regeneration of limb and tail. How does this external surface layer act to prevent regeneration? Experiments on Tubularia with vital dyes show that neutral red penetrates rapidly into cells at an ex- posed surface but very slowly into cells covered with the perisarc. Thus the perisarc acts as a barrier to free diffusion of neutral red and perhaps other substances. Now it is known that the excretory products of Tubu- laria inhibit regeneration completely. A glass INVERTEBRATES tube containing a bubble of oxygen placed over the end of the stem inhibits regenera- tion (Rose and Rose, ’41). This inhibition is not caused by a lack of oxygen, since oxygen is supplied and therefore inhibition must be caused by the accumulation of excretory substances within the glass tube. In the case of the ascidian stolon, either carbon dioxide or urea or uric acid will inhibit zooid formation but allow stolons to grow. In summary, we may conclude that in some organisms, at any rate, a surface layer prevents free diffusion of substances which inhibit regeneration. When the surface layer is removed by cutting, then carbon dioxide, urea and uric acid escape more readily and the exposed region is able to undergo regen- eration. POTENCIES OF CELLS During early embryonic development an important question is, In how many ways can the various cells differentiate? Similarly we may ask whether the cells participating in regeneration have wide potencies or lim- ited potencies. This problem is difficult to solve and perhaps we can do little more here than to state the problem. In the first place we may visualize a num- ber of possibilities as regards the potencies of the regenerating cells. 1. Imagine that all of the cells at the cut surface proliferate and form a mass of un- differentiated cells with wide potencies. Then some inductive influence from adjacent structures must induce differentiation. 2. The cells at the cut surface dediffer- entiate morphologically but retain their specificity as regards cell types. Then a muscle cell which dedifferentiates must re- differentiate into a muscle cell. Cells retain their limited potencies. 3. Cells at the cut surface do not prolifer- ate but rather transform directly into the regenerating structures. This process may or may not involve changes in cell type. Some cells must have a number of potencies to supply missing structures. 4. The cells at the cut surface take little part in regeneration and the main source is from undifferentiated cells which migrate to the site of the wound. Missing structures would thus differentiate from reserve cells (embryonic cells, neoblasts, formative cells, eleocytes) which are present throughout adult tissues and have wide potencies. The whole problem of the origin of the cells which differentiate during regenera- 667 tion and the very important implications in regard to a possible reversal of biochem- ical differentiation will be a challenge to investigators for some time. A_ histological dedifferentiation is clearly shown by Hux- ley’s histological studies on ascidians. How are we going to trace cells during regener- ation? Can we trace a nerve cell from the adult ascidian zooid into the stolon during regression and then trace this same cell back into the new zooid which forms from this stolon? Does this nerve cell redifferentiate into a nerve ganglion cell or may it form an intestinal cell? Are cell types immutable? For those who maintain the affirmative to this latter question I would point out that the atom was once considered immutable. As to those who take the negative, I must remind them that conclusive evidence of a major change in cell type is still wanting. Certain facts are clear and must be con- sidered before reaching a judgment on this question. In forms such as Tubularia the hydranth develops directly from the coeno- sarc near the cut surface. There is no growth and little cell division. A section of the stem about 2 mm. in length transforms directly and quickly (24 hours) into a hydranth. There can be no question here of reserve cells migrating in from the rest of the stem and differentiating into a hydranth. Indeed, a section of the stem 2 mm. long will trans- form completely into a bipolar form with a hydranth at each end. Nor is there any proliferation of reserve cells accompanied by a degeneration of cells of the stem. There- fore a cell which is part of the stem is converted directly into a cell forming part of a tentacle. During the regression of an ascidian zooid into the stolon, the latter becomes packed with cells which dissociate from the tissues of the zooid (Huxley, ’26). These cells flow through the circulatory system. They are alive as witnessed by observation of the living stolon and by histological examina- tion. Therefore nerve, heart and intestinal cells pass from the regressing zovid into the stolon. This stolon may then form zooids when cut into short sections. What happens to the nerve, heart and intestinal cells when the new zooid forms? Do they undergo cytolysis or do they take part in the forma- tion of the zooid? There is no evidence of extensive cytolysis and thus the cells prob- ably enter the new zooid. If they do, will a former nerve cell enter the intestinal tract and become a secretory cell or must it rediffer- entiate into a nerve cell? This question can- 668 not be answered definitely. Recall, however, that isolated parts of the adult zooid will regress and form a mass of morphologically dedifferentiated cells. This mass can then form a complete zooid of reduced size. In this instance when a region lacking the nerve ganglion is isolated, where does the new nerve ganglion come from? I would like to urge a study of ascidian cell types and their behavior during regression of the zooids, during dedifferentiation of the stolon and during regeneration of the parts of the zooid. This material is relatively easy to handle and changes are relatively rapid, an important consideration where so much ex- ploratory research still remains. Consulta- tion of the papers of Driesch (’06), Spek (27), Huxley (26), Deviney (34), Berrill and Cohen (736), and Goldin (’48) will be valuable in this connection. Another form which appears admirable for studies on potencies of cells is the sponge. Here cells may be dissociated and they will aggregate and reconstitute a sponge. The nature of the cells which take part in the formation of the new sponge is still a matter for investigation (Galtsoff, ’25; Wilson and Penney, °30; Penney, °33; de Laubenfels, 34). However, since the adult sponge may be broken up into individual cells, there may be a means of separating the various cell types and examining each type for its ability to differentiate. If the various cell types differ by some property such as spe- cific gravity, then centrifuging the dissoci- ated cells in sugar solutions would separate the cells into layers. CORRELATIVE DIFFERENTIATION As in embryonic development the differ- entiation of tissues during regeneration de- pends on factors located in adjacent tissues. The organizer phenomenon has been inves- tigated in flatworms (Santos, ’31; Miller, 38) and in coelenterates (Child, ’29; Li and Yao, ’45). In Dugesia the head region acts as an organizer inducing changes in the surrounding tissues. A small fragment of the head region transplanted to the body region will induce an outgrowth that de- velops into a head and in addition induces the host tissue to form a pharynx. What are the similarities between the organizer in Dugesia and the amphibian organizer? 1. A small piece of transplanted tissue induces a large part of the host to differentiate. 2. Dugesia organizer is not species specific and heteroplastic induction occurs. 3. The organ- REGENERATION izer is localized. In planarians it is restricted to the head region while in an amphibian gastrula only cells from the dorsal and lat- eral lips of the blastopore will induce. In flatworms and annelids the nervous system may play a special role in regenera- tion. The nerve tissue appears to act as an organizer for head structures. However, re- generation of heads will still occur in absence of nerve cord at the cut surface. These ob- servations are in keeping with the properties of the amphibian organizer phenomenon where, although nerve tissue is a good organ- izer, the chorda-mesoderm is the primary organizer. For a discussion of the role of the nervous system in various invertebrates, see Child (42, pp. 338-341). Another expression of correlative differ- entiation is found in the regeneration of the hydranth in Tubularia. If a piece of stem is isolated, the apical (distal) cut end differ- entiates into a hydranth while the basal (proximal) cut end forms a stolon. The apical end exercises an inhibition over the regeneration of the basal end, a fact which can easily be demonstrated by preventing the apical end from regenerating with the result that the basal end then regenerates a hydranth instead of a stolon. Thus differ- entiation of the basal end into stolon or hydranth is correlated with the presence or absence of differentiation at the apical end. The mechanism by which the regener- ating apical end influences basal regenera- tion has been investigated from two points of view, that of electrical differences in potential and that of competition. 1. Correlations have been shown to exist between the electrical differences in potential and the behavior of the two cut surfaces of the hydroid stem (see Child, *42, and Lund and others, ’47, for literature). If electrodes are placed on the two ends of the stem a P.D. can be measured. If a constant electric current is passed through sea water contain- ing stems, regeneration is inhibited at one pole but normal hydranths regenerate at the opposite pole. Thus an electric current may determine polarity. The question as to whether the electrical potential difference measured between the two ends gives rise to a current within the stem of sufficient intensity to account for the inhibition exer- cised by the distal end over the proximal end is still a matter for investigation. The possibility that P.D.’s are effective in other forms such as the earthworm has been more recently studied by Moment (746). The general hypothesis of the E.D.P. as INVERTEBRATES controlling cellular differentiation is an at- tractive one but beset with great difficulties which should be a challenge to investigators. The electrical difference in potential must have its origin in concentration differences Region A O2 External Via le NL Circulation Internal I Surface layer removed, Surface layer impermeable to I Regeneration inhibited 669 chemical differences and without signifi- cance in regeneration? For a complete dis- cussion of the problem see Lund (747). 2. Quite a different hypothesis is that of the existence of competition among groups Electrical Potential Difference Region B O2 ee aes External Circulation Circulation I diffuses out, regeneration proceeds ROLE OF INHIBITOR [E) yt [Ty coe [ET y- [Substrates], = Hydranth Hydranth forms in presence of [I], because enzymes are at [E]y. [E}, + (1y] == [EN], [Substrates], <_— __ Hydranth Hydranth fails to form in presence of [I], because at low concentration [E],, enzyme is removed. ROLE OF SUBSTRATES [Ely [Substrates], a Hydranth forms because: [EJy « [Substrates], =K,y which is above threshold for hydranth regeneration. Hydranth fh [Substrates], _——> Hydranth Hydranth fails to form because: [E], -[Substrates], =K, which is below threshold for hydranth regeneration. EFFECT OF BLOCKING INTERNAL CIRCULATION Rapid utilization of substrates by high concentration of [E], lowers concen- tration to [S]y-;. However, [E]y- [Sly-1= Ky-4 which is still well above threshold for regeneration. When Region A is no longer utilizing sub- strates from Region B, the concentration of S rises to [S]y+1- Then [E],-(Sly+4=Ko which is above threshold for regeneration. Fig. 230. arising from differences in either the kind of metabolism or the rate of metabolism. What is the primary difference? May not the differences in metabolism themselves be sufficient to account for the polarity phe- nomena and control of regeneration without any intervention of electrical potentials? Are these potentials merely the consequence of of cells such that some groups more ad- equately endowed or situated are able to inhibit cells living under less favorable conditions (Barth, °38; Spiegelman, °45). The general assumption is made that differ- entiating cells require energy for structural changes and need materials for synthesis of chemical constituents. In addition, differ- 670 entiation results in the formation of excre- tory products such as carbon dioxide and nitrogenous compounds and these substances have inhibitory effects. Therefore any two groups of cells potentially able to regenerate may be competing with each other by uti- REGENERATION ganism or by an environmental difference imposed from without. The situation is out- lined in Figure 230. Numerous simple experiments are consist- ent with the competition hypothesis. In Tubularia it is possible to select pieces of CONTROL OF REGENERATION BY INHIBITION By contact Io Inhibitor accumulates at surface of contact and hydranth develops at free surface. By mutual inhibition Zooid Stolon Inhibitor accumulates in center, C, where four sources are in close proximity. By COz Stolon reservoir COs, urea, uric acid permit stolon growth but inhibit zooid formation. Fig. 231. lizing common substrates and by excreting common inhibitory substances. Now, if the two groups of cells, A and B, differ initially by some quantitative factor in their chemical make-up, then A may be better equipped to utilize substrates and resist inhibitory com- pounds as compared with B. Therefore, A may inhibit the regeneration of B and differ- ent structures will result at A and B. The initial difference between A and B would be derived from a gradient in the whole or- stem of such a length and age that regenera- tion always occurs at the distal (apical) end but not at the proximal (basal) end. It is only necessary to make a ligature at the middle of such stems and then both ends form hydranths. What are the effects of tying a ligature? The tissue is constricted and separated at the ligature so that circu- lation between the two ends is stopped. Neither inhibitory substances nor substrates can pass from the apical end to the basal INVERTEBRATES end. Of course any electric current flowing through the stem would also be broken by the ligature since there is no connection between the ends save through the non- cellular perisarc. In an effort to see whether the circulation or the cellular connections transmit the inhibition, two kinds of experi- ments were carried out. 1. Circulation was blocked by means of an oil droplet or an air bubble (Fig. 230). The basal end then was freed of inhibition and regenerated a hy- dranth. 2. A glass tube was inserted into the coelenteron and the tissue ligatured around this tube. The circulatory fluid would pass through the tube but the cellular con- nections between the two ends was broken. Under these conditions the basal end was inhibited. Evidently the inhibition is trans- ported through the circulation (Barth, ’40; Rose and Rose, ’41). Continuing with the analysis, now using ascidians, it can be shown that excretory products inhibit regeneration of zooids from pieces of the stolon (Barth and Barth, ’50). Four to six pieces were placed with one of their ends in close proximity, but not touch- ing, in the manner of spokes in a wheel (Fig. 231). Thus one end of each piece was exposed to the accumulated excretory prod- ucts of all the others while the opposite end was not so exposed. The ends in close proximity were inhibited and zooids did not form. The opposite free ends developed zooids. Next carbon dioxide was applied in graded concentration to one end of the piece and the end at the highest concentra- tion of carbon dioxide grew out as a stolon while the end at the lowest concentration formed a zooid (Fig. 231). Urea and uric acid also inhibit zooid formation in a like manner. Analyzing these simple observations we see that substances pass through the circulation between one end, A, of a regenerate to the other end, B, and that the inhibition of end B is released by isolating it from end A. A therefore either (1) takes away from B sub- stances necessary for the regeneration of a particular structure C; or (2) A produces inhibitory products, 7, in such amounts that the concentration of J is too high for the regeneration of C at B. Therefore some other structure, D, regenerates at B instead of C. We have advanced no argument for (1). Indeed the arguments for this possibility are as yet too tenuous (Fig. 230) since sub- strate concentration has not been controlled. However, we have direct evidence that ex- cretory products do inhibit the differentia- 671 tion of C but permit the differentiation of D. These observations and experiments make the competition hypothesis attractive for future studies, but it seems only reasonable to investigate the effects of the above ex- perimental treatments on electrical potential differences. ENVIRONMENTAL FACTORS For regeneration to proceed the external medium must meet the usual requirements of temperature adjustment, hydrogen ion con- trol, gaseous exchange, salt balance and osmotic pressure. In addition, in marine organisms especially, the rate of circulation of the sea water and the population density are important factors. The presence or ab- sence of a surface for attachment is a factor in sessile animals. In general the require- ments for regeneration in regard to en- vironmental factors are more rigid than for simple maintenance. For example, Tu- bularia stems remain healthy at low oxygen tension, but fail to regenerate a hydranth. In certain ascidians, stolons will maintain themselves in standing sea water at 28° C., but regeneration of a zooid does not occur unless the temperature is reduced to 23° C. (Jaeger and Barth, ’48). Temperature is a very important factor in marine organisms during the summer months when much of the work on regener- ation is done. Laboratory temperatures are often too high for regeneration and a lower- ing of the temperature by a few degrees centigrade is then necessary. Temperature effects are related to rate of circulation of sea water in such a way that regeneration will occur at higher temperatures in running sea water as compared with standing sea water. Certain ascidian stolons will form zooids at 28° C. in running sea water but not in standing sea water. Hydroid stems show a similar behavior. A reasonable in- terpretation of the above observations is that inhibitory products accumulate in the cells faster in standing sea water, and pos- sibly the oxygen tension within the cell is lower, than in circulating sea water. If this is so then lowering of the temperature will, by lowering the rate of metabolism, decrease the rate of accumulation of excretory prod- ucts and will also increase the internal oxygen tension. Population density. The numbers of indi- vidual regenerating parts, or perhaps better the total mass of cells per unit volume of external medium, is a factor in regeneration. 672 A mutual inhibition of regeneration is ob- tained when many stems or stolons are present as compared with few. Widely sep- arated parts regenerate better than crowded parts. This phenomenon is best explained as a more rapid accumulation of inhibitory substances by many as compared with few regenerates. The optimum number will de- pend on temperature and rate of circulation of sea water and on proximity of regener- ates as discussed above. Gaseous exchange. It is seen from the fore- going that temperature, circulation and pop- ulation density are related to gaseous exchange and excretions. Oxygen stimulates regeneration in Tubularia and carbon diox- ide inhibits it (Goldin, °42). Increasing hydrogen-ion concentration also inhibits. In ascidians, carbon dioxide, urea and uric acid inhibit. The inhibitory effect on regeneration in Tubularia of accumulated carbon dioxide at high population densities is shown by the use of Warburg manometers. With large numbers of stems of Tubularia in the War- burg flasks and no potassium hydroxide to absorb carbon dioxide, no regeneration occurs. The same number of stems in a flask with potassium hydroxide regenerate com- pletely. Since the respiratory quotient dur- ing regeneration is 1.0 it may be that carbon dioxide and ammonia are the chief excretory products. Salt balance. Calcium salts are particularly necessary for cell aggregation of the disso- ciated cells of the sponge (Galtsoff, ’25). In absence of the Ca** ion, dissociated cells do not aggregate and therefore no regeneration is possible. Ca** no doubt acts on the cell surface to maintain the intercellular matrix as it does in the dividing egg. Amino acids. The specific quantitative and qualitative effects of various amino acids have been investigated for a number of years by Hammett (’43) and his co-workers on regeneration of hydroids and growth of other forms. The process of regeneration is broken up into a number of individual processes and the papers of this school should be consulted for details. Radiations often inhibit regeneration with- out interfering with maintenance of life. METABOLISM AND REGENERATION Here is an attractive new field for investi- gation. The ultramicrochemical and ultra- micromanometric methods of the Linder- strém-Lang and Heinz Holter school plus the methods of Kirk and his co-workers REGENERATION have made possible exact studies on masses of tissue of the order of magnitude provided by regenerates. The cytochemical methods for phosphatases and nucleoproteins offer the opportunity for studying phosphate trans- fer and the role of nucleic acid during re- generation. Some progress has been made and has been reviewed by Brachet (’50). The analysis of the problem appears to be as follows: Energy is required for the differentiation of cells. The initial source of this energy comes from cellular oxida- tions. The free energy of cellular oxidations is transferred and conserved in an energy- rich phosphate bond. Compounds containing the energy-rich phosphate bond transfer phosphate to proteins and the free energy of the splitting of an energy-rich bond is uti- lized in performing work. The work may be lifting a weight, as in muscular contrac- tion, or the work may consist in change in cell shape and the chemical constitution of the cell. In this analysis the critical linkage between the oxidations and the performance of work appears to be through phosphate compounds and the controlling factors may be the enzymes which split and transfer phosphate. In this connection Jaeger and Barth (48) have shown that the undifferentiated stolon cells of ascidians have no water-extractable apyrases while the zooids do. As the stolon differentiates into a zooid the apyrases ap- pear. The initial step in any differentiation may be the formation of a new apyrase which transfers the energy of oxidations to the specific work required for the formation of a specific cell type. REFERENCES Barth, L. G. 1938 Quantitative studies of the factors governing the rate of regeneration in Tubularia. Biol. Bull., 74:155-177. 1940 The process of regeneration in hydroids. Biol. Rev., 75:405-20. , and Barth, Lucena J. 1950 The control of differentiation by external factors. Anat. Rec., 108:587. Berrill, N. J. 1951 Regeneration and budding in tunicates. Biol. Rev., 26:456-475. , and Cohen, A. 1936 Regeneration in Clavellina lepadiformis. J. Exp. Biol., 13:352- 362. Brachet, J. 1950 Chemical Embryology. Inter- science Publishers, New York. Child, C. M. 1929 Lateral grafts and incisions as organizers in the hydroid Corymorpha. Physiol. Zool., 2:342. 1942 Patterns and Problems of Develop- ment. University of Chicago Press, Chicago. INVERTEBRATES Deviney, E. M. 1934 The behavior of isolated pieces of Ascidian (Perophora viridis) stolon as compared with ordinary budding. J. Elisha Mitch- ell Sci. Soc., 49:185-224. Driesch, H. 1906 Skizzen zur Restitutionslehre. Roux’ Arch. Entw.-mech., 20:21-29. Galtsoff, P. S. 1925 Regeneration after dissocia- tion (an experimental study of sponges). J. Exp. Zool., 42:183-256. Godlewski, E. 1928 Untersuchungen itiber Aus- lésung und Hemmung der Regeneration beim Axolotl. Roux’ Arch. Entw.-mech., 774:108-143. Goldin, A. 1942 A quantitative study of the in- terrelationship of oxygen and hydrogen ion con- centration in influencing Tubularia regeneration. Biol. Bull., 82:340-46. 1948 Regeneration in Perophora viridis. Biol. Bull., 94:184-193. Hammett, F.S. 1943 The role of the amino acids and nucleic acid components in developmental growth. Growth, 7:331-399. Horstadius, S. 1935 Uber die Determination im Verlaufe der Eiachse bei Seeigeln. Pubbl. Staz. Zool. Napoli, 74:251-429. Huxley, J. S. 1926 Studies in dedifferentiation. VI. Reduction phenomena in Clavellina lepadi- formis. Pubbl. Staz. Zool. Napoli, 7:1—-36. Jaeger, Lucena, and Barth, L.G. 1948 The rela- tion of apyrase activity to differentiation in an ascidian. Jour. Cell. & Comp. Physiol., 32:319- 330. de Laubenfels, M. W. 1934 Physiology and mor- phology of Porifera exemplified by Jotrochota birotulata Higgin. Tortugas Lab., Carnegie Inst. Washington., 28:37-66. Li, H. P., and Yao, T. 1945 Studies on the or- ganizer problem in Pelmatohydra oligactis. J. Exp. Biol., 27:155-160. Liebmann, E. 1943 New light on regeneration of Eisenia foetida (Sav.). J. Morph., 73:583-610. 673 Lindahl, P. E. 1936 Zur Kenntnis der physiolo- gischen Grundlagen der Determination im Seei- gelkeim. Acta Zool., 17:179-366. Lund, E. J. 1925 Experimental control of or- ganic polarity by the electrical current. J. Exp. Zool., 39:357-380. , and others 1947 Bioelectric Fields and Growth. University of Texas Press, Austin. Miller, J. A. 1938 Studies on heteroplastic trans- plantation in triclads. Physiol. Zool., 17:214—-247. 1939 Experiments on polarity determina- tion in Tubularia regenerates. Abstr. Amer. Soc. Zool., Anat. Rec., 75, 4, Suppl.: 38-39. Moment, G. B. 1946 A study of growth limita- tion in earthworms. J. Exp. Zool., 103:487-506. Penny, James T. 1933 Reduction and regenera- tion in fresh water sponges (Spongilla discoides). J. Exp. Zool., 65:475-497. Rose, S. M. 1944 Methods of initiating limb re- generation in adult Anura. J. Exp. Zool., 95:149- 170: , and Rose, F. C. 1941 The role of a cut surface in Tubularia regeneration. Physiol. Zool., 14:328-343. Santos, F. 1931 Studies on transplantation in Planaria. Physiol. Zool., 4:111-164. Spek, J. 1927 Uber die Winterknospenentwick- lung, Regeneration und Reduktion bei Clavellina lepadiformis und die Bedeutung besonderer “‘om- nipotenter” Zellelemente fiir diese Vorgange. Roux’ Arch. Entw.-mech., 777:119-172. Spiegelman, S. 1945 Physiological competi- tion as a regulatory mechanism in morphogenesis. Quart. Rev. Biol., 20:121-146. Wilson, H. V., and Penney, James T. 1930 The regeneration of sponges (Microciona) from dis- sociated cells. J. Exp. Zool., 56:73-147. Zwilling, E. 1939 The effect of removal of the perisarc on regeneration in Tubularia crocea. Biol. Bull., 76:90-103. Section XTIT CHAPTER 2 Vertebrates J. S. NICHOLAS INTRODUCTION REGENERATION as defined by Nussbaum (1886) is limited to the replacement of lost parts by an individual during either its lar- val or its adult existence. This would mean that a structure must have completed its morphological differentiation in order that we may speak of its replacement as regener- ation; otherwise it is replaced by the organ- ization of an undifferentiated cell mass, the cells of which are, in a non-specialized sense, formative and, in another, equipotential in the capacity for forming parts which are quite different from their original prospective sig- nificance. When one arbitrarily states that an embryo has no capacity for regeneration, as the definition above would indicate, it is simply an expression of a limitation to which we cannot now rigidly adhere. An undetermined embryonic structure does not regenerate because its prospective signifi- cance has not yet been realized; it therefore retains its prospective potency for the pro- duction of structures with which it would ordinarily not be concerned. A restatement of the original usage of the term regenera- tion would imply simply that regeneration is the replacement of lost parts, no matter when in the history of the individual this occurs. Such usage would make unnecessary the various loosely used terms such as res- titution, reconstitution, replacement and reg- ulation, as well as those peculiarly signifi- cant to undetermined embryonic structures such as super-regeneration or post-regenera- tion. It is evident from this short discussion that a definition of regeneration might be laid down arbitrarily and therefore exactly, but qualifications of the definition increase in direct proportion to the study of the organ- ism in its various phases. There is a con- tinuity of the regenerative process—the discontinuous elements are postulates not actual. With this limitation in mind it is clear that there must be a continuous overlap between the processes to which we refer as embryonic, differentiative and regenerative. They may be considered essentially as parts of the continuum which is constantly present as a potential process and which is most significantly evident in those organisms which completely replace parts lost by accident and design; they also are less spectacular but probably more important in the gen- eralized process of growth and repair. The first we might catalogue as reconstitutive regeneration, the second as_ physiological regeneration. These in turn overlap as proc- esses within the organism; e.g., the pan- creas, a doubly secreting structure, shows both types of regeneration. The multiple acini of the ducts are constantly changing both in their cellular secretion and in the cells themselves. In the ductless part of the pancreas there is total destruction of huge areas of islet cells, a period of quiescence and later reestablishment of function. In general, the entire process might be said to be limited, so far as reconstitutive regeneration is concerned, to active replace- ment of peripheral parts, either on the out- side of the body where more than one tissue is usually concerned, or internally where a part of an organ is reconstituted. The physiological regenerative process is much more continuous in the vertebrates than is the reconstitutive, and quantitatively ac- counts for the immediate repair of large areas of tissue reconstituted after their reg- ular and sometimes cyclic loss. In addition to a high degree of physio- logical regeneration, vertebrates possess ad- ditional regulative mechanisms in a physio- logical adaptation, accompanied, of course, by morphological changes, termed ‘“‘compen- satory regulation.” Under this heading come hypertrophy and hyperplasia, which when combined affect the regulation of other portions of the system. In addition to the compensatory regula- tory processes, there is constant replacement of integumentary parts in nearly all forms. 674 VERTEBRATES This may show superficially as an inter- mittent type, as in the amphibians and snakes. The process, however, when analyzed is a continuum with intervening preparatory processes. Birds and mammals are constantly repairing integumentary structures; glands and teeth also come under the replacement category, as do the linings of the intestinal and reproductive tracts. In the face of all these facts we certainly cannot regard the higher vertebrates as lacking in regenerative capacity. AMPHIOXUS AND PISCINE FORMS Amphioxus (Biberhofer, 06) will regen- erate anteriorly. If a cut is made across the funnel just anterior to the anteriormost ex- tent of the notochord, the funnel and oral cirri reform and the anterior end is recon- stituted. No posterior regeneration occurs. This work needs systematic repetition and should be performed with a sterile technique, for many of the animals died of infection after operation. Little work has been done in either cyclo- stomes or elasmobranchs. The horny corneal teeth of the cyclostomes are known to re- generate and the ovary likewise undergoes striking cyclic replacement. Studnicka (712) has given the most complete study of one case of an animal injured in nature. The complete tail end was missing. He traced the amount of regeneration by the differ- ences in the pigmentation of the epithelium, since the line of injury was marked by a very light pigment. The entire series of tissues reconstituted a rounded mass in which the nervous system extended to a point just proximal to the tip. The muscles were re- constituted as were the bony elements of the vertebrae. Larval lampreys have been found with two tails, indicating that either in the egg or soon thereafter a process of redupli- cation occurred. The teleosts are known to have some power of regeneration in early stages, and in later stages possess the capacity to re- generate some organ systems either partially or completely. The first recorded work on these goes back to Spallanzani (1768), who recorded that when a fin was completely removed it did not regenerate. Broussonet (1786) discovered that if a part of the fin is removed it will regenerate the distal parts. This was again demonstrated by Philippeaux (1867). Morgan (’00, ’02) used several kinds of fishes and worked on the tail fin. The killi- 675 fish, scup, and goldfish all gave approxi- mately the same result. If the tail of Fundu- lus heteroclitus is cut obliquely with the greater excision ventral, new material at first appears uniformly along the whole mar- gin of the injury, but the growth and prolif- eration tend to be faster at the basal part. Where the angle of obliquity is reversed, so is the growth. When a double cut is made, the proximal tissues grow faster causing a rounding out at the base and a morphallaxis distally (cf. Nabrit, °38). The scup (Steno- pus chrysops) normally has a bifid tail. When a square cut is made, the swallow- tail is regenerated by a different zone of BX Fig. 232. Regeneration of the tail fin in Fundulus. A, After injury by cutting; B, after injury through a distal and more proximal cutting. (From Morgan, ’00, ’02.) growth in two proliferative points, one dorsal and one ventral. The bifid condition is in- dicated before the level of the original bi- furcation has been reached by the regenerate. The rate of maximum proliferation is limited to those parts of the regenerate which give the regenerate the form of the original part. This type of form regulation is difficult to explain. Why does it form a specific type of tail fin and the one tail fin so characteristic of the organism? It is hard to define this as a genetic result; the materials here are not plastic, as in the embryo, and an organizer activity is an untenable explanation. The thin rays of the fin are really dermal bones, as shown in Harrison’s (1893) paper on the development of the fins. If the cut is made below the basal plate, there is no re- generation; distal to it, there is. When the tail is cut squarely the rays regenerate along their old axis. When the cut is oblique, the rays orient perpendicularly to the cut. After a time they tend to regulate to a radiate condition. Barfurth (1899) stated this origi- nally, “the axis of the tissue in a regenerating structure is at first perpendicular to the cut.” This is sometimes termed Barfurth’s law. Beigel (710) gave a very complete account 676 of the histological processes. The new rays are formed by osteoblasts which form at the tip of the cut rays, and the horny rays, which in development are simultaneous with the dermal rays, are formed later. While the experiments have been per- formed usually on the tail fins or dorsal fins, the anal and paired fins give the same result. Scott (07) showed that in young fish the capacity for regeneration is greater and the time during which it is completed REGENERATION faster and complete the process much more quickly both actually and relatively. Sum- ner (04) had shown that the embryo would repair parts of the neural plate after injury, as would also the shield. Hoadley (28) later repeated this experiment and likewise re- moved blastomeres with no defect produced in the later embryo. Lewis (712) found that midline injuries in the plate tended to leave definite defects in the resulting nervous system, which finding Hoadley later corrob- SCHEMA OF RECONSTITUTION | PLOTTED AGAINST STAGE OF DEVELOPMENT INJURY MATERIAL REMOVED CELLULAR MULTIPLICATION WITHOUT FIXED LOCALIZATION OF POTENCY PROLIFERA- TION uw ° WwW Ww a oO WwW a a a z fo} - « ° a °o a a BASED ON QUANTITY OF 2CELLED S8CELLED 4CELLED I6CELLED OPPENHEIMER 3- 4 -S — 6 STAGES IN FUNDULUS SPECIFIC ORGAN TRUE FORMATION |REGENERATION FIXITY REPAIR OF PARTS |LIKE FROM LIKE BLASTULA |GASTRULA SHIELD/ PREMOTILE RING Fig. 233. is less than in older fish. Nusbaum and Sidoriak (00) studied trout fry, in which regeneration of the entire posterior end of the body will occur if cuts are made anterior to the anus. This process is complete and regulatory in detail even to the openings of the ducts into the cloaca. Duncker (05, ?07) showed in lophobranchs that skeleton, muscle, fin and dermal rays can be regener- ated in remarkably constant form. Morrill (06) worked on the operculum and reported negatively. Beigel (10) found that regener- ation was complete but extended over a very long period. The regeneration of skin and scales was described in the papers of Fraisse (1885) and Beigel (12). The central nervous system was most thoroughly treated in Hooker’s (30, °32) experiments performed on Lebistes, in which the spinal cord regenerates with as great or greater completeness than is found in frog tadpoles (Hooker, ’25) and in much the same manner. If young fry regenerated fast, younger animals should regenerate orated. Oppenheimer (’36) and Sumner (’04) both got perfect embryos, Sumner after in- juries to the shield in the region of the blastopore, while Oppenheimer removed all of the gastrular dorsal lip and still got a normal embryo. Some of this information has been added since the original experi- ments were planned, and the exceedingly critical and precise work of Oppenheimer (36) has given a practical cell lineage which shows what some of these cells nor- mally do in this makeup of the embryo. Nicholas (27) and Nicholas and Oppen- heimer (42) removed the ear, eye, fin, and nerve cord at various levels, and finally the tail. These operations were performed on non-motile forms at the 12-somite stage. None of the structures regenerated. At this particular stage and for a short while there- after, there is no demonstrable regenerative potency; the organism is mosaic and any part removed is missing, and no initiation of the reconstitutive process occurs (Fig. 233). After the later differentiation of definite VERTEBRATES structures, the capacity for regeneration again appears. The nervous system which regenerates in the early shield stages fails to regenerate in later formative stages. AMPHIBIA Regenerative processes have been studied in amphibians over a very long period. Spall- anzani (1768) recorded many definite ex- periments in the Prodromo, and the lit- erature of the nineteenth century is full of allusions to naturally occurring forms which came to notice because of the abnormal- ity which appeared either during or after the regenerative process. These forms which appear with a fair frequency in nature often are exceedingly difficult to interpret and in some cases impossible to duplicate in the laboratory. Hellmich (’30), for example, figured a salamander with a single but huge limb near the dorsal midline at the general limb level. In this particular location an embryonic transplant would have to be made early in the embryo and _ probably would resorb in a great proportion of such cases. It is impossible to interpret the abnormali- ties found in nature except hypothetically. Bateson (1894) gave a rough classification of abnormalities, as did Przibram (’26) in his later catalogue. Tornier (’06) showed that all the forms produced could be called ex- amples of the multiplicities occurring after regeneration by a single but drastic experi- ment. With a single cut through a tadpole extending through the forming hind limbs, Tornier secured animals with 6 to 8 mul- tiple limbs. The situation is considerably clouded in amphibians, owing to the life histories of various forms. Not all amphibians have an equal power for regeneration at all stages in their history; the anurans normally are restricted in regenerative capacity to larval stages. The length of the embryonic and larval stages varies tremendously; in some of the tropical forms metamorphosis is al- most part of the embryonic period; in others, as in the bullfrog, two full years may lapse before metamorphosis is complete. These variabilities when correlated with the capac- ity to regenerate offer little in the way of a constant around which we can collect the varied miscellany of events. Kammerer (’05) discussed an absolute age and claimed that this, rather than the stage, determines re- generative capacity. We find this same general idea in general growth studies when 677 the necessity for evolving a base level has given rise to all sorts of postulated constants. Embryonic regeneration constitutes dan- gerous ground, for it is difficult to speak of regeneration in a structure which is going through its formative stages. Nevertheless, the term has been used frequently and the issue was clearly raised when the Roux- Driesch controversy on capacity of the half egg was occupying the center of biological attention. The circumstances of this situa- tion are well known. Suffice it to state that Roux’s idea of the egg as a mosaic was found- ed upon the half embryo developing from a single blastomere. Schultze (1894) turned the egg after killing one blastomere and secured a small but complete embryo. When Roux restudied the question he was faced with a negation of the mosaic idea, but in- stead of retracting he called the process by which more than a half embryo is formed post-regeneration. Now it is well known that the amphibian egg is capable of a fairly complete internal reorganization; it has been proved most conclusively that its capacity to continue development and _ ultimately produce a complete and normal embryo can hardly be blocked by any experimental contingency. In the face of these facts, it might be better to assume that the reconsti- tution is a form of regulation which is pos- sessed to an unusual degree by the em- bryo. This is shown clearly in some of the limb experiments (Harrison, 18). In some cases of incomplete duplicity the manus de- veloping after rotation of the embryonic limb disc may be represented only by a small spear of tissue. If this be cut after it has developed to the late embryonic or early larval stages, the regenerating appendage frequently is much more complete than that originally secured and will show sufficient morphological criteria for an accurate diag- nosis of the asymmetry. Swett (724) also used this method in studying the process of reduplication. In both these cases the in- vestigator is using regenerative processes to explain embryonic reactions during the proc- ess of reorganization of the limb. These processes differ only in degree from the original formation of the end organ. The embryonic structure is drawn from a wide embryonic field, while the regenerated struc- ture is derived from a much smaller and more localized field. The problem of the field of regenerative organization has been ably discussed by Weiss (’26a) and Guyénot (’29). The essential point which the field concept 678 has added is the postulation of a directive force orienting particles in a field. In the urodeles the speed of regeneration is sometimes much faster in the larvae than in adults. In older larvae and in neotenous forms there is sometimes a noticeable failure in the capacity of the organism to regenerate after minor operations, e.g., a slight V- shaped piece of tissue may be removed from the dorsal fin extending to the upper surface Fig. 234. Regeneration of the salamander ex- tremity (Triturus taeniatus). A, a-h, Varying stages in the regeneration of the forelimb; B, differentia- tion of skeletal parts in the right hind leg; C, further advanced differentiation in the left forelimb. (From Fraisse, 1885.) of the myotome. The edges heal but the tissue, which is of a rather nondescript and unorganized type, may not be replaced. In fishes when the fibrillae alone are removed from the dorsal fin, there is no reconstitu- tion. An entirely different situation occurs in the anurans where normally the limbs re- generate only during larval stages. The exceptions will be considered below. Bar- furth (1895) worked on the limb problem in Rana fusca. When the limb was ampu- tated while still in the condition of a small bud (no real limb), regeneration occurred and was complete in detail. When an ampu- tation was performed upon a paddle-shaped appendage when the digits were just begin- ning to show, imperfectly formed appendages were secured, the deficiency being marked in the foot. After the stage where digitation is complete and the knee bud is marked, an amputation provokes little or at most imperfect regeneration. These results were REGENERATION repeated by Byrnes (04) upon the forelimb with nearly identical results. Guyénot (27) studied the decrease in capacity to regenerate. The capacity to re- generate lasts longer in the tail than in the lmb. If a limb is_heteroplastically grafted from a toad to a salamander it fails to regenerate in its new location. If, however, a tail is similarly grafted, it will regenerate. The failure of the limb is due to something intrinsic in the limb tissue and is not due to the internal medium. The transplanted urodele tail will regenerate even though retrogressing during metamorphosis. Harrison’s (’21) thorough analysis of the problem of asymmetry with the embryonic forelimb has been adequately supplemented by Weiss’ (’26b) results with larval and adult limbs. Weiss cut the limb so that he got two definite surfaces, from each of which a complete limb developed. The regenerative blastema is a harmonious equipotential sys- tem in the Driesch sense. It is possible, in the light of Holtfreter’s (47) experiments, that the blastema may serve as an organiz- ing center and that the results obtained in some transplantation experiments may be secured from host tissue by induction. This is particularly true in early larval stages. The active regeneration capacity of newts and salamanders is almost unlimited, par- ticularly with respect to the limb. This will regenerate at any time and at any level (Fritsch, °11). The limb and its girdle can be removed and regeneration of the limb will take place. In other words, no matter how complete the removal, the limb is re- constituted and is a complete functional limb. Braus (’09) showed the independence of limb formation in the embryo from skeletal parts. Wendelstadt (04) thought, on the basis of his experiments, that the presence of the skeleton was essential for regeneration. A part of the bone when removed was re- generated, but when the whole structure was carefully disarticulated no regeneration occurred. Strasser (1879), Goette (1879) and later Morrill (18) had shown that when the limb was sectioned, perichondrium and peri- ostium became very active and made up a substantial part of the blastema. Weiss (’25a) repeated and extended Wendelstadt’s experi- ments and showed that removal of the skele- ton did not prevent the appearance of skele- ton in the newly regenerating limb, but in confirmation of Wendelstadt there was mo regeneration of a completely extirpated skele- tal element in its old bed. Only if a remnant VERTEBRATES of a skeletal element is left behind will this regenerate in its own right. Bischler (726) showed that Wendelstadt’s results were faulty, for when repeated in a large variety of experiments, she secured regeneration of both the free extremity and the skeleton. Milojevic (24) thought he had determined a polarization of the materials in the blaste- ma. By interchanging the cap of fore and hind limb regenerating blastemas he found that during the first ten days the basal part PERIOD OF WOUND HEALING (%o) 10 PERIOD OF BLASTEMA FORMATION 679 nutritional conditions, the interaction of other systems with their effects together with degenerative phenomena and capacities of regrowth. In this chaotic milieu the inves- tigator is always trying to find order and certainly reaches a point when his own observations are coalesced into a subjective unit which appears satisfactory. So far, our ideas of the formation of the blastema can be outlined in much oversimplified form under the following headings: PER/OD OF DIFFERENTIATION 15 DAYS OF REGENERATION Fig. 235. Changes in the epidermal mitotic index (E.M.I.) and mesenchymal epidermal ratio (M.E.R.) during a 28-day period of regeneration. (From Manner, 753.) influenced the distal part no matter what the orientation of the graft. After this indif- ferent period the distal part develops accord- ing to its origin with respect to both its axes and the form of the appendage. This work may be correct but the criteria used are not as reliable as Milojevic thought them to be. In the early regenerates it is very difficult to differentiate the fore from the hind limb and the bones themselves are too similar in form to permit an accurate interpretation. The origin of the blastema has proved a fascinating enigma. So far there has not been a single absolutely critical experiment. Every worker in the field of limb regenera- tion has honestly tried to come to grips with this question, which is an exceedingly im- portant one since it involves tissue reactions, 1. Regeneration by extension, in which after the healing of the wound and the de- differentiation of tissues the regeneration blastema is formed and then becomes or- ganized with the parts remaining, each giving rise to like tissues and the structures which grow to normal limits in replacement of the missing parts. This type of regenera- tion seems to occur in the tail but is not clear-cut in the limb. 2. Blastema formation, by the invasion of the wound area by new elements, generally from the blood stream, which act in the organization of the blastema and join with the cells there in the new growth process: hematogenic origin. 3. The participation of the epidermis, through dedifferentiation and a direct amal- gamation of the epidermal elements into 680 blastema. This idea will need discussion since it has been revived by Rose’s (748) work. 4. The origin of the blastema from reserve cells, similar to the archeocytes of the worms Fig. 236. Transition from fibroblast to fibrocyte. FB, Fibroblast; 7, an intermediate type; FC, fibro- cyte. Camera lucida drawing. * 450. (From Man- ner, *53.) or from cells which retain their embryonal capacities and potentialities. Unfortunately, most workers in the field usually adopt an intermediate position with regard to the above headings, e.g., Puckett (34) considers the blastema as a cell mass composed of materials secured by complete dedifferentiation of all the formed elements. This logically means that the parts so gener- ally derived have to undergo a complete sorting-out process to attain their later def- inite structural relationships. It is a com- bination of like from like with an inter- mediary dedifferentiation. The process of this sorting out is of course very difficult to follow and a causative mechanism so far has not been ascertained. Manner (753), like many others, has studied in detail the three phases which have to be considered in regeneration of the limb of Triturus, using as indices the changes in mitotic activity of the epidermal and mes- enchymal tissues during the 28 days follow- ing amputation. (See Fig. 235.) In securing his figures for the mitotic index he has been most careful to avoid some of the pitfalls of this method, e.g., Litwiller (39) has shown that the mitotic index varies with the time of day; in recog- nition of this, Manner has removed the blastemata used for his study at a fixed time for all cases. Since the epidermis is the only tissue which can be counted accurately, he counts the mitotic mesenchymal cells REGENERATION and compares this figure with the total number of epidermal cells at any given stage. A total count of all cells in the mes- enchyme is practically impossible, since the amount of debris resulting from amputation is an obscuring factor which he recognizes. Nevertheless, by this comparison he secures a relative value which can be employed. Manner draws a rather sharp line in def- inition between the fibroblast shown by Maximow and Bloom (752) to possess the capacity for the formation of all connective tissue elements and the fibrocyte which has lost its embryonal capacities. The evidence from this paper, as well as from others (Nassonoy, ’30; Thornton, ’38b; Needham, °42; Liebman, *49), all points to a fibroblast cell as the effective agent in blastema formation. The possible influences engendered in the degeneration after am- putation and the relationship of nerves and muscles are taken up as logical steps in the proximodistal progress of regeneration. It is fairly clear that the like from like prin- ciple does not hold here and that the sorting- out process is conditioned by the nerve sup- Fig. 237. General diagram of a regenerating fore- limb. M, Epidermal mitotic cell; ¥, mesenchymal mitotic cell; 7, the epidermal cells on one side, in a single section, from the level of amputation to the middle of the blastema. (From Manner, ’53.) ply in the presence of muscle. Epidermis certainly is not primarily needed (Weiss, 27); bone may be elimated from the stumps (Weiss, ’25b; Thornton, ’38b), and still re- generation takes place. Other evidence such as the regeneration VERTEBRATES 681 Tasie 31. Summary of the Major Histological Changes Occurring in the Adult Urodele Limb During a 28-Day Period of Regeneration (from Manner, ’53) DAYS OF REGENERATION MAJOR HISTOLOGICAL OBSERVATIONS 1 Day 2 Days The amputation wound is closed over by the migrating epidermal cells. Numerous phagocyte cells are present in the wound area. The disintegration of cartilage begins. The continued migration of the epidermal cells to the distal end of the regenerating limb results in an accumulation of these cells at this point. 5 Days The disintegration of the striated muscle begins. The epidermal cells continue to accumulate at the distal end of the regenerating limb. 6 Days The connective tissue, with its component cells, migrates between the cut end of the bone and the overlying epidermis. The disintegration of the cartilage and bone continues, resulting in cartilage and muscle fragments which presumably are incorporated into the blastema. 16 Days There is a maximal accumulation of the epidermal cells at the tip of the regenerating limb. The disintegration of the cartilage and muscular tissue continues. 20 Days The first finger makes its appearance. The fibroblasts are beginning to differentiate into chondroblasts. The disintegration of the cartilage and muscle ceases. 28 Days entiate. The second finger makes its appearance. The new cartilage continues to differ- of a haploid chimaeric limb (Hertwig, ’27), or Butler’s (35) transplantations of an un- irradiated limb to an irradiated host which had no capacity for regeneration, or the histological studies showing that any or all of the mesodermal tissues can give rise to the formative elements of the blastema—all of these point to mesenchymal components as the essential elements in blastema for- mation. A different line of evidence is secured in the work of Heath (53). He replaced the ectoderm of the embryonic limb buds in two slow-growing species of salamander. When A. tigrinum ectoderm (fast-growing) is grafted over Triturus torosus limb bud mesoderm there is first a slight retardation in development, later an acceleration re- sulting in a larger than normal limb. When T. torosus ectoderm (slow-growing) is used Fig. 238. 7, 48 hours after amputation. E, Epidermis; C, cartilage; M, muscle; C.T., connective tissue. Camera lucida drawing. < 100. 2, 6 days after amputation. E, Epidermis; C, cartilage; M, muscle; C.T., connective tissue; Bl, blastema. Camera lucida drawing. * 100. 3, Epidermis at 16 days after amputation. HE, Epidermis. Camera lucida drawing. < 100. 4, 20 days after amputation. EZ, Epidermis; B/., blastema; C.7., connective tissue; C, cartilage; P.C., precartilaginous tissue. Camera lucida drawing. < 100. (From Manner, ’53.) 682 REGENERATION fé Fig. 239. Amblystoma punctatum larvae; 6m sections of regenerating forelimbs, stained for alkaline phosphatase (from Karczmar and Berg, 751). 1, Normal enzyme distribution. Healing on the first day after amputation. E, Epidermal cap, + (single plus); M, muscle, + (single plus); C, diaphysis and marrow of humerus, + + (double plus); H, hypodermis. < 150. 2, Dedifferentiation and first increase in enzyme activity, third day after amputation. Hypodermis (H), perichondrium (P) and weakly staining muscle (M) dedifferentiating distally into + -+ (double plus) mesenchyme. 3, Accumulation of enzyme-rich mesenchyme: fifth day after amputation. Bm, Mesenchyme, staining + + (double plus) to + + + (triple plus); C, proximal epiphysis of humerus with + + + (triple plus) matrix; Ch, hyaline cartilage of shoulder joint with enzyme-free matrix; M, muscle, + (single plus); H, hypodermis, + + + (triple plus). x 80. 4, Early blastema with an apical gradient of enzyme activity, 9% days after amputation. This section was incubated 20 hours beyond the optimum time, to test for the leakage of stain. The reference VERTEBRATES 683 to cover the limb mesoderm of A. tigrinum site of presumptive limb components. In the result is a smaller than normal limb.' stages 51 and 52, enzyme activity decreases When the chimaeric larval limbs are am- locally and the differentiating tissues sepa- putated they regenerate according to the rate regionally into loci of lower and higher growth rate of the original mesodermal enzyme activity. components. If the fast- or slow-growing This brief description is applicable to the species epidermis was an essential part in regional development of the differentiating the formation of the blastema, it did not limb components. When, however, the spe- register in the size of the regenerate although cific tissue differentiations occur there is a de- it did have an effect upon the original limb crease in the alkaline phosphatase activity growth. This indirect evidence suggests during myogenesis and by the time the muscle strongly that there is no epidermal contri- bundles are formed the phosphatase de- bution to the blastema. creases to the relatively low larval level. A Another line of evidence for the impor- similar condition prevails in cartilage for- tance of mesoderm in the formation of the mation, although there is a_ slight but blastema is found in the work of Karczmar perceptible rise at the beginning of bone and Berg (51). They have studied the occur- differentiation. Throughout limb develop- rence of alkaline phosphatase during both ment, there is no demonstrable localization embryonic development and regeneration in of alkaline phosphatase in the epidermis of Amblystoma. Brachet (746), Krugelis (50), the body wall. Lindeman (749) and Moog (44) have cor- Karczmar and Berg (51) (see photomicro- related the alkaline phosphatase content graphs 1 to 7, Fig. 239) have divided the with various steps in differentiation phases regeneration process into three overlapping of normal ontogeny. The importance of the phases: (1) dedifferentiation, (2) growth and study is that it tests whether normal onto- (3) differentiation. They have traced the genic processes, in which alkaline phospha- localization of the alkaline phosphatase in tase levels are low, are duplicated in the each of these stages. There is an increase in dedifferentiation subsequent to injury. level about three days after amputation, Histochemical localization of alkaline rising through the growth period between the phosphatase was made (1) on 25 young larvae, fourth and fifth day and becoming region- 13 to 17 mm. long, (2) on 60 regenerating ally localized in the regenerate on the fif- larvae, 25-47 mm. long, and (3) on adult teenth day (see photomicrograph 7, Fig. regenerating limbs. In ontogeny the hind 239). After this the enzyme distribution limb primordium (stage 47) shows cells with follows the ontogenetic pattern. a high alkaline phosphatase content. As The evidence on the epidermal contribu- the limb bud (stages 49 and 50) begins to tion to the blastema seems clear. The epi- project from the body wall, the phosphatase dermis covers the wound in a few hours, but activity is strongest at its apex. This relation is phosphatase-negative until six days after persists and increases as the limb parallels amputation, by which time the blastema is the trunk (stage 50); later (stage 50+) the well formed and growing. The localization stain shows a definite concentration at the of the enzyme is variable, after this period base of the elongating limb, marking the stronger at the apex of the epidermis over structures still provide an accurate measure of enzyme activity: muscle stains + (single plus); acellular diaphyseal shaft + -++ (double plus). Blastematic nuclei (B) stain + + + (triple plus). « 20. 5, Conical blastema, staining + -+ + (triple plus), 10 days after amputation. This section through the margin of the perichondrial sheath of the humerus, P + -++ (double plus) demonstrates the enzyme distribution in muscle. H, Hypodermis, + -+ + (triple plus); D, enzyme-free distal edge of collagenous derma, + (single plus) marking the level of amputation; Ch, hyaline cartilage, with + (single plus) matrix. Note the patches of enzyme activity on the epidermis. x 20. 6, Differentiating blastema with regions of high and low enzyme activity, 13 days after amputation. D, Newly secreted matrix of the differentiating derma, + (single plus); Pu, presumptive perichon- drium, + + + (triple plus) of humerus and ulna—the parallel, + -+ (double plus) area is the differentiating muscle of upper arm and forearm; B, finger-bud blastema, + + -+ (triple plus). « 80. 7, Detail of a differentiating blastema, 12 days after amputation. C, Rudiment of humerus, illustrating from right to left, the primary differentiation of chondrocytes accompanied by a drop in enzyme activity from -+- + + (triple plus) to + (single plus), and the secondary increase of intra- and extra-cellular enzyme activity in the hypertrophic chondrocytes; M, differentiating muscle, staining + -+ (double plus); P, differentiating perichondrium, staining + + + (triple plus); V, blood vessel with + +--+ + (triple plus) walls. 684 the blastema and weaker at the base of the limb. This brief increase in level extends over the fifth and ninth days, after which there is a decrease, and no alkaline phospha- tase is demonstrable after the twelfth day. As is apparent in Figure 240, changes in the phosphatase level occur with or before the first histological criteria of differentiation, PHOSPHATASE REGENERATION It is essentially a formation from mesen- chyme derived from fibroblast cells rich in alkaline phosphatase or by similar cells secured from dedifferentiation of muscle and cartilage. The possibility of an epidermal contribu- tion to the blastema seems at present most unlikely on the basis of former histological DAYS Fig. 240. both in embryonic limb formation and in larval limb regeneration. The cells compos- ing the regeneration blastema react dis- tinctively to phosphatase stains. Fibroblast and dedifferentiating muscle appear to fur- nish the chief components contributing to the blastema, a finding in agreement with those of Butler (33), Thornton (’38a,b, ’42) and Forsyth (46). The contribution of the epidermis to the blastema, as advanced by Godlewski (’28) and Rose (’48), receives no support from the studies on phosphatase localization and level. The phosphatase in the epidermis appears after the preliminary phases of regeneration are complete and it is seldom that the epidermis elements merge with the blastema. The problem of blastema origin then stands at present on a fairly substantial accumulation of material which clearly supports the older histological observations. studies which show the non-participation of epidermal structures as well as the later studies of Thornton (’38a,b), Manner (53), Heath (53) and Karczmar and Berg (’51). The descriptions of careful observers cannot easily be thrown aside and the indirect evidence on mitotic loci and the abundance of embryonal fibroblasts, combined with chimaeric limb regeneration studies and the alkaline phosphatase correlation, all add to the fund of knowledge which places the burden of proof upon those who claim an epidermal contribution to the blastema. On the basis of the evidence, it is much more likely that the epidermis is either passive or an inhibitor of regeneration. This idea is supported by Rose’s (44, °45) observations on regeneration of the limb in adult anurans. These forms are recalcitrant to limb regeneration, but limb development can be initiated by preventing wound heal- VERTEBRATES ing, either by treatment with strong sodium chloride solutions, by stripping the epithe- hum from the amputation surface, or by causing a temporary vitamin deficiency. He suggests that the dermal layer of the skin prevents the regenerative processes which lead to the formation of a rapidly growing blastema. Gidge and Rose (44) suggest again that timing of healing is a factor in the normal failure of regeneration in frogs. The dermis redifferentiates before the blastema cells can grow, and the dedifferentiated elements vicar- iously become scar tissue which acts further in inhibiting the regenerative growth. A stripped wound is covered by epidermis but not by dermis. The regenerate consists of cartilage only. When, however, the wound is covered with larval skin, a distal blastema develops and a regenerate composed of all limb tissues is formed. This work forms a promising lead for future work in the puzzling field of anuran limitations for regeneration. In the urodele embryo the ectoderm sometimes inhibits rather than assists in differentiation, but the facts here seem quite different from those needed to explain the regenerative inhibition in anurans. The one dominant factor in limb regenera- tion is the import of the nervous system. This has been shown beyond peradventure by Singer (42~49) and by Singer and Egloff (49). There is a basic quantitative minimum of innervation required for limb regeneration. This can come from the motor, sensory or sympathetic nervous system. If this basic minimum is not present limb re- generation will not occur. The limitation of space prevents the discussion which this im- portant work deserves and the reader is re- ferred to the original sources for this material. Weissfeiler (24) and Vallette (26) worked on the regeneration of the amphib- ian head region. Vallette (’26) found that if the cut is anterior to the nasal capsule, a complete regeneration occurs. If it is pos- terior to it, the wound heals but regenera- tion does not occur. This is true also in the embryo, where May (27) found that in the early stages the nasal capsule can re- generate but that it is one of the first of the organs to become fixed. Weissfeiler found that the forebrain will regenerate if the nasal capsule is not dis- turbed, but not if the capsule is removed. This is an interesting corollary to Burr’s (16) experiment in which, when the em- 685 bryonic olfactory placodes are removed, the forebrain fails to develop to normal size. Vallette transplanted the jaw to various regions, then cut the jaw and followed the regenerative process. It reconstituted a typ- ical jaw no matter where located. Hooker (’25) studied the regeneration of the spinal cord after section in various stages. While the time is progressively longer as the anurans approach metamorphosis, A B Fig. 241. Triton head regeneration. A, The levels of the cuts 7 and 2; B, normal; C, regeneration after cut at level 7; D, regeneration after cut at second level. (From A. v. Sziits, 14.) the process is complete both morphologically and physiologically. These studies and those of Guyénot and Schotté on the influence of the nervous system will be considered in greater detail under that head. ENDOCRINE EFFECTS The majority of the workers dealing with the interaction of the endocrine secretions upon the process of regeneration have em- ployed fishes; Grobstein (’42—’48) has con- tributed papers dealing with regeneration of the gonopod under varying conditions. The gonopodium of the male Platypoecilus has a declining regenerative capacity which is due to the inhibition of regeneration of circulating androgen; testosterone propionate changes the female gonopod to form a re- 686 generate of the male type. A masculinized female has a declining regenerative capacity from which it does not recover after the withdrawal of androgen. In Gambusia, Turner (’47) has given the results on normal and castrated males. In the castrate the rate of regeneration is the same as for juvenile males or females. Hop- per (49a) has performed the same experi- ments on both males and females of Lebistes where the castrate females regenerate nor- mally while the castrate males regenerate a female type of fin. When he exposed the fishes to ethynyl testosterone in two con- centrations (Hopper, ’49b) the females de- veloped and also could regenerate typical anal fins; the males showed no loss of re- generative capacity. These three investiga- Fig. 242. Vertebral breaking point in the lizard tail (Lacerta muralis) (from Slotopolsky, ’21-22). tors, using similar methods for the study of three different teleost fishes, show how very different the regenerative behavior can be in respect to hormonal activity. In the urodeles, Schotté and Hall (752) have tested the effects of hypophysectomies performed at varying times on the formation of the regenerating forelimb. The wound- healing phase was severely affected by the hormonal imbalance, dedifferentiation was affected but not so severely, while blastema and growth phases were hardly affected at all. They propose that the growth factor is not the acting one, but that probably it is the ACTH factor acting through cortisone to regulate the wound epidermis action upon the cut stump tissues. Considerably more in- formation will be needed before this mecha- nism of action can be confirmed. Richardson (’45) has studied the effect of the interrelationship of the thyroid and hypophysis upon hind limb regeneration. The hypophysis in the absense of the thyroid has a greater inhibiting effect than thyroid acting alone in the absence of the hypophy- sis. When Antuitrin-G and thyroxine were supplied as substitutes for the hypophysis and thyroid, respectively, good cartilaginous skeletal regeneration was secured. Regenera- tion without removal of hypophysis and thyroid is accelerated by Antuitrin-G. REGENERATION REPTILIAN REGENERATION Fraisse (1885) listed thirty or forty cases of regeneration observed chiefly on lizards, with a few observations on the serpents. There are also a few scattered observations upon some of the turtles. The double- and triple-tailed lizards, however, held most of the attention, but in the category of abnor- mality rather than regeneration. These are naturally occurring forms. Cuvier (1829) commented on the neces- sity for the study of regeneration in the lizard and remarked concerning his interest in the abnormal development of bone after autotomy. Gachet (1833) found that regen- eration was not confined to the lizards Lacerta agilis and L. viridis, but that it was possessed in varying degrees in others of the reptiles. He examined five lizards with double tails, four with regenerating tails and several with single or double regen- erating tails. Of the native forms, L. muralis, L. viridis and L. ocellata were represented, and also Anolis iguana. His findings were all verified by dissections and are com- pletely described in Fraisse’s (1885) mono- graph. Gachet gives the literature to his date, together with a description of cases under discussion. Guyénot and Matthey (28) and Guyénot and Ponse (730) traced the regeneration of the limb and repeated Fraisse’s work on the formation of the blastema and the regenera- tion of the tail. Weiss (’23) tested the effect of transplantation on the regenerate as well as the effect of the whole upon the graft and the graft upon the whole. A tail blas- tema transplanted to the limb region regen- erates a miniature tail, not a limb. Occasion- ally when a small strip of tissue or a small incision is made near the hind limb, a small but imperfect tail results. Slotopolsky (2122) restudied the mech- anism of autotomy in the lizard Lacerta and found the explanation of the evenness of level with which it occurs, for there are two vertebrae behind the pelvic girdle in which the midpoint of the vertebra is an unfused part, giving a rough articular sur- face which, because of strong muscular con- traction, serves as a line of cleavage. While other forms of autotomy occur in the higher vertebrates, this is the only series recorded in which the mechanism is similar to the much more efficient mechanism of the ar- thropods. Many rodents autotomize the tail but the epidermis is all that is lost. Sumner and Collins (718) described this process for Peromyscus. Other forms tend to chew parts VERTEBRATES which are caught in traps, but this is dis- tinctly not an automatic reflex type. The various tissues regenerated within the tail structure are far different from its original composition. The external form of the tail, including its scaly covering, is fairly 687 tebrae are not segmented but form a con- tinuous tube which, because it is thin walled and little calcified, has the same or greater degree of flexibility as had the segmented structure. The muscle mass regenerates but shows little of the original segmentation and Fig. 243. X-rays of regenerates of the tail in Lacerta muralis. The right-hand illustration shows a regenerate replacing the normal. The stimulation in this case was by cutting the ninth tail vertebra. whereas the left figure was cut at the fourteenth with regeneration occurring at the breaking point of the eighteenth tail vertebra. (From Slotopolsky, ’21—’22.) completely reconstituted, but the internal relations are not normal. The spinal cord regenerates incompletely; the membranes extend down a rather amorphous type of tube formed of procartilage, the nervous tissue itself being reduced to a thin strand chiefly glial rather than neuronal. Seldom are spinal ganglia formed, for the nerve trunk gives rise to no nerve roots except at the anterior end of the regenerate. The ver- the organ has a flaccidity which is quite different from the original state. Woodland (21) gave an excellent picture of this con- dition in the gecko (Hemidactylus flavivir- idis). Limbs are shown by Egger (1888) and Marcucci (730) to have the ability to regen- erate in part, but here as in the tail region the skeletal elements are imperfect and, be- cause of this, many of the appendages re- 688 generated are tail-like in appearance. The scales are much more like scales which cover the tail than those ordinarily found on the limb. The character of the scales, their size and location, has frequently been used as species diagnostic, and here the form of the regenerate seems to change the scale pattern; e.g., in Weiss’ (730) case where the tail was transplanted heterotopically to the limb region, the tail was later covered by the scutes which formed from the material which originally would have covered a limb, but the scales were definitely tail scales. Noble REGENERATION an observation by H. Miiller (1864) that temperature had a decided effect upon re- generation in lizards. Noble and Clausen found that when the temperature was kept constant the skin of any definite area regen- erated the same type of scale. They found no evidence of reversion to a common type. Moreover, the scales of the head and back are regenerated in a form similar to the original. The size of the regenerate has something to do with form, for small areas of epidermis can be removed even on a regenerating tail and will be restored to the original scalation. 0 PO : rr Fig. 244. Tail regenerate in the gecko (Hemidactylus flaviviridis). K, Reconstituted cartilaginous tube surrounding central canal with poorly regenerated nervous system. (From Woodland, ’21.) and Bradley (733), and later Noble and Clausen (’36), ran a series of experiments to determine what the factors might be which are involved in the pattern, form and color of the scales. While it has been known for a long while that the lizard scales in a regenerate may be quite different in arrangement and form, Boulenger (1888) pointed out that the re- generate might be a much less special or individualized type of scale and that there- fore it was closer to the ancestral type. Wer- ner (1896) confirmed this idea and intro- duced the idea of reversion to the ancestral condition. Tornier (’06) pointed out that this type of reasoning was slightly unsafe, but the idea had caught and was extended by Werber (’05) and carried to its logical reductio ad absurdum by Barbour and Stet- son (729), who compared the scale pattern in an extant regenerating form with that of an “undoubted ally of Jurassic age.” Marcucci (’26) and Avel and Verrier (’30) gave data which showed that the shape and size of the regenerate determine the scale form. When Noble and Clausen (’36) de- cided to repeat this work they made use of The work of Bovet (’30) and Marcucci (732) seems to indicate that the factors con- trolling scale form lie in the mesoderm of the blastema, but this can be changed by temperature effects (Noble and Clausen, 36). Noble and Clausen have studied the formation of the relatively broader ventral scalation—this type of scale can be caused to form on the dorsal side of the tail by implanting muscle from the ventral half of the tail. The converse experiment also holds. The pigment or shade of the scale is deter- mined by the scale location and is not af- fected by the transplant. Barber (744) has studied the healing of the wound over the tail and compared this with the healing in the limb region. She concludes that the heal- ing is responsible for the restriction of re- generation in the limb. PHYSIOLOGICAL REGENERATION We can now turn from the active or acute type of restorative process to the repetitive or physiological process which is character- istic of both the preceding group and the subsequent group of birds and mammals. VERTEBRATES Here it would seem possible to differentiate sharply, but the line is not so clear as one would think, as the subject when analyzed yields a difficult series of interrelations, some of which occur during all the normal states of use and repair, shading to an acute res- toration of parts which are badly damaged if not completely ruined by a physiological overload. The step between a normal and a pathological process is so minute that it is hard to determine except by arbitrary rule which one belongs in which category. The same condition exists here, and in the following evaluation of the repetitive proc- esses it is by no means simple to place them under a hard and fast series of definitions. Epidermal Replacement. The stratum germ- inativum, or malpighian layer, is constantly giving rise to new elements which are re- placing the superficial parts. In reptiles the replacement may be gradual as in the Che- lonia and Crocodilia, or in rapid bursts of cornified epithelium as in some lizards and serpents, in which the surface cornification may be a complete cast rigidly mirroring every detail of the underlying parts. Even the corneal layer of the eye is sloughed at an ecdysis. Here again the process has a superficial resemblance to that occurring in arthropods. The glands in association with the cuticle, and the individual scales of which the cuticle are composed, all have their basic levels of regeneration through the activity of the stratum germinativum. The skin of mammals has a similar con- stitution and replacement power but the surface epithelium is sloughed slowly, and the replacement is not noticeable unless there is an actual denudation of a large area with subsequent restoration. Bishop (745), attack- ing this problem experimentally, finds by biopsies of varying thickness that the limi- tation of healing and scar formation can be fairly definitely marked out. Complete re- generation results when removal does not go below the reticular layer. Fibrosis is in- hibited if the papillary layer is left intact. In the birds the situation is quite different, for they seem to have the attributes of both groups. Feather regeneration occurs after feather removal, and the process of reforma- tion is essentially similar to the embryonic process. Samuel (1870) gave the first nearly complete picture of the process. Lillie and Juhn (32) showed the complete picture of feather regeneration and the pattern of the regenerating feathers. In addition to the usual process, birds have a periodic moulting in many cases and 689 this process is similar to the ecdytic phe- nomena in reptiles. They have, however, little replacement of skin as the keratiniza- tion does not seem to reach the same degree as it does in mammals. Moreover, birds possess in the beak an unusual integumen- tary appendage, with the capacity to regen- erate a large proportion of this structure after injury or removal. Scales and spurs also can be replaced. Hair growth and that of claws, hooves and nails are relatively constant and do not call for special consideration. The growth of horns, however, proceeding as it does in a very definite way in the forms which possess the epidermal types, is really a remarkable case of repetitive restitution. It was studied in detail by William Harvey and John Hun- ter. The correlation between the testis and the antlers was made very clearly because castrates in which the horns are shed do not regain the increasing branched antler but the rudimentary form. The glands in connection with the in- tegument—sebaceous, sweat, ceruminous, mucous and the necrobiotics—all have the potentiality for comparatively rapid regen- eration. The mammary gland was investi- gated quite early, and it was found that the nipple does not regenerate, nor will the gland regenerate if the organ is completely removed. If a part is left, however, the work of Gardner and Chamberlin (’40—’41) shows the amount of reorganization which can occur. The tubular ducts and the glandular epithelium both undergo amazing phases of growth and reconstitute an active gland from its involuted condition in a short time. While considering the general situation in epidermal structures, it might be well to digress for a moment to see what the super- ficial reactions are during wound healing. Epithelium in general gives rise to both the glandular and the surface types. Occasion- ally one type may form the other, but they usually run true to rule. Peters (1885) found that after the cornea is scraped the wound is first covered by lymph which forms a clot through which the epithelium becomes ac- tive by (1) spreading movement and (2) cell proliferation. Movement by spreading is the initial reaction to the produced defect. In the entire contiguous cell area the cells are amoeboid and spread over the denuded area. Mitosis resulting in cellular prolifera- tion begins about 20 hours after the injury, and here also the reaction extends to cells quite removed from the area of injury. The same process seems nearly universal in 690 epithelial wounds. Barfurth (1891) showed a similar process in frog larvae, and Poynter (22) in chick embryos; Oppel (12) traced the movements of corneal epithelium of the cat, but missed completely the amoeboid stage. Matsumoto (718) repeated Oppel’s tissue culture experiments using intra vital stain to trace the reaction and confirmed Peters. Rose (48) has reinvestigated the role of the epidermis and returns to the older idea of Godlewski that the blastema is formed from differentiated epidermis. Muscle Tissues. Muscle does regenerate from muscle remaining after injury. The difficulty with reconstitution here seems to be the infiltration of connective tissue which prevents the slower muscle elements from entering. It was formerly thought that myo- fibrillae had no capacity for regeneration but that the entire replacement was a fibrous con- nective tissue. This is true for heart damage late in life. Cartilage and Bone. Cartilage does not re- generate readily in the higher vertebrates, in which it is formed from perichondrium. Bone regeneration is a complex process, de- pending first upon the formation of cartilage with the appearance of callus, then cartilage and the chondroclasts, osteoblasts and osteo- clasts, resulting in the typical bone recon- stitution. The time required for complete repair is variable and depends to a large extent on calcium utilization. There is an important line here between the vitamin complex and other internal factors. The Gut. Desquamation is constantly oc- curring from the entire alimentary mucosa, probably greater in degree than that occur- ring in the epidermis. There is a constant and consistent replacement of the glandular content of the salivary, esophageal, gastric and intestinal regions. The proliferation of the mucosa is not of the pronounced cyclic type, but is a continuous process. When a part of the gut is removed, the intestine re- maining does not reconstitute the part. The gut remains in its reduced condition. Flint (10) removed 80 per cent of the intestine in dogs, with viability; in such cases the extent of villation is markedly increased, and physio- logical readjustment takes place by a com- pensating hypertrophy of the villi which makes possible an increase in the absorptive area. Grant (745) has experimentally re- moved the epithelial cells of the gastric mucosa and finds that the cells are replaced within a few hours, provided that the under- lying gland cells are not disturbed. Glands. There have been many studies on REGENERATION the regeneration of glands. Podwyssozki (1881) studied the regeneration of liver in the rabbit. The bile ducts give rise to liver cells. The salivary glands and kidney have also been studied. The kidney is an example of compensating hypertrophy, but the re- generative phase is in many cases a very active one. In these three big classes of vertebrates, there is by far more regeneration than one would suspect in a superficial examination of the field. Hyperplasia and compensatory hypertrophy play an important part in these forms. The repetitive type of regeneration is a vital and probably one of the chief sur- vival values of these forms. GENERAL HISTOLOGY The early studies on the histology of the regenerative process interpreted it in the light of embryonic development (Goette, 1879; Strasser, 1879). Barfurth (1891) was the first to point out the wandering character of the epithelial cells and their stretching quality which covers the wound before ac- tive cell divisions occur. At the same time Peters (1885) showed that this same process occurred after wounds of the cornea. Since then it has been found frequently (Poynter in chick embryos, and numerous individuals working on amphibians). In the tail, regeneration is brought about by a typical succession, each tissue regen- erating its like. Nerve cord gives rise to nerve cord, but the regenerated nerve cord is frequently different from the original, as are also the spinal ganglia and spinal nerves. The situation is entirely different in the limbs, and here we fortunately have ex- cellent studies of Naville (24), Béhmel (29) and Hellmich (730). THE EYE AND LENS The experimental attack upon the eye and its regenerative capacity goes back to Bonnet and Blumenbach, who recognized that a large part of the eye could be regenerated provided that a small portion remained after the original operation. Colucci (1891) showed that the lens regenerated from the edge of the iris. This paper is a remarkably well done piece of work which was com- pletely neglected until brought to attention by Emery (1897). The workers during the intervening period, however, were moved by the controversial points which were raised by Gustav Wolff (1895). His paper stressed VERTEBRATES the origin of the regenerating lens from the iris epithelium instead of the corneal epi- thelium—a distinct change from its ontog- eny. This viewpoint was a rebellion against the Haeckel dominance and was primarily directed against Darwinism as championed by Haeckel. Wolff’s findings were imme- diately in the spotlight; how much so is shown by the attachment of his name in- stead of Colucci’s to this type of regeneration. Since the conclusions were advanced against Darwinism, the cudgels of the scientists were immediately brought to the attack. The re- sults were immediately attacked and experi- ments repeated. Everyone was sceptical, but the repetition of the work and its confirma- tion by E. Miller (1896) and others, with the discovery by Emery of Colucci’s work, made Fischel’s (700) later confirmation a redundancy. Wolff showed that the dorsal border of the iris was the formative area in lens re- generation. From his viewpoint this was the most advantageous place for it to form, for as it increased in size it practically fell into place in normal location. He followed also the extrusion of pigment which occurs at the border of the iris during lens formation. Wolff's observation was an important one. He was opposed by Weismann, who took the stand that regeneration was a power acquired by natural selection. Wolff attempt- ed to show (1) that the reaction was pur- poseful, and (2) that purposefulness is a common property, primitive in nature, which can be explained neither by heredity nor by natural selection. Fischel took the view that the process had no purpose, citing the many anomalies which occurred in the nature of imperfect lenses or double formations. To him the process was a mechanical one. The limitation of regeneration to the dorsal border of the iris was explained by gravity. He invoked tur- gor acting from the region of the vitreous as inhibiting regeneration. He did prevent regeneration in one series of experiments by transplanting cornea into the eye, thus in- creasing the internal pressure. When, how- ever, he introduced foreign bodies such as bread pills and potato, his results were un- successful for lens inhibition, but were highly successful in attracting Wolff’s rid- icule. The controversy grew from the acid- ulous to the philippic, for Wolff was clever in both argument and experiment. He cut the spinal cord of his animals to inactivate them, then placed them on their backs until regeneration had occurred. Regeneration oc- 691 curred as before from the original dorsal border of the iris and Fischel’s idea of the action of gravity was exploded. In the course of these experiments he found that Triturus completed the process faster than Sala- mandra; the axolotl was very slow in re- generation if it occurred at all. Wachs (714) recalled the problems of the eye. His paper gave an excellent review of the background of the work and laid out clearly many of the problems which could be approached. He worked out the variations of the process in different species and the details of factors, such as the age of the animal at the time of regeneration, the de- gree of differentiation at the time of con- striction from the iris, and its relative size. He gave proof to Fischel’s contention that turgor was a controlling factor in regenera- tion. He supported and extended Wolff’s findings. The lens regeneration occurs with- out injury to the iris, since it occurs if the lens is removed through the roof of the mouth instead of through the cornea. When the iris is injured, regeneration is slowed and the lens may be less perfectly formed than when the iris is intact. Injury is not necessarily a factor, since the lens may be removed and replaced after which no regen- eration takes place. If a small lens from a younger animal is placed in a host, regen- eration is inhibited. The small lens grows more rapidly (younger tissue) until it reaches the size appropriate for the host eye. Wolff used grafting experiments to study the effects of the lens upon regeneration. These are partly mechanical and _ partly chemical, reacting to a stimulating chemical effect from the retina. If the lens is pushed posteriorly into the vitreous, there is an initial reaction all around the border of the iris as though regeneration were about to occur; however, it never is completed. If the lens is removed and minced lens placed in the eyeball, no regeneration occurs until the minced lens is resorbed. The first experi- ment might be due to a mechanical effect, but the second seems to have removed this completely and indicates the chemical con- trol of the reaction. If the eye receives the grafting of an extra dorsal border of the iris, it will not form a lens until the host lens is removed, but when this is done two lenses regenerate, one from the grafted iris and one from the normal one. If the dorsal border of the iris is transplanted into the vitreous humor after lens removal, it forms a lens more quickly than it would in its original location. Lenses are not formed in 692 head connective tissue, auditory vesicle, or any other region unless the retina is present. If the retina with iris is transplanted and rounds up to form a vesicle, the lens regen- erates. Sato (30, °33) showed that there is a gradient of material in the iris. Sato divided REGENERATION lens capsule is left behind. In rabbits (Ran- dolph, ’00), complete regeneration may occur after injury, but here the regeneration is from old lens epithelium, not from the bor- der of the iris. Within recent years studies have appeared (Harrison ’29, °33; Twitty, 34; Twitty and Fig. 245. Reconstitution of the eye in Triturus taeniatus. A, Before the operation; B—F, after 2, 14, 21, 28 and 49 days. (From H. Wachs, 714.) the iris into six parts with definite locali- zation of potency. Beckwith (’27) had shown that if the eye is rotated early the choroid fissure develops ventrally. Sato turned it in later stages and the choroid fissure retains its rotated position. In this case the lens regenerates from the ventral margin of the iris opposite to the choroid fissure. If the lens is removed early and replaced by in- different ectoderm, the eye may never have a lens, in which case the vitreous humor may be both incomplete and imperfect. The majority of these experiments have been performed upon amphibians. The ex- periments have at one time or another been duplicated on teleosts, lizards and other forms. The removal of an opaque lens is a frequent operation, but in the mammals the Schwind, °31; Stone, ’30) in which, by heteroplastic transplantation, the effects of the graft upon the host and the host upon the graft have been studied. In general these studies give a mass of evidence to show that ultimately the eye tends to regulate to an average condition; a large eye on a small host tends to conform to its surroundings— it is larger than the eye removed to the host but smaller than it would have been upon the donor. Reports by Schotté (38) and his co- workers show that lens may be formed from tissues foreign to the eye. The chief outlook of this work has been not the regeneration of the lens but the transformation of other tissues into lens. This links the eye to the problem of determination. It likewise adds VERTEBRATES another link from a different angle to the chain of circumstances which tie so many ideas with the embryonic nature of the blas- tema. The eye is being used as a test for the reaction of the new forming tissue. Schotté adopts the thesis that the fate of a regenerate depends upon the inductive action of the organ used which is based on Weiss’ (26a) experiment in which the tail blastema was substituted for a leg (Guyénot and Schotté, °27). Schotté and Hummel (’39) used tissues from larvae throughout one series of experiments, working with the limb blas- tema. They made four types of transplant: (1) ectodermal cap, (2) cap plus mesen- chyme, (3) mesenchyme separated from cap, and (4) regenerating mesenchyme attached to old tissues of the limb. They used regen- erates not older than 5 to 6 days in order to have indifferent tissues. The blastemal tissue was placed in the left eye of the host after removal of the lens, while the right eye served as the control after its lens had been removed. Regeneration may occur from fragments of lens epithelium left behind and here it is exceedingly rapid. In their control series regeneration occurred in one case in two days, in another in three, both of which they attributed to lens epithelium left be- hind. Regeneration changes as modifications of the transplant occurred in 69 per cent of the xenoplastic cases and in 33 per cent of the homoplastic. Lens formation in the total series was 23 per cent, with 46 per cent occurring in the xenoplastic series. These figures Schotté and Hummel were at a loss to explain and they relied upon a statistical study which was to some extent vitiated by their footnote (4), in which they stated that the errors decreased as experimentation proceeded, which indicates that both criteria and observations of completeness of opera- tive removal were inadequate at the time of the first operation. The confusing numerical differences be- tween the homoplastic and xenoplastic com- binations seem quite significant and can be interpreted in the light of Wachs’ (714) experiments. The xenoplastic graft is much more frequently resorbed and in the case of some anuran species we know definitely that resorption takes place with extreme rapidity. Let us postulate for the sake of argument that the degenerating blastema liberates sub- stances into the eye which may stimulate lens development—the chemical factor. These substances are not liberated while the blastema is intact. Further, xenoplastic 693 tissue during degeneration sometimes forms epithelial pearls which might easily simulate the histological picture of early lens forma- tion. Schotté and Hummel may be right, but the persistence of qualifying footnotes throughout their discussion indicates that all possible factors have not adequately been weighed. Their best case does not depend upon the statistical interpretation but upon structural organization—blood and notochordal ele- ments were found in one case (tail blas- tema) surrounded by capsular fibers. Either the blastema was being transformed, or else the lens capsule was proliferating so fast that it circled the resorbing blastema before complete internal disintegration occurred. The conclusion of Schotté and Hummel that the regenerating tissues of urodeles and anurans are totipotent, in the sense that they are capable of differentiations which are normally observed only in embryonic tissues, will certainly have to await acceptance sub- ject to a more critical experimental analysis. The original work on the eye by Colucci (1891) was to determine how much regen- erative capacity the eye would show. The whole eye will regenerate provided some of it has been left behind. Regeneration may not occur if only the choroid is left behind. Fuchs (24) found that tadpoles could re- generate the whole eye from the optic stalk. Reyer (48) has brought together all the work upon the embryos and larvae of Tri- turus, studying carefully the regeneration of the lens from the dorsal iris in five age groups. The process in every case is similar although varying in minor details, and swell- ing and depigmentation occur in the older animals. The reader is referred to Reyer (54) for a complete review and discussion of the lens problem and its general relation- ship to definitive potentiality. CONCLUSION The present discussion of regeneration aims merely to point out a few of the high points of what has been done as well as to point to a few of the things which need to be done. Since the work has been centered about the limb, eye and tail these have re- ceived the greatest amount of attention. The regenerative capacities of internal organs have been scantily treated, although the recent literature on liver regeneration shows that many of the fundamental potentialities of organ-forming tissues still remain for discovery. The fragmentary work which has 694 been done on many of the other glands is a case in point. These problems are more difficult technically than the relatively sim- ple ones so far studied, but with care and patience they should yield much in the next decade. It is evident that regeneration is a com- plex primarily depending upon the two cardinal things peculiar to tissues: differen- tiation, and proliferation resulting in growth. The determination of these tissues is usually and generally uniformly controlled by po- sition in the total organism. The potential- ity of a cell, whether embryonic or blas- temal, to differentiate into either a structure harmonious to its region or the replacement of such, is the resultant of many factors such as the compatibility of the tissues, the re- lationship of the proliferating parts, and the interplay of growth inhibitors and regula- tors. The solution of this complex still re- mains to engage our study and attention. REFERENCES Avel, M., and Verrier, M. L. 1930 Un cas de ré- génération hypotypique de la patte chez Lacerta vivipara. Bull. Biol. France et Belg., 64:198-204. Barber, L. W. 1944 Correlations between wound healing and regeneration in forelimbs and tails of lizards. Anat. Rec., 89:441-453. Barbour, T., and Stetson, H.C. 1929. The squa- mation of Homoeosaurus. Bull. Mus. Comp. Zool., 69:99-104. Barfurth,D. 1891 Zur Regeneration der Gewebe. Arch. mikr. Anat., 37:406-491. 1895 Die experimentelle Regeneration iiberschiissiger Gliedmassentheile bei den Am- phibien. Roux’ Arch. Entw.-mech., 7:91-116. 1899 Eine Larve von Petromyzon planeri mit drei Schwanzspitzen. Roux’ Arch. Entw.- mech., 9:27-31. Bateson, W. 1894 Materials for the Study of Variation. Macmillan and Co., London. Beckwith, C. J. 1927 The effect of the extirpa- tion of the lens rudiment on the development of the eye in Amblystoma punctatum, with special reference to the choroid fissure. J. Exp. Zool., 29: 217-260. Beigel, C. 1910 Zur Regeneration des Kiemen- deckels und der Flossen der Teleostier. Bull. in- ternat. d.l’Acad. d. Cracovie, Ser. B., Sci. Nat., July, 1910:655-690. 1912 Regeneration der Barteln bei Silu- roiden. Roux’ Arch. Entw.-mech., 34:363-370. Biberhofer, R. 1906 Uber Regeneration bei Am- phioxus lanceolatus. Roux’ Arch. Entw.-mech., 22:15-17. Bischler, V. 1926 L’influence du squelette dans la régénération, et les potentialités des divers ter- ritoires du membre chez Triton cristatus. Rev. suisse Zool., 33:431-560. REGENERATION Bishop, G. H. 1945 Regeneration after experi- mental removal of skin in man. Am. J. Anat., 76: 153-182. Bohmel, W. 1929 Regeneration nach Entnahme von Skeletteilen beim Axolotl. Roux’ Arch. Entw.- mech., 115:464-509. Boulenger, G. A. 1888 On the scaling of the re- produced tail in lizards. Proc. Zo6l. Soc., London, 1888:351-353. Bovet, D. 1930 Les territoires de régénération; leurs propriétés étudiées par la méthode de déviation du nerf. Rey. suisse Zool., 37:83—145. Brachet, J. 1946 Localisation de la phosphatase alcaline pendant le développement des Batraciens. Experientia, 2:143. Braus, J. 1909 Gliedmassenpfropfung und Grundfragen der Skelettbildung. I. Die Skelettan- lage vor Auftreten des Vorknorpels und ihre Beziehung zu _ spateren Differenzierungen. Morph. Jahr., 39:284—430. Broussonet, M. 1786 Observations sur la régén- ération de quelques parties du corps des Pois- sons. Hist. l’Acad. Roy. des Science (de Paris), Amsterdam, 1786:684—688. Burr, H. S. 1916 The effects of the removal of the nasal pits in Amblystoma embryos. J. Exp. Zool., 20:27-57. Butler, E.G. 1933 The effects of X-radiation on the regeneration of the forelimb of Amblystoma larvae. J. Exp. Zool., 65:271-315. 1935 Studies on limb regeneration in X- rayed Amblystoma larvae. Anat. Rec., 62:295- 307. Byrnes, E. F. 1904 Regeneration of the anterior limbs in the tadpoles of frogs. Roux’ Arch. Entw.- mech., 78:171-177. Colucci, V. S. 1891 Sulla regenerazione parziale dell’ occhio nei Tritoni. Istogenesi e sviluppo. Mem. della R. Acad. Sc. Ist Bologna, Ser. 5, 7: 593-629. Cuvier, G. 1829 The Animal Kingdom. Vol. 2. G. B. Whittaker, London. Duncker, G. 1905 Uber Regeneration des Schwanzendes bei Syngnathiden. Roux’ Arch. Entw.-mech., 20:30-37. 1907 Uber Regeneration des Schwanzen- des bei Syngnathiden. Roux’ Arch. Entw.-mech.. 24:656-662. Egger, E. 1888 Ein Fall von Regeneration einer Extremitat bei Reptilien. Arbeiten zool.-zootom. Inst. Wiirzburg, 8:201-212. Emery, C. 1897 Wer hat die Regeneration der Augenlinse aus dem Irisepithel zuerst erkannt und dargestellt? Anat. Anz., 73:63-64. Fischel, A. 1900 Ueber die Regeneration der Linse. Anat. Hefte, 74:1-256. Flint, J. M. 1910 Compensatory hypertrophy of the small intestine following resection of large portions of the jejunum and ileum. Trans. Conn. State Med. Soc., 1910:283-335. Forsyth, J. W. 1946 The histology of anuran limb regeneration. J. Morph., 79:287-322. Fraisse, P. 1885 Die Regeneration von Geweben und Organen bei Wirbeltieren. Cassel, Berlin. Fritsch, C. 1911 Experimentelle Studien iiber VERTEBRATES Regenerationsvorgange des Gliedmassenskeletts der Amphibien. Zool. Jahrb. (Abt. f. allg. Zool. u. Physiol.), 30:377-472. Fuchs, F. 1924 Augenregeneration nach Entfer- nung des Bulbus bei Alytes und Bufo. Zool. Jahrb. (Abt. f. allg. Zool.), 47:121-178. Gachet, M.H. 1833 Mémoire sur la reproduction de la queue des reptiles sauriens. Actes d. 1. soc. Linnéenne d. Bordeaux, nr. VI, 7833:213-259. Gardner, W. U., and Chamberlin, T. L. 1940-41 Local action of estrone on mammary glands of mice. Yale. J. Biol. & Med., 13:461-465. Gidge, N. M., and Rose, S. M. 1944 The role of larval skin in promoting limb regeneration in adult Anura. J. Exp. Zool., 97:71-85. Godlewski, E. 1928 Untersuchungen iiber Aus- losung und Hemmung der Regeneration beim Axolotl. Roux’ Arch. Entw.-mech., 174:108-143. Goette, A. 1879 Ueber Entwickelung und Re- generation des Gledmassenskeletts der Molche. Leopold Voss, Leipzig. Grant, E.R. 1945 Rate of replacement of the sur- face epithelial cells of the gastric mucosa. Anat. Rec., 97:175-186. Grobstein, C. 1942 Endocrine and developmental studies of gonopod differentiation in certain poeciliid fishes. II. Effect of testosterone propio- nate on the normal and regenerating anal fin of adult Platypoecilus maculatus females. J. Exp. Zool., 89:305-328. 1947a Decline in regenerative capacity of the Platypoecilus maculatus gonopodium during its morphogenesis. J. Morph., 80:145— 160. 1947b The role of androgen in declining regenerative capacity during morphogenesis of the Platypoecilus maculatus gonopodium. J. Exp. Zool., 106:313-314. 1948 Optimum gonopodial morphogene- sis in Platypoecilus maculatus with constant dos- age of methyl testosterone. J. Exp. Zool., 109:215- 333}. Guyénot, E. 1927 La perte du pouvoir régén- érateur des Anoures, studiée par la méthode des heterogreffes, la notion de territoires. Rev. suisse Zool., 34:1—53. 1929 La notion de territoires en biologie. Actes de la Soc. Helvétique des Sci. Nat., Davos, part II, pp. 81-91. , and Matthey, R. 1928 Les processus ré- genératifs dans la patte postérieure du lézard. Roux’ Arch. Entw.-mech., 773:520-529. , and Ponse, K. 1930 ‘Territoires de ré- genération et transplantations. Bull. Biol. France et Belg., 64:251-287. , and Schotté, O. 1927 Greffe de régén- érat et différenciation induite. Compt. Rend. Soc. Phys. et Hist. Nat. Geneve, 44:21-23. Harrison, R.G. 1893 Ueber die Entwicklung der nicht knorpelig vorbildeten Skelettheile in den Flossen der Teleostier. Arch. mikr. Anat., 42: 248-278. 1918 Experiments on the development of the forelimb of Amblystoma, a self-differentiating equipotential system. J. Exp. Zool., 25:413-462. 695 1921 On relations of symmetry in trans- planted limbs. J. Exp. Zool., 32:1-136. 1929 Correlation in the development and growth of the eye studied by means of hetero- plastic transplantation. Roux’ Arch. Entw.-mech., 120:1-55. 1933 Heteroplastic grafting in embry- ology. Harvey Lectures, 1933-34, Series 29, pp. 116-157. Heath, H. D. 1953 Regeneration and growth of chimaeric amphibian limbs: limb regeneration and growth. J. Exp. Zool., 122:339-366. Hellmich, W. 1930 Untersuchungen tiber Her- kunft und Determination des regenerativen Ma- terials bei Amphibien. Roux’ Arch. Entw.-mech., 121:135-203. Hertwig, G. 1927 Beitrage zum Determinations- und Regenerationsproblem mittels der Trans- plantation haploidkerniger Zellen. Roux’ Arch. Entw.-mech., 777:292-316. Hoadley, L. 1928 On the localization of develop- mental potencies in the embryo of Fundulus heteroclitus. J. Exp. Zool., 52:7-44. Holtfreter, J. 1947 Neural induction in explants which have passed through a sublethal cytolysis. J. Exp. Zool., 706:197-222. Hooker, D. 1925 Studies on regeneration in the spinal cord. III. Reestablishment of anatomical and physiological continuity after transection in frog tadpoles. J. Comp. Neur., 38:315-348. 1930 Physiological reactions of goldfish with severed spinal cord. Proc. Soc. Exp. Biol. & Med., 28:89-90. 1932 Spinal cord regeneration in the young rainbow fish, Lebistes reticulatus. J. Comp. Neur., 56:277-297. Hopper, A. F. 1949a Development and regener- ation of the anal fin of normal and castrate males and females of Lebistes reticulatus. J. Exp. Zool., 110:299-319. 1949b The effect of ethynyl testosterone on the intact and regenerating anal fins of normal and castrated females and normal males of Le- bistes reticulatus. J. Exp. Zool., 111:393-413. Kammerer, P. 1905 Uber die Abhangigkeit des Regenerationsvermégens der Amphibienlarven von Alter, Entwicklungsstadium und spezifischer Grésse. Experimentelle Studie. Roux’ Arch. Entw.-mech., 79:148-180. Karczmar, A. G., and Berg, G.G. 1951 Alkaline phosphatase during limb development and regen- eration of Amblystoma opacum and Amblystoma punctatum. J. Exp. Zool., 717:139-164. Krugelis, E. J. 1950 Properties and changes of alkaline phosphatase activity during amphibian development. Compt. Rend. Lab. Carlsberg, Ser. Chim., 27:273-290. Lewis, W. H. 1912 Experiments on localization and regeneration in the embryonic shield and germ ring of a teleost fish (Fumndulus heterocli- tus). Anat. Rec., 6:325-334. Liebman, E. 1949 The leucocytes in regenerat- ing limbs of Triturus viridescens. Growth, 13: 103-118. Lillie, F. R., and Juhn, M. 1932 The physiology 696 of development of feathers. I. Growth-rate and pattern in the individual feather. Phys. Zool., 5: 124184. Lindeman, V. F. 1949 Alkaline and acid phos- phatase activity of the embryonic chick retina. Proc. Soc. Exp. Biol. & Med., 77:435-437. Litwiller, R. 1939 Mitotic index and size in re- generating amphibian limbs. J. Exp. Zool., 82: 273-286. Manner, H. W. 1953 The origin of the blastema and of new tissues in regenerating forelimbs of adult Triturus viridescens viridescens (Rafi- nesque). J. Exp. Zool., 122:229-257. Marcucci, E. 1926 Rigenerazione degli arti nei Rettili. Boll. Soc. Nat. Napoli, 38:8-19. 1930 II potere rigenerativo degli arti nei Rettili. Ricerche sperimentali sopra alcune specie de Saurii. Arch. Zool. Ital., 14:227-252. 1932 Trapianti di pelle e rigenerazione in Lacerta muralis. Arch. Zool. Ital., 17:435-447. Matsumoto, S. 1918 Contribution to the study of epithelial movement. The corneal epithelium of the frog in tissue culture. J. Exp. Zool., 26:545—- 563. Maximow, A. A., and Bloom, W., 1952 A Text- book of Histology, 6th ed. W. B. Saunders Co., Philadelphia. May, R. M. 1927 1re Thése: Modifications des centres nerveux dues a la transplantation de l’oeil et de Vorgane olfactif chez les embryons d’an- oures. 2e Thése: Propositions données par la Faculté. (Théses presentées a la Faculté des Sciences de Paris.) Arch. de Biol., 37:337-395. Milojevic, B. D. 1924 Beitrage zur Frage iiber die Determination der Regenerate. Roux’ Arch. Entw.-mech., 703:80-94. Moog, F. 1944 Localizations of alkaline and acid phosphatases in the early embryogenesis of the chick. Biol. Bull., 86:51-80. Morgan, T. H. 1900 Regeneration in Teleosts. Roux’ Arch. Entw.-mech., 70:120-134. 1902 Further experiments on the regen- eration of the tail of fishes. Roux’ Arch. Entw.- mech., 174:539-561. Morrill, C. V. Jr., 1906 Regeneration of certain structures in Fundulus heteroclitus. Biol. Bull., 12:11-20. 1918 Some experiments on regeneration after exarticulation in Diemyctylus viridescens. J. Exp. Zool., 25:107-126. Miiller, E. 1896 Uber die Regeneration der Augenlinse nach Exstirpation derselben bei Tri- ton. Arch. f. mikr. Anat., 47:23-33. Miiller,H. 1864 Uber die Regeneration der Wir- belsaule und des Riickenmarkes bei Tritonen und Eidechsen. Frankfurt a.M. Nabrit, M. S. 1938 Regeneration in the tail fins of embryo fishes (Opsanus and Fundulus). J. Exp. Zool., 79:299-308. Nassonoy, N. V. 1930 Die Regeneration der Axolotlextremitaten mach Ligaturanlegung. Roux’ Arch. Entw.-mech., 121:639-657. Naville, A. 1924 Recherches sur l’histogenése et la régénération chez les Batraciens anoures (Corde dorsale et téguments). Arch. de Biol., 34:235-343. REGENERATION Needham, J. 1942 Biochemistry and Morpho- genesis. Cambridge University Press, London. Nicholas, J. S. 1927 The application of experi- mental methods to the study of developing Fundulus embryos. Proc. Nat. Acad. Sci., 13:695- 700. , and Oppenheimer, J. M. 1942 Regula- tion and reconstitution in Fundulus. J. Exp. Zool., 90:127-157. Noble, G. K., and Bradley, H. T. 1933 The rela- tion of thyroid and the hypophysis to the molting process in the lizard (Hemidactylus brooki1). Biol. Bull., 64:289-298. , and Clausen, H. J. 1936 Factors con- trolling the form and color of scales on the regen- erated tails of lizards. J. Exp. Zool., 73:209-229. Nusbaum, J., and Sidoriak, S. 1900 Beitrage zur Kenntnis der Regenerationvorgange nach kiin- stlichen Verletzungen bei alteren Bachforellen- embryonen (Salmo fario L.). Roux’ Arch. Entw.- mech., 10:645-684:. Nussbaum, M. 1886-1887 Die Teilbarkeit der lebenden Materie. Arch. mikr. Anat., 26 (1886): 485-538; 29 (1887): 265-366. Oppel, A. 1912 Causalmorphologische Zellen- studien. V. Die aktive Epithelbewegung, ein Fak- tor beim Gestaltungs- und Erhaltungsgeschehen. Roux’ Arch. Entw.-mech., 35:371-456. Oppenheimer, J. 1936 Processes of localization in developing Fundulus. J. Exp. Zool., 73:405- 444. Peters, A. 1885 Uber die Regeneration des Epithels der Cornea. Inaugural Dissertation, Bonn. Philippeaux, J. M. 1867 Régénération des mem- bres chez l’Axolotl et la Salamandre. Compt. Rend. Acad. Paris, 1867, pp. 1162-1163. Podwyssozki, W., Jr. 1881 Experimentelle Un- tersuchungen iiber die Regeneration der Driisen- gewebe. II. Die Regeneration des Nierenepithels, der Meibom’schen Driisen und der Speicheldrii- sen. Beitrage zur path. Anat. u. Physiol., 2:1-28. Poynter, C. W.M. 1922 The effects of ultraviolet rays on developing mollusks (Limnaeus). Anat. Rec., 23:32. Przibram, H. 1926 Regeneration und Trans- plantation bei Tieren; in Handb. der norm. u. pathol. Physiologie, Vol. 14, part I, pp. 1080-1113. Puckett, W. O. 1934 The effects of X-radiation on limb development in Amblystoma. Anat. Rec., 58:32-33. 1936 The effects of X-radiation on limb development and regeneration in Amblystoma. J. Morph., 59:173-213. Randolph, R. L. 1900 The regeneration of the crystalline lens. Johns Hopkins Hosp. Rpts. (Contr. to Med. Sci. . . . dedicated to William Henry Welch . . .), 8:237-263. Reyer, R. W. 1948 An experimental study of lens regeneration in Jriturus viridescens viridescens. I. Regeneration of a lens after lens extirpation in embryos and larvae of different ages. J. Exp. Zool., 107:217-268. 1954 Regeneration of the lens in the am- phibian eye. Quart. Rev. Biol., 29:1-46. VERTEBRATES Richardson, D. 1945 Thyroid and pituitary hor- mones in relation to regeneration. 2. Regenera- tion of the hind leg of the newt, Triturus viri- descens, with different combinations of thyroid and pituitary hormones. J. Exp. Zool., 700:417- 429. Rose, S. M. 1944 Methods of initiating limb re- generation in adult Anura. J. Exp. Zool., 95:149- 170. 1945 The effect of NaCl in stimulating regeneration of limbs of frogs. J. Morph., 77:119- 140. 1948 Epidermal dedifferentiation during blastema formation in regenerating limbs of T7i- turus viridescens. J. Exp. Zool., 108:337-361. Samuel, S. 1870 Die Regeneration. Virchow’s Arch. f. pathologische Anat. u. Physiol. u. f. klinische Medicin, 50:323-354. Sato, T. 1930 Beitrage zur Analyse der Wolff’- schen Linsenregeneration, I. Roux’ Arch. Entw.- mech., 722:451-493. 1933 Beitrage zur Analyse der Wolff’- schen Linsenregeneration, II. Roux’ Arch. Entw.- mech., 730:19-78. Schotté, O. E. 1938 Induction of embryonic or- gans in regenerates and neoplasms. Collecting Net, 73:1-6. —, and Hall, A. B. 1952 Effect of hypo- physectomy upon regeneration in progress (7'ri- turus viridescens). J. Exp. Zool., 121:521-556. , and Hummel, K.P. 1939 Lens induction at the expense of regenerating tissues of amphib- ians. J. Exp. Zool., 80:131-165. Schultze, D. 1894 Die kimstliche Erzeugung von Doppelbildungen bei Froschlarven mit Hilfe abnormer Gravitatswirkung. Roux’ Arch. Entw.- mech., 7:269-305. Scott, G. G. 1907 Further notes on the regenera- tion of the fins of Fundulus heteroclitus. Biol. Bull., 72:385—-400. Singer, M. 1942 The nervous system and regen- eration of the forelimb of adult Triturus. I. The role of sympathetics. J. Exp. Zool., 90:377-399. 1945 The nervous system and the regen- eration of the forelimb of adult Triturus. III. The role of the motor supply. J. Exp. Zool., 98:1-21. 1946a The nervous system and regenera- tion of the forelimb of adult Triturus. IV. The stimulating action of a regenerated motor supply. J. Exp. Zool., 701:221-239. 1946b The nervous system and regenera- tion of the forelimb of adult Triturus. V. The in- fluence of number of nerve fibers including a quantitative study of limb innervation. J. Exp. Zool., 101:299-337. 1947a The nervous system and regenera- tion of the forelimb of adult Triturus. VI. A fur- ther study of the importance of nerve number, including quantitative measurements of limb in- nervation. J. Exp. Zool., 104:223-249. 1947b The nervous system and regen- eration of the forelimb of adult Triturus. VII. The relation between number of nerve fibers and surface area of amputation. J. Exp. Zool., 104: 251-265, 697 1949 The invasion of the epidermis of the regenerating forelimb of the urodele, Tri- turus, by nerve fibers. J. Exp. Zool., 777:189-209. , and Egloff, F. R. L. 1949 The nervous system and regeneration of the forelimb of adult Triturus. VIII. The effect of limited nerve quan- tities on regeneration. J. Exp. Zool., 177:295— 314: Slotopolsky, B. 1921-22 Beitrage zur Kenntnis der Verstiimmelungs- und Regenerationsvorgange am Lacertilierschwanze. Zool. Jahrb., 43:219- B22: Spallanzani, L. 1768 Prodromo di _ un’opera sopra le riproduzioni animali. Modena. (Math. physikal. Abhandl., Leipzig, 1769). Stone, L. S. 1930 Heteroplastic transplantation of eyes between the larvae of two species of Am- blystoma. J. Exp. Zool., 55:193-261. 1952 An experimental study of the in- hibition and release of lens regeneration in adult eyes of Triturus viridescens viridescens. J. Exp. Zool., 121:181-223. Strasser, H. 1879 Zur Entwicklung der Extrem1- tatenknorpel bei Salamandern und _ Tritonen. Morph. Jahrb., 5:240-315. Studnicka, F. K. 1912 Uber Regenerationser- scheinungen im caudalen Ende des Koérpers von Petromyzon fluviatilis. Roux’ Arch. Entw.-mech., 34:187-238. Sumner, F. B. 1904 A study of early fish devel- opment: experimental and morphological. Roux’ Arch. Entw.-mech., 17:92-149. , and Collins, H.H. 1918 Autotomy of the tail in rodents. Biol. Bull., 34:1-6. Swett, F.H. 1924 Regeneration after amputation of abnormal limbs in Amblystoma. Anat. Rec., 27:273-288. Sziits, A. von 1914 Beitrage zur Kenntnis der Abhangigkeit der Regeneration vom Zentral- nervensystem. Roux’ Arch. Entw.-mech., 38:540- 545. Thornton, C.S. 1938a The histogenesis of muscle in the regenerating forelimb of Amblystoma punctatum. J. Morph., 62:17-47. 1938b The histogenesis of the regener- ating forelimb of larval Amblystoma after exar- ticulation of the humerus. J. Morph., 62:219- 241. 1942 Studies on the origin of the regen- eration blastema in Triturus viridescens. J. Exp. Zool., 89:375-389. Tornier, G. 1906 Der Kampf der Gewebe im Regenerat bei Begiinstigung der Hautregenera- tion. Roux’ Arch. Entw.-mech., 22:348—369. Turner, C. L. 1947 The rate of morphogenesis and regeneration of the gonopodium in normal and castrated males of Gambusia affinis. J. Exp. Zool., 106:125-143. Twitty, V.C. 1930 Regulation in the growth of transplanted eyes. J. Exp. Zool., 55:43-53. 1934 Growth correlations in Amphibia studied by the method of transplantation. Cold Spring Harbor Symposia on Quantitative Biology, 2:148-156. , and Schwind, J. L. 1931 The growth of 698 eyes and limbs transplanted heteroplastically be- tween two species of Amblystoma. J. Exp. Zool., 59:61-86. Vallette, M. 1926 Mécanisme de la régénération du museau chez les Urodéles. Arch. Sci. phys. et. nat., Géneve, 8:28—32. Wachs, H. 1914 Neue Versuche zur Wolffschen Linsenregeneration. Roux’ Arch. Entw.-mech., 39:384451. Weiss, P. 1923 Die Transplantation von entwick- elten Extremitaten bei Amphibien. I. Morphol- ogie der Einheilung. Roux’ Arch. Entw.-mech., 99:150-167. 1925a Unabhangigkeit der Extremita- tenregeneration vom Skelett (bei Triton cris- tatus). Roux’ Arch. Entw.-mech., 104:359- 394, 1925b Die seitliche Regeneration der Urodelenextremitat. Roux’ Arch. Entw.-mech., 104:395-408. 1926a Physiologie der Formbildung (Entwicklung und Regeneration). Jahresbericht ges. Physiologie f. d. Jahr 1926, pp. 77-112. 1926b Ganzregenerate aus halbem Ex- tremitatenquerschnitt. Roux’ Arch. Entw.-mech., 107:1-53. 1927 Die Herkunft der Haut im Ex- REGENERATION tremitatenregenerat. Roux’ Arch. Entw.-mech., 109:584—610. 1930 Potenzpriifung am Regenerations- blastem. II. Das Verhalten des Schwanzblastems nach Transplantation an die Stelle der Vorderex- tremitat bei Eidechsen (Lacerta). Roux’ Arch. Entw.-mech., 722:379-394. Weissfeiler, J. 1924 Régénération du cerveau et du nerf olfactif chez les batraciens urodéles. Rev. suisse Zool., 32:1—44. Wendelstadt, H. 1904 Experimentelle Studie liber Regenerationsvorgange am Knochen und Knorpel. Arch. mikr. Anat., 63:766—795. Werber, I. 1905 Regeneration der Kiefer bei der Eidechse Lacerta agilis. Roux’ Arch. Entw.-mech., 19:248-258. Werner, F. 1896 Uber die Schuppenbekleidung des regeneration Schwanzes bei Eidechsen. Sitz- ber. Ak. Wiss. Wien, math.-nat. K1., 705:123-146. Wolff, G. 1895 Entwicklungsphysiologische Stu- dien. I. Die Regeneration der Urodelenlinse. Roux’ Arch. Entw.-mech., 7:380-390. Woodland, W. N. F. 1921 Some observations on caudal autotomy and regeneration in the Gecko (Hemidactylus flaviviridis, Riipel) with notes on the tails of Sphenodon and Pygopus. Quart. J. Micr. Sci., 65:63-100. Section XIV TERATOGENESIS EDGAR ZWILLING INTRODUCTION Previous chapters in this book have demon- strated that normal development depends on a harmonious sequence of closely inter- dependent events. Such development is the end product of the expression of the intrinsic potentialities of cells or groups of cells as conditioned and modified by their relation- ship to each other and to the rest of the embryo. What the student of embryology sees under ordinary circumstances are the visible results of these expressed potencies. Prior to this there are intracellular re- arrangements and localization of formative materials. As development progresses masses of tissues migrate from one region of the embryo to another. Having arrived at their destinations they in turn may influence ad- jacent tissues and initiate definite develop- mental tendencies. In this manner a pattern of organ-forming regions is laid out. Within each of these regions, then, we have seen how similar dependencies exist in which later parts require the influence of adjacent regions for their normal formation. Subse- quent to the elaboration of the basic pattern for a structure there is a period which is characterized by cellular differentiation and growth. All of these events presumably are consequences of prior chemical and physical processes. In a situation where such a vast array of orderly interactions must occur in order that a normal structure be formed it is not surprising that deviations from the normal are frequently encountered. Dis- turbances in either the spatial or the tem- poral synchronization of the many develop- mental interrelationships may lead to abnor- mal individuals and structures. Many minor deviations which occur are considered to be within the range of normality. Only the more extreme deviants are regarded as aber- rant. Variants from the normal, monstrosities of every sort, have fascinated students of biology and medicine from ancient times to the present. Many superstitions have been built around them and they have occasioned much fantasy and speculation. However, to modern students of embryology these terata (the name given the abnormal individuals) provide further material for the study of development. Not only has the application of principles derived from other experimen- tal procedures been instructive in explaining the nature of terata, but the terata themselves have been utilized in elucidating normal developmental relationships. Hypotheses concerning the causation of human terata have been numerous and have included such diverse agencies as celestial influences, gods, devils, hybrids (i.e., human < some other animal), maternal impressions and imagination, mechanical pressure (extra- abdominal, intra-abdominal and amniotic), amniotic and umbilical strangulation, dis- ease, faulty implantation, etc. [For an ex- cellent account of the older history the reader is referred to Ballantyne (’04) and Schwalbe (06-37).] Recent years have witnessed a revival of interest in these problems in medical circles, partly as the result of an increased attention to inherited anomalies but largely because of the discoveries con- cerning the possible role of viruses, blood factor antagonisms, nutritional factors and radiations in the etiology of congenital mal- formations (for recent bibliography see Gruenwald, ’47; Morison, ’52). Embryological studies must eventually provide the basis for an analysis of the prob- lems in human teratology. Since most in- vestigations of the experimental production of monsters have been undertaken on inver- tebrates and lower vertebrates one of the problems has been to what extent the ob- servations and conclusions derived from these studies may be applicable to humans. In- creasing evidence indicates that no unique principles apply to mammals—that any observed differences result from differences 699 700 in the details of development and the mode of nutrition found in placental embryos. We feel, therefore, that until more controlled and reliable data on human material are forthcoming one must accept the generali- zations derived from studies of other forms. What, then, are the major problems in teratology? Briefly they may be separated into two classes: (1) What are the causal agents? and (2) How do these agents produce their effects? THE CAUSAL AGENTS In their quest for information about the etiology of congenital malformations many of the teratologists of the past have become strong proponents of either environmental or hereditary factors as the exclusive cause of the abnormalities. Mall (’08), one of the most influential American teratologists of his time, emphasized that all anomalies were the result of external influences on normal ova. These abnormal influences were caused by faulty nutrition, the consequence of poor implantation. Streeter (730, 31), on the other hand, has placed extreme emphasis on the importance of genetic factors not only for the embryonic phases of development but for all subsequent stages of life; everything depended on one’s being a “good egg” to start with. Observations of the past few decades indicate quite strongly that a com- promise between the two is more represen- tative of the facts—that not only may anomalies be mediated by both hereditary and environmental factors, but similar kinds of anomalies may be produced by either. Hereditary Factors. A wide variety of con- genital anomalies, in many kinds of animals, has been found to be transmitted from gen- eration to generation in a regular mendelian fashion. There is no doubt about the genetic basis for these conditions [the reader is re- ferred to Griineberg (752), Gruenwald (747), and Landauer (51) for bibliographies on this subject]. The mode of inheritance is varied. Characters may be transmitted as simple re- cessives, i.e., the condition is not expressed unless the individual receives a mutant gene for the character from each parent. Other characters may be transmitted as dominants. In this situation a single factor, from one parent, will produce the anomaly. The actual expression of either type of inheritance may vary from relatively minor to severe and often lethal maldevelopment. Many of the dominant mutations are pe- TERATOGENESIS culiar in that the presence of two mutant genes (homozygotes) produces effects which are more severe than found in heterozygotes. Heterozygotes of the Creeper mutation in fowl (Dunn and Landauer, ’26; Landauer and Dunn, °30; Landauer, ’32) produce a characteristic chondrodystrophy. The long bones, especially of the legs, are dispropor- tionately shortened. However, most of the heterozygotes are quite viable. Homozygotes, on the other hand, never survive to hatching. Most of them are retarded and die during the third or fourth day of development. Some survive to later stages and show a typi- cal phokomelic condition—extremely short legs, deformed beaks and eyes and a general dwarfing. Another example of this sort is the muta- tion Kinky in the mouse (Caspari and David, 40). In the heterozygotes the only expression is an absence or fusion of tail vertebrae. Gluecksohn-Schoenheimer (’49a) has shown that the homozygotes are, for the most part, lethal and evince, before death, a wide variety of duplication and twinning. That this may be a more severe and possibly early expression of a similar tendency in hetero- zygotes is indicated by forking of the distal part of the tail and occasional occurrence of duplication of the vagina in the latter. Lethal effects, however, are not invariable consequences of homozygosity in dominant mutations. In many cases the heterozygous and homozygous conditions are distinguish- able only by breeding tests. On the other hand there are mutations like Yellow (Rob- ertson, 42) and the spotting genes in mice (Russell and Fondal, 51) in which there are relatively minor effects on pigment or color pattern in heterozygotes but in which the homozygous condition leads to severe and frequently lethal defects. Essentially the same syndrome of effects may be produced in a given species by en- tirely independent recessive or dominant mu- tations. The type of inheritance cannot be determined from the appearance of the in- dividual animals. Conditions similar to the chondrodystrophy produced by the Creeper factor have been caused by at least five dis- tinct recessive lethals in fowl (Landauer, ’35; Asmundson, ’39, 42; Landauer, 41; Lamor- eux, 42; Hays, 44). Likewise rumplessness may result, in fowl, from the action of two independent mutations—one dominant (Dunn, ’25), the other recessive (Landauer, 45a). While the recessive rumpless chickens may have other associated anomalies (acces- TERATOGENESIS sory ribs, scoliosis, lordosis), frequently they cannot be distinguished from the dominant rumpless by appearance alone. The genetic situation is further compli- cated in that both the “penetrance” (i.e., incidence of effected individuals) and the “expressivity” (degree of the effect) may be altered by both genetic and environmental factors. Both of the rumpless mutations (Dunn and Landauer, °34, °36; Landauer, ’45a) may be, by proper selection experi- ments, modified strongly so that there is a high incidence of “normal” tails even though the mutant genes are present. Embryological studies (Zwilling, 45a) demonstrate that minor anomalies, which have no permanent morphological effect, occur early in develop- ment of some genetic rumpless chicks. Such altered expression of a mutation is a result of the genetic background upon and with which a given gene must operate. In some instances (Landauer, *33) only a single modi- fying gene may be responsible, in others it is evident that a complex of modifying fac- tors is involved. Occasionally, as in species or strain crosses, the modifying backgrounds are so diverse that the entire pattern of dominance may be reversed. Reed (737) has reported a dominant mutation in Mus mus- culus which causes anomalies which vary from fusion of vertebrae to absence of the tail (ribs may be absent or fused). When the Fused gene is introduced into another species, Mus bactrianus, it behaves as a recessive. Many other examples of such phenomena could be cited. Extra-genetic factors may alter the expres- sion of a given mutant. Some of the early experiments with Drosophila (Goldschmidt, 38) have shown that the expression of many mutants may be altered by increased or low- ered temperatures at critical stages. The an- alyses of Wright (’34) indicate that such factors as age of mother and season (which seem to affect the condition of the mother) may influence the expression of mutant genes in mammals. In view of the variable situa- tions produced by both genetic background and extra-genetic factors it is not surprising that there are many cases, especially in hu- man heredity, in which the mode of trans- mission of a suspected hereditary character is not certain. Environmental Factors. Virtually every en- vironmental factor has been instrumental in producing anomalies in some organism or other provided that the factor is modified in the proper way and at the proper time. Tem- 701 perature variations, mechanical disturbances (vibration, pressure), a host of chemicals, irradiations with hard rays (x-rays) and ultraviolet and modification of the gaseous environment have all been shown to produce effects when properly applied. [Gruenwald (47) has compiled most of the important references to that date. | During the early part of the present cen- tury efforts were made to relate more di- rectly with mammalian terata the results of experiments on anamniotes. For a long time these attempts were equivocal. In most in- stances it was demonstrated that toxic sub- stances (heavy metals, etc.) increased the abortion rate, but not, at least to any great extent, the rate of production of monsters. Essentially the same conclusions were reached when surveys were made of human alco- holics or of people engaged in occupations involving exposure to toxic compounds. In recent years the unequivocal relation to mam. malian terata of at least four factors has been demonstrated: 1. Considerable evidence has accumulated which shows that dietary deficiencies of preg- nant mothers may be responsible for anoma- lous development in their young. Much of these data are in the agricultural literature. One of the most conclusive analyses has been contributed by Warkany and Schraffenberger (43). These authors produced a syndrome, including micromelia, in rats by means of a riboflavin-deficient diet. All symptoms were eradicated by replacement therapy if the riboflavin-containing diet was fed prior to the thirteenth day of gestation. 2. A number of authors have described the effects of irradiations (mostly x-rays) on the development of mammals. Russell (°50) has published a thorough analysis of the effects on mouse embryos of whole body x-irradia- tion of the mother. Dose and time after copulation were varied. The data were an- alyzed in terms of the incidence of various anomalies as related to the time of irradia- tion. Pre-implantation stages were non-re- active. In implanted embryos the incidence of most abnormalities rose gradually, reached a peak, then fell off abruptly. A somewhat different situation was revealed (Wilson, Brent and Jordan, *53) in a less extensive study with rats. These authors irradiated only the exteriorized embryos. There was an abrupt transition: 8-day embryos were merely retarded in growth while those exposed on the ninth and tenth days showed a number of anomalies (see also p. 715). 702 3. It has been shown that blood factor incompatibilities between a mother and a fetus may result in abnormal development. The best known case is that involving Rh factors [see Levine (’48) for review of the literature]; in this situation Rh positive blood from a fetus may produce antibodies in an Rh negative mother which may in turn have a deleterious effect on the fetus. In practice it is found that subsequent fetuses are more severely affected and have more severe symptoms of erythroblastosis. (Note that when several such cases occur in one family the records may resemble those for direct inheritance, yet these conditions arise only indirectly through the blood factor in- compatibilities. ) 4. Considerable evidence has been pre- sented (Gregg, ’41; Erickson, ’44) which in- dicates that various congenital anomalies (cataract, microcephaly, heart disease, den- tal defects, etc.) occur after women contract rubella (German measles) during the early months of pregnancy. It has been assumed by most people that the virus crosses the placenta and attacks the fetal tissues directly. Hamburger and Habel (’47) have shown that some viruses may produce teratological effects when applied directly to a chick em- bryo. Gillman et al. (48) have produced a number of congenital anomalies (hydroceph- alus, cleft palate, eye and tail defects, etc.) by injecting trypan blue into rats during pregnancy or prior to conception. The dye itself does not reach the embryo; instead, ac- cording to these authors, it produces meta- bolic disturbances in the mother which in turn affect the developing embryos. These authors believe that it is not necessary for viruses or other disease-causing organisms to act directly on the fetus. They may produce their effects indirectly, through an accumula- tion in the blood of products of disturbed maternal metabolism. This concept has far reaching implications for further studies on the etiology of human congenital malforma- tions. Two generalizations from these studies of environmental effects should be emphasized at this time. (1) Essentially the same effects may be produced by a variety of seemingly unrelated treatments. (2) In many instances the morphological consequences of environ- mental interventions are essentially the same as those caused by genetic factors. Phenocopies. When experimental treatment of a genetically normal embryo modifies its development so that its final appearance du- plicates that of a mutant of the same species TERATOGENESIS it is called a phenocopy (Goldschmidt, ’38). Goldschmidt was the first to recognize that phenocopies could be of great value in elu- cidating gene action. His comparison of the effects of temperature variation on wing de- velopment with mutant wing conditions in Drosophila is now classic. A high incidence of a particular type of wing defect was ob- tained when the larvae were subjected to the proper temperature at the proper time (i.e., phenocritical period). It was reasoned that the temperature alteration modified the rate of processes important for wing develop- ment, that these were probably the same processes disturbed by mutant genes which produced wing defects, and that the pheno- critical period for a given condition was the same as the period when the gene producing a similar condition was operative. This reasoning is very suggestive, but one must, in order to have valid conclusions of this sort, verify each step. Even in Drosophila Henke, Fink and Ma (’41) have shown that phenocopies may be produced by treating larvae at periods other than the critical one for the mutant copied. An example from vertebrate material may be used to illustrate some of the problems in studies of pheno- copies. Tail reduction occurs in different ways in the two rumpless mutations of chick- ens. Presumptive tail tissue degenerates (Zwilling, °42), in dominant rumpless em- bryos, prior to its incorporation into the tail (i.e., during the second and third days). The degree of tail reduction is correlated with the amount of degenerate tissue. In recessive rumpless embryos (Zwilling, °45a) degen- eration does not occur until after rather ab- normal tail structures have formed. Abnor- mal tail morphogenesis may occur at any stage between the third and sixth days and the degree of tail reduction is correlated with the time when the abnormalities appear. Structures proximal to the involved regions continue to develop normally.* Phenocopies * Several mutations in mice cause taillessness of one degree or another. Chesley (’35) and Glueck- sohn-Schoenheimer (’49b) have described the devel- opmental events leading to tail defects in a number of them. In one case (Brachyury heterozygotes) the tail may develop normally at first and then de- generate distal to a given point. In other genetic combinations (T/t°, T/t!) the tail may, at first, form normally except that the notochord is missing. After a while the tail in these degenerates com- pletely. Here we see at least two more genetic mechanisms which result in tail loss. At least two groups (Hamburgh, 52; Waddington and Carter. 52) are studying tail defects (phenocopies) which follow injection of trypan blue into pregnant mice, TERATOGENESIS of rumplessness occur sporadically (Landauer and Dunn, ’25), may be produced by me- chanical jarring (Landauer and Baumann, 43), by transection of the posterior part of an early embryo’s body (Zwilling, ’45b), by local irradiation (Wolff, 736) and by topical application of several chemicals (Ancel, ’50). After the discovery (Landauer, *45b) that a high incidence of rumplessness is a consistent consequence of injection of insulin into the yolk sac of early embryos (prior to 72 hours) a series of studies was undertaken to estab- lish any relationship between these pheno- copies and the mutants. Moseley (’47) made a very careful study of the morphology of insulin-treated embryos at various stages. Roughly 20 per cent of her rumpless embryos resembled the recessive condition. The majority of the rumpless em- bryos (2/3), however, achieved the tailless condition in a manner which was unlike that seen in either of the mutant types. Tail de- fects in this large group followed an abnor- mal deviation of caudal structures which forced them into the cloaca where, after a variable time, they degenerated. Only an embryological study could reveal these facts, since an examination of advanced embryos or hatched chicks merely indicated that there was a marked resemblance to the mutant types. Evidence (Landauer and Rhodes, *52) indicates that insulin rumplessness is medi- ated via an interference with anaerobic gly- colysis. Pyruvic acid, injected somewhat before or simultaneously with insulin, mark- edly decreased the incidence of rumplessness. Mortality was lowered. Similar treatment of embryos of both mutant strains (Landauer, 54) had no effect on incidence of rumpless- ness in either. Embryological study thus re- veals that the final adult phenocopy condition may be reached by morphogenesis which is unlike that found in mutant forms. The work by Landauer indicates that even the morpho- logical similarity between some of the in- sulin-treated embryos and the recessive mutants is probably preceded by different metabolic disturbances. The extent, there- fore, to which phenocopies can be used in elucidating the action of genetic factors is limited. It must be borne in mind that no a priori conclusions can be drawn about gene action from the mere production of a phe- nocopy. One must, without additional evi- dence, be literal and insist that, by defini- tion, only the appearance of a mutant has been duplicated. These points will be stressed again below in our discussion of micromelia. 703 HOW AGENTS PRODUCE THEIR EFFECTS To say that genetic or environmental fac- tors are the causal agents of anomalies does not, however, explain how these agents pro- duce their effects. In the absence of more specific information rather general hypoth- eses have been advanced to explain the action of agencies which produce terata. One of the oldest of these (Ballantyne credits Harvey’s work of 1651 with its foundation) is the concept of arrested development. In- itially this hypothesis was utilized to de- scribe the fact that in many anomalies development seems to have stopped at an early stage in the formation of a structure and that primitive features are retained; for example, a cleft palate results from a failure of the two palatal primordia to fuse and, thereby, an early embryonic condition per- sists. Later authors have extended the im- plications of this concept to include situations in which one does not necessarily find per- sistent embryonic conditions. In such in- stances the anomaly is supposedly preceded by temporary arrest at some stage and this arrest is the prime contributing factor to the subsequent abnormal development. Follow- ing the St. Hilaires and Dareste, Stockard (21) was one of the strongest proponents of this concept. He argued that a slowing or virtual stopping of development (by way of low temperature, oxygen lack or various chemical and physical interventions) was the primary result of the treatment and that the type of deformity depended on the time in development when this occurred. E. Wolff (48) considers arrests of development of suffi- cient importance as a first step in production of terata that he postulates this as the first of his laws of teratogenesis. Attempts to establish the mechanism of action of teratogenic agents, genetic or other- wise, have been very fruitful on the morpho- logical level. Careful embryological studies have provided considerable information which is basic for causal analyses. In some cases unsuspected relationships have been revealed. Griineberg’s (738) grey lethal mu- tation in rats is a case in point. Skeletal anomalies, excess of erythrocytes and hemo- globin, heart enlargement, lung emphysema, etc., are all related by means of a “pedigree of causes” to an early anomaly of cartilage. However, just as in normal development, there is a deficiency in our knowledge of the metabolic events (and possibly unknown cellular relationships) which cause the visi- 704 ble morphogenetic interactions. We know little or nothing about the events which lead to the initial cartilage abnormalities in the grey lethal rat, which cause presumptive tail tissue to degenerate in rumpless chicks, etc. Most authors still accept Stockard’s (’21) four principles as explanations of causation. These are, in his own words: TERATOGENESIS Probably the most completely studied tera- tological condition is that of micromelia (leg shortening) in chickens. It may be profitable to discuss this condition in some detail to evaluate possible causal relations. Chondrodystrophy-like micromelias occur frequently. They are found in many animals besides chickens (man, cattle, dogs, rabbits, Fig. 246. A, An example of inherited chondrodystrophy in a 21-day chick embryo (courtesy of Dr. W. F. Lamoreux, Cornell University, and the editors of the Journal of Heredity). B, A markedly similar condition in an embryo from a hen which was fed a biotin deficient diet (courtesy of Dr. J. R. Couch and the editors of the Anatomical Record). 1. “. . . Every type of developmental mon- ster known in the literature may be produced by one and the same experimental treatment. 2. “. . . The same structural abnormality may be induced in the embryos of various species by a great number of different ex- perimental treatments. 3. “. . . In all cases the initial effect of the experimental treatment is a lowering of the developmental rate, and the resulting deformity is always secondarily due to this slow rate of development. 4. “. . . The type of monster or deformity is determined by the developmental period during which the slowing in rate is experi- enced.” These principles preclude specificity of action of the agents. They also preclude specific tissue or rudiment requirements which, in the presence of metabolic disturb- ances, may result in an anomalous structure. salamanders). The condition is characterized by a disproportionate shortening of the Jong bones of the extremities, due largely to growth retardation which is a consequence of hypoplasia of epiphyseal cartilages. Fre- quently the chondrocranium, especially the jaw elements, is also distorted. As men- tioned above, there are a number of inde- pendent mutations in domestic fowl which are responsible for this type of condition. Similar chondrodystrophy-like syndromes oc- cur sporadically (Landauer, ’27) and have been produced experimentally in chickens. Byerly et al. (35) reported the development of short-legged embryos from eggs of chick- ens on a deficiency diet which could be mitigated by wheat germ. A similar anomaly was reported by Lyons and Insko (37); in their case the dietary imbalance was re- versed by manganese supplements. Roman- off and Bauernfeind (’42) demonstrated that TERATOGENESIS a riboflavin deficiency caused micromelia. Essentially similar abnormalities were pro- duced by biotin deficiencies by Couch et al. (48). These were eliminated by injections of biotin into the deficient eggs. Essentially the same syndrome has been produced by sulfa- nilamide and eserine sulfate (Ancel, ’45a), by insulin (Landauer, ’47a), thallium (Kar- nofsky, Ridgeway and Patterson, ’50), boric acid (Landauer, 52) and pilocarpine (Lan- dauer, 53). With all of these substances, in- jected at the proper time, both the extremi- ties and the head are affected and the limbs of both sides are involved symmetrically (there are frequently other associated anoma- lies but these will not be discussed here). Usu- ally the most severe cases of micromelia are found in embryos which are quite dwarfed. Even though a similar syndrome is pro- duced by all of these treatments a number of significant facts are revealed by a close inspection of the data. The critical period is different for different substances. Insulin must be injected at 120 hours (Landauer, 47a), sulfanilamide at 48 hours (Zwilling and DeBell, 50) and boric acid at 96 hours (Landauer, 52) for maximum incidence and severity of micromelia. This does not neces- sarily mean that sulfanilamide and_ boric acid are inactive until a later stage. Rather the evidence points to their continued activ- ity over a period of time which, for maxi- mum effect, must include the earlier stages. While the micromelias produced by the various substances are superficially similar, a detailed analysis reveals that there are differences in detail. With insulin (Land- auer, 54) and sulfanilamide (Zwilling and DeBell, *50) all of the long bones of the leg are shortened and the tibiotarsus shows the greatest relative reduction. With boric acid (Landauer, *52) and pilocarpine (Landauer, 54) the femur and tibia are relatively nor- mal but the tarsometatarsus is greatly re- duced. There are, in addition, differences in digital involvement; insulin and_ sulfanil- amide cause little or no toe reduction while boric acid does result in shortening or ab- sence of the toes. Although differences in incidence may result from injecting these substances at times other than the optimum, the morphological details remain constant for a given chemical. The various dietary deficiencies which re- sult in micromelia involve substances (bi- otin, riboflavin) which are generally accepted as components of co-enzymes involved in car- bohydrate cycles. It is thus of extreme im- portance that carbohydrates are involved in 705 some way in the action of most of the mi- cromelia-inducing chemicals (thallium is ex- ceptional). The teratogenic effects of insulin may be almost completely eliminated by simultaneous injections of nicotinamide (Landauer, *48a). In addition, insulin pro- duces hypoglycemia (Zwilling, 48, °51) and other pronounced carbohydrate disturbances. Both severity and incidence of hypoglycemia correlate well with the degree and incidence of micromelia. In like manner the teratogenic effects of sulfanilamide (Zwilling and De- Bell, ’50), of eserine sulfate (Landauer, *49) and of pilocarpine (Landauer, *53) may be more or less completely mitigated by injec- tions of nicotinamide (which is a component of co-enzyme I). The effects of boric acid (Landauer, 52) are most convincingly re- lated to its capacity to complex with ribo- flavin and render the latter metabolically inactive. Nicotinamide is not effective in alleviating boric acid defects. This last point is of great significance, since it might be con- sidered that nicotinamide merely provides an alternate source of energy which may be used to overcome general depressing effects of the chemicals. That it does not do so with boric acid points to a more specific action of the chemicals on one or another link in the carbohydrate chain. It is of great importance that these diverse chemicals have been shown to have some relation to carbohydrate metab- olism. This provides the first basis for the elimination of some of the vagueness in pre- vious discussions of etiology of anomalies. The fact that different substances differ in the details of their effects points to rather subtle and specific metabolic requirements of the components of the limb rudiments. This material also allows us to evaluate the importance of lowered developmental rate as a prime factor in teratogenesis. At least two substances [adrenal cortical ex- tracts (Landauer, ’47b; Karnofsky, Ridgeway and Patterson, 51) and para-aminobenzoic acid (Zwilling and DeBell, °50)] cause dwarfing of chick embryos after injection at 5 days. The size reduction is perfectly pro- portionate, with no special effects on the limbs; it is of the same order of magnitude as that produced by sulfanilamide and in- sulin and occurs over the same period of time. Moreover, there is no mitigation of the dwarfing when the micromelia produced by sulfanilamide and insulin is eliminated by nicotinamide therapy. These facts indicate that a substance may produce quite general effects (i.e., retardation) as well as specific effects but that the latter need not be causally 706 related to the former. On the basis of results from the mere completely analyzed cases, most embryologists have abandoned the idea that retardation per se may be the cause of anomalies. Finally, the analyses of micromelias in chicks have implications for an interpreta- tion of phenocopies. There are, at present, no indications that the effects of the Creeper mutation are altered by nicotinamide therapy (Landauer, 54). The blood-sugar levels are quite normal in this stock (Zwilling, unpub- lished). Despite the similarity in appearance of the induced and genetic micromelias, ap- parently the metabolic derangements leading to the morphological condition are quite dif- ferent. This is not surprising. The abnormal carbohydrate metabolism probably has its ultimate expression in quantitative or quali- tative alterations of protein synthesis. Similar changes in proteins may be produced by the genetic factors by considerably different pathways. In this case, again, the phenocopies have not copied the initial metabolic action of the genes. A CLASSIFICATION OF TERATA BASED ON EMBRYOLOGY Rather than repeat the usual teratological classification the writer will present a new type of classification of terata based on knowl- edge of inductive and morphological relation- ships. It will be shown how dislocations in normal processes may give rise to various kinds of familiar malformations and, at the same time, by drawing on both for examples, it will be shown that there are no basic dif- ferences, at least on the morphological level, between genetic and experimental anomalies. This classification is not final; there are, doubtless, instances of over-simplification or omission. The intent here is to be provoca- tive rather than definitive. Many of the cases have been placed in a given category even though their analyses are not complete from the present point of view. Future studies may very well indicate the proper position of these examples. It must be borne in mind that many cases of teratological development may involve more than one of the categories which we present. Whenever the information is at hand we have included some of the data in regard to the physiological disturbances which precede the morphological ones. Teratological development may _ result from: 1. Abnormal initial stimulus. a. Initial stimulus absent. TERATOGENESIS b. Deficient initial stimulus. c. Excessive initial stimulus. 2. Abnormal response of reacting tissues. a. Absence of response. b. Partial or incomplete response. c. Excessive response. d. Mechanical interference with re- sponse. 3. Abnormality of both initial stimulus and responding tissue. 4. Abnormal differentiation of component tissues. Abnormal growth of structures. . Degenerative processes. a. Abnormal degeneration. b. Excessive “normal” degeneration. c. Failure of degeneration to occur. 7. Abnormality of functional activity or regulatory mechanisms. oe ABNORMAL INITIAL STIMULUS Initial Stimulus Absent. In most cases of agenesis it is very difficult to establish whether the initial stimulus is lacking or whether the reacting tissues are unable to respond to the stimulus. Probably cases of anidian development in amniotes (in which cells of the blastoderm may divide and even form blood islands, but in which embryonic tissue never develops) represent instances in which the primary organizing stimuli are absent; but this has never been definitely demonstrated. There are, however, well es- tablished cases involving secondary inductors which show that the stimulus for develop- ment may be lacking. Boyden (’27) and Gruenwald (°37, 42) have demonstrated by experimental procedures that the elaboration of a metanephric kidney (in chick embryos) depends upon the prior formation of a ureteric bud. Normally this bud grows up from the cloaca to the metanephrogenous blastema and stimulates the latter to elab- orate tubules, etc. The blastema does not differentiate if the ureteric bud fails to reach it after a surgical block (see Section VII, Chapter 6). Substantiating evidence for this relationship has been found in the case of the wingless mutation of chicks (Waters and By- waters, 43; Zwilling, 49). In the homozy- gous recessives the metanephrogenous tissue does not develop beyond the blastema stage; this is associated with absence of the ureteric bud. Deficient Initial Stimulus. Probably the best example of this category is microcephaly (reduced head). Excessive Initial Stimuli, An increase in the TERATOGENESIS number of induction centers or fields is one category of excess of stimulation. No excess of material may be involved, merely the iso- lation (physiologically or otherwise) of inde- pendent organization centers. The result may be partial or complete duplication of a struc- ture or organism. Such duplication may occur spontaneously or as the result of many experimental treatments (Stockard, ’21; Hin- richs and Genther, ’31; Torrey and Brene- man, °41; Pasteels, 47; Lutz, ’49, are but a few of the references). Another type of excessive development re- sulting from excessive induction would be the increase in size of structures. Few of these instances have been sufficiently analyzed to ascertain that the increase is definitely due to the evocating rather than the responding tissues. Ranzi, Tamini and Offer (46) have described what they call a hyperinduction of this sort in Rana esculenta and the axolotl of Amblystoma tigrinum. When blastulae of these species were kept in a solution of so- dium thiocyanate for periods ranging to 48 hours there was considerable enlargement of the notochord. Associated with this were en- largement of the rhombencephalon, duplica- tion of the epiphysis and other hyperdevelop- ments. (They point to the similarity of these malformations to the “Arnold-Chiari” syn- drome of human teratology. ) ABNORMAL RESPONSIVENESS OF REACTING TISSUE Absence of Response. See Section VI, Chap- ter 1. Partial or Incomplete Response. In this cate- gory we place those cases in which the initial stimulus for development is present and ap- parently normal. The responding tissues, however, cannot react completely, with the result that there is a deficiency, either of size or of form, in the ensuing structure. In some instances of this type of aberrant de- velopment the incomplete response may be due to failure of the subsidiary inductors, in others the abnormal response may be more directly linked to the initial evocator. An ex- ample for this group is the recessive “wing- less” mutation in fowl (Waters and Bywaters, 43). Initial wing induction is apparently normal, since wing buds are present in all cases. However, in homozygotes, the apical ectodermal thickening usually present along the ridge of limb buds is missing from the wing buds (Zwilling, ’49). Saunders (748) has demonstrated that the apical ectoderm plays an important role in the elaboration 707 of the wing. Surgical removal of this struc- ture at 72 hours resulted in complete absence of distal parts of the wing. Excessive Response. While there is ample evidence from transplantation experiments that some tissues of embryos have far greater potencies than are ever expressed under nor- mal conditions, it is relatively rare to find expression of these excess potencies in intact embryos which are subjected to abnormal physical or chemical conditions. True, one finds frequent duplications of entire regions of the body. However, these must be ascribed, for the most part, to excessive or accessory inductions and not to excessive response of the reacting tissues. Accessory invaginations have been observed following many types of experiments (Pasteels, "41; Atlas, 735). There are a number of anomalies which probably represent the result of excessive re- sponse to secondary inductors. Hyperpha- langy (enlarged digits due to excess number of phalanges), polydactyly (excess number of digits), macrodactyly, hypermasticism (su- pernumerary mammae), cases of supernu- merary ribs and other similar anomalies probably fall within this grouping. Patten’s (52) case of neural tube overgrowth in hu- man embryos may be a case in point. Mechanical Interference with Response. Un- der certain circumstances it is likely that both the initial inductive stimulus and the reactivity of the tissue may be normal, yet, owing to a mechanical alteration of the spatial relationships, a structure may de- velop abnormally. One must bear in mind that the mechanical alterations themselves represent the expression of earlier derange- ments, so that if the entire causal sequence is to be determined one must reconstruct a complete chain of events leading to a par- ticular anomalous development. There are numerous instances in the literature on trans- plantation and explantation in which such effects of alteration of the mechanical con- ditions have resulted in distortions of de- velopment. When an inductor-reactor system is separated the structure involved usually fails to develop. The best example of this is the exogastrulated amphibian embryo dis- cussed in previous chapters. Here the mor- phogenetic movements fail to establish the proper relation between the roof of the arch- enteron and presumptive neural tissue. The consequence is a complete absence of any complex differentiation of the ectoderm. In addition, there may be considerable distor- tion of an organ as the result of disturbed mechanical relationships. In the case of the 708 “shaker-short” mutation of mice, Bonnevie (34) has traced such an inherited situation back to its apparent origin. Shaker-short mice are characterized by a number of abnormali- ties, the most striking of which are their erratic waltzing movements. These aberra- tions of movement, and a deafness, have been linked to abnormal development of the ears. An embryological study revealed that the inner ears in these mice are normal until the ninth day of gestation. The various append- ages which should be formed in subsequent stages are not elaborated despite the fact that differentiation of the epithelium of the vesi- cle is normal and various regional character- istics are present in approximately the usual location in the distorted ears. The onset of these deviations is correlated with a failure of the myelencephalon to expand as it does normally. Largely as a result of lack of pres- sure on the ear vesicle, according to Bon- nevie, the aural appendages are inhibited. Thus, despite the capacity for normal cellu- lar differentiation the abnormal mechanical conditions interfere with normal form devel- opment to such an extent that severe anom- alies and aberrant functioning result. ABNORMALITY OF BOTH INITIAL STIMULUS AND RESPONDING TISSUE In theory it is possible that both the in- ductive stimulus and the reacting tissue may be defective in some respect. This may be the situation in many of the experiments in which the entire embryo is subjected to the influence of an abnormal chemical or phys- ical milieu for protracted intervals. It is also probably true in many cases of gene-medi- ated anomalies. Too frequently investigators are prone to implicate an inductor without considering that the reacting tissue may also be involved. Recently, just this situation has been illustrated by the experiments of Moore (46, ’47, 48). In hybrids of R. pipiens 9 x R. sylvatica § development proceeds nor- mally to the initial stages of gastrulation. The majority of the embryos cease their de- velopment at this point and continue as “blocked” gastrulae for several days before they cytolize. Moore has tested both the com- petence of the ectoderm and the inductive capacity of the dorsal lips of such gastrulae. This was accomplished by heteroplastic and xenoplastic transplants in which the ecto- derm was placed in specified regions of the host and the dorsal lip in the blastocoele of another host. Similar transplants of normal R. pipiens gastrular tissue were used as con- TERATOGENESIS trol. By comparing both the number and the size of the induced structures in the two situations Moore has concluded that both the competence of the ectoderm and the inductive capacity of the dorsal lip are defective in hybrids. It is of interest that Barth (46) and Barth and Jaeger (47) have shown that respiration, lactic acid production and apy- rase activity are interfered with in these hybrid gastrulae, while Brachet ('47b) has demonstrated that there is an inhibition of ribonucleic acid synthesis in similar blocked hybrids. Sze (53) has shown that respiration is depressed in all regions of the ectoderm of the hybrids. ABNORMAL DIFFERENTIATION OF COMPONENT TISSUES There are a number of anomalous situa- tions in which the tissues either do not differ- entiate normally or differentiate in another direction than expected. These might be con- sidered cases of abnormal response of react- ing tissues, although the exact circumstances involved in these cases are, for the most part, unknown. One of the best known examples of a tissue differentiating atypically is found in Leh- mann’s (738) lithium-treated Triton embryos, in which the presumptive chorda material becomes continuous with and differentiates as mesoderm. [Cohen (738) believes that it becomes entoderm in his experiments.] An- other instance is that of an inherited anomaly in mice (Hovelacque and Noel, ’23) in which the distal end of the precartilaginous con- densation of the tibia differentiates into fibrous ligaments instead of cartilage. In the case of Griineberg’s (738) grey lethal mutation in rats the cartilage cells, especially of the ribs and trachea, are larger than normal and form thick, enlarged cap- sules; in addition the perichondrium is ab- normally thickened and active. The result is a hyperplasia of the tracheal and costal car- tilages which is instrumental in causing an emphysema of the lungs. This eventually leads to the death of the animal. Hyperac- tivity of the cartilage continues even when it is transplanted to a normal host but does not do so when grown in tissue culture (Fell and Griineberg, °39). Normal cartilage does not become abnormal in a lethal host. These experiments led the authors to the conclusion that the abnormal condition is intrinsic to the cells which bear the lethal genes. While the phenomenon is rare, embryonic tissue may be involved in tumor formation. TERATOGENESIS Witschi (22, °30, °34:) described several types of tumors which developed in embryos from overripe eggs of Rana temporaria. Briggs (41), and Briggs and Berrill (41), from a study of similar embryos, reported the ap- pearance of several epidermal growths, one of which was a papilloma that was benign in normal hosts. It grew progressively only in hosts which happened to be retarded. ABNORMAL GROWTH OF STRUCTURES AND INDIVIDUALS Into this category we place cases in which there is an anomalous size increase of the organism or its parts but in which cell size and differentiation are not affected. There are many cases in which normal growth pat- terns are distorted, either owing to deficien- cies in growth-promoting substances or be- cause of an excess of such factors. This topic is covered in Sections X and XI, and at this point we shall merely mention that such growth-disturbing conditions may affect one part of the body quite differently than others and thus lead to marked disproportions (see also Huxley, 732). DEGENERATIVE PROCESSES Many types of anomalous development are characterized, in one phase or another, by degeneration of tissues. It must be borne in mind here again that the destructive changes are merely visible expressions of aberration in some physiological process. The investiga- tions of Mitchell (43) and of Bodenstein and Kondritzer (48) demonstrate that at least one of the disturbances preceding cellular degeneration, after both x-ray exposure and mustard gas treatment, is the inability of the cells to synthesize desoxyribonucleic acid. Ribonucleic acid formation is not interfered with. Degeneration may be extensive and result in the destruction of the entire indi- vidual. Sometimes it may be sharply localized and involve only a given region or structure. In some instances it has been shown that an initially restricted focus of degeneration may spread and involve cells which have the capacity to survive when removed from the degenerative influence. Baltzer (30) and Hadorn (’32) have found this to be the case in their merogonic hybrids of Triton. Abnormal Degenerations. The heading “ab- normal degenerations” seems redundant. It is used here to emphasize the fact that degen- erative processes occur quite frequently as a step in normal morphogenesis (see below). 709 Terata may arise either as an extension or exaggeration of a normal degeneration or as the result of degeneration where none is nor- mally involved. There are many instances in which degenerative changes follow a gross distortion of morphogenesis. In the examples which we shall give for this section we should like to disregard these and include cases in which the degeneration is one of the first evidences of abnormality. Many instances of genetic incompatibility are characterized by degenerative alterations. When the genic disturbance is considerable, as in the case of the merogonic hybrids men- tioned above, large areas may become pyk- notic at relatively early stages and eventu- ally lead to the death of the individual. In some cases the degeneration is more sharply restricted. Greene and Saxton (739) have de- scribed a recessive mutation in rabbits which causes brachydactyly (and other abnormali- ties of the extremities) in homozygous reces- sives. The anomalous condition was preceded by dilatation and swelling of distal blood ves- sels of affected parts on the sixteenth to seventeenth days of development. Inman C41), in a more detailed examination of earlier stages, reported that the first evi- dences of this gene’s expression were de- generative changes in the endothelial cells of these vessels as early as the thirteenth day. On the eighteenth day red patches were grossly visible. Hemorrhage and necrosis be- gan on this day and became progressively more marked. There was usually a sharp line of demarcation between normal and necrotic regions which, finally, formed an annular constriction. The necrotic tissue sloughed off and the stubs of the limbs were “healed” by the twenty-fifth day of gestation. Cells may degenerate very rapidly, with- out any apparent preliminary events, follow- ing certain treatments. Mustard gas caused this sort of degeneration in certain non- dividing cells of the eye in Amblystoma (Bo- denstein, 48). In other cells, in the same work, degeneration occurred after a delay which was characterized by inhibition of mitosis, cell enlargement, increased baso- philia and inhibited desoxyribonucleic acid synthesis. Excessive Normal Degeneration. As men- tioned above, cellular destruction may be a normal part of a morphogenetic process. Scat- tered cells whose large chromophilic droplets are evidence of degeneration are encoun- tered in many places in embryos (Ernst, ’26). More concentrated foci of such cells are found to occur quite regularly in certain 710 instances: the cloacal region of chick embryos (Boyden, °22), limbs in mice (Chang, *40) and other embryos, the brain of human and chick embryos (Gruenwald, °45), etc. The morphogenetic importance of these foci has occasioned some speculation but is, in most instances, still obscure (Ernst, ’26; Peter, 36; Gliicksmann, *51). It is likely that these foci are, under normal conditions, kept localized by inhibitory factors. It is conceivable that the inhibition may be lost under certain cir- cumstances and that the degenerative foci may involve more extensive areas and cause more permanent destruction than one en- counters normally. An example which seem- ingly involves this type of excessive destruc- tion is found in the dominant rumpless mu- tation of chicks discussed above (p. 703). In normal embryos there is a focus of degener- ating cells posterior to the tail bud which represents the remnant of the primitive streak. As a result of the formation of the tail fold this group of cells comes to lie in the region of the cloaca and concomitantly there is an increase in the number of de- generate cells. The latter become, during the third and fourth days, involved in the events leading to the formation of the cloaca and anus (Boyden, ’22). In the dominant rump- less embryos it appeared that the degenera- tive process did not remain restricted to the anal plate but spread to the undifferentiated tail bud. Failure of Degeneration to Occur. A final possible involvement of degenerative proc- esses in the production of anomalies is their failure to occur as part of a normal develop- mental sequence. All students of anatomy are familiar with such instances. It will suf- fice to mention a few of the well known ex- amples of this sort. Persistence of the thymus gland beyond the stage when fatty degener- ation normally sets in is a familiar case. There are instances in which various of the primitive aortic arches fail to degenerate. Probably the persistence of the right fourth aortic loop with the resulting two aortae is one of the most striking of these. In like man- ner the failure of the ductus arteriosus to undergo its fibrotic degeneration (and the resultant “blue baby”) is classic. For details of this type of anomaly see Patten (46) and Arey (54). ABNORMALITY OF FUNCTIONAL AC- TIVITY OR REGULATORY MECHANISMS Under this heading are included anomalies which result directly from a gross disturb- TERATOGENESIS ance of funciion and from hormonal imbal- ance. It has been pointed out previously, and should be emphasized again, that all of the malformations described in this chapter must be the visible expressions of disturbances of physiological processes. Those discussed here are cases in which these disturbances are more obviously related to the teratological development. Bonnevie ('36, °43) has investigated two mutations in mice in which excess of fluid production by the choroid tissue of the brain produces different effects. In the first (Bagg and Little’s x-ray—induced mutation) the ex- cess fluid escaped from the brain and, atter flowing under the epidermis along the con- cavities of the body, formed blebs in various places (over the eyes, at the distal end of the limbs, etc.). Eventually the blebs became filled with clots of extravasated blood. The abnormal local pressures caused by the clots result in various anomalies [club-feet, syn- dactyly, hypodactyly and congenital ampu- tation (Bagg, ’29)]. In the other mutation the choroid plexi start producing the excess fluid during the twelfth day of gestation. The result is a fairiy typical hydrocephalus, since the fluid does not escape (see Mechanical In- terference with Response, above, for effects of deficient pressure in the brain and its re- lation to defects of the ears). The well known effects of hormonal im- balances have their parallel in embryonic life. Fugo (40) has demonstrated that when the pituitary gland was removed (from chick embryos of 33 to 38 hours), growth was re- tarded. The retardation became very marked after the sixteenth day of incubation. In ad- dition the feathers, thyroid glands, testicu- lar intertubular tissue and yolk-sac with- drawal were abnormal in the later stages of development. The general similarity of the effects of this experimentally produced de- ficiency to hereditary cases of pituitary mal- function is marked despite the fact that in the best known cases of the latter (pituitary dwarfism of mice) it is evident that the aber- rant development does not begin until after birth (Francis, 44). GENERALIZATIONS In this section are indicated certain gen- eralizations which may be drawn from a study of experimental teratogenesis. It is quite evident that this information may relate to some of the occasional or sporadic terata. However, this section is chiefly in- tended to clarify some of the problems and TERATOGENESIS indicate some of the sources of confusion in this type of experimentation. CONDITION OF GAMETES AND ZYGOTES One of the most consistent observations made by investigators in the field of experi- mental teratology is the extreme variability in the reactions of fertilized eggs or embryos to the treatments to which they are subjected. Zygotes from the same set of parents do not all respond equally to a given set of conditions, and the incidence of response differs in offspring from different parents. Variations in reactivity of zygotes and em- bryos may be due to alterations in their physiological condition or to their genetic constitution. Physiological Condition. It is quite evident that normal development can be expected only from zygotes that have been produced and raised under normal conditions. Some effects may be produced by factors acting before the onset of embryonic development. The experiments of Witschi (’22, °30, ’34) on eggs which were retained in the female’s uterus for 3 to 5 days prior to fertilization are familiar examples. Such over-ripe eggs are frequently polyspermic. Among eggs fer- tilized by one sperm Witschi found axial duplications, supernumerary appendages, pigment changes and a tendency towards the production of neoplasms. According to Wit- schi such abnormalities result from a lack of “coordination” in the egg and embryo, probably due to alterations in the cortical layer. In chick eggs it has been known for a long time that proper storage conditions are essential for good development (Landauer, 51). Proper temperature and humidity must be maintained. Extended storage, even under optimum conditions, may also result in im- paired development. Poor pre-incubation conditions result in poor hatchability, early mortality and anomalies. In experiments in which the effects of some particular treat- ment of eggs are to be evaluated these con- ditions must be controlled carefully. The reactions of eggs which have been im- properly stored cannot be compared with those which have been handled carefully. It is likely that at least some of the contradic- tions in the literature are due to these factors. Even under proper conditions of storage one may encounter differences in susceptibility of eggs from the same hens at different times of the year. Landauer and Baumann (’43) found that the incidence of rumplessness ul resulting from mechanical jarring of eggs was greater during the summer months than in the spring. In another study Landauer (43) found that there was a greater in- cidence of micromelia during the fall and early winter than in the late winter and early spring. The reverse was found for eye defects (micro- and anophthalmia). These few illustrations are presented to indicate that one must know the condition of material with which one performs experiments. There is little doubt that similar variations in gametes and zygotes exist in other forms. The researches of the Hertwigs and others (see Frets, ’31, for bibliography) demonstrate that the fertilization of eggs with sperm previously subjected to treatment with var- ious chemicals (methylene blue, chloral hy- drate, strychnine, etc.) results in the forma- tion of many terata. Similar results may be obtained after irradiation [it is important to note that after exposure of sperm to increased doses the incidence of abnormalities decreases because of the ability of the otherwise in- activated sperm to initiate parthenogenetic development (Hertwig, 28; Rugh and Exner, ’40)]. More clearly related to the practical consideration of physiological conditions are observations on the effects of fertilization of eggs by “stale” sperm. Nalbandov and Card (43) have obtained data of this sort for chickens. They removed the males from pens of tested hens and collected eggs for as long as 35 days following this separation. They found that eggs fertilized with older sperm had a lower hatchability and greater early embryonic mortality, with embryonic death occurring earliest in those eggs fertilized by the oldest sperm. There was no increase in incidence of terata. Aberrant development was found in a similar experiment (Dhar- marajan, *50). Some observations (unpub- lished) made by the writer at Storrs indicate that embryos from eggs fertilized by stale sperm are much more susceptible to various treatments (both in regard to mortality and incidence of abnormalities) than are more vigorous embryos. Genetic Constitution. The importance of genetic uniformity in experimental material is well established at the present time. There is still, however, a tendency among some investigators to neglect this factor in com- paring their results with teratological agents with those of others. Stockard emphasized this point by indicating the difference be- tween trout and Fundulus embryos in regard to the incidence of twinning. Trout embryos not only yielded a much higher percentage of iz these abnormalities after similar treatment, but also showed a much greater incidence of spontaneous duplicate embryos. Holtfreter (45) has demonstrated that a difference in neuralization response between the ectoderm of Amblystoma punctatum and Triturus toro- sus is the result of an increased susceptibility of the former to hypertonic salt solutions. This susceptibility is in turn related to differences Ty 2 cS (2) % Incidence of Abnormal Ww oO LOD ZO 30 40 50 GO 20 S80 in the resistance of the cell membrane of the outer ectodermal cells. The above examples refer to differences between genera. Similar differences may be found between different strains within a species. Warkany et al. (42) demonstrated a strain difference between Sprague-Dawley and Wistar rats in regard to the production of skeletal abnormalities as a result of dietary deficiencies (later shown to be riboflavin). The former strain had a higher incidence of abnormalities. However, this was shown to be due to its greater resistance to the deficiency, since the Wistar rats were so affected by this treatment that they failed to produce any young. An even more strik- ing demonstration of genetically conditioned response to experimental conditions has been provided by Landauer (48b). White Leg- horn embryos yield a considerably higher percentage of rumpless embryos after insulin TERATOGENESIS is injected into the yolk sacs of unincubated eggs than do embryos of other breeds of chickens. Moreover, it is possible (Landauer and Bliss, ’46) to select a “high” and a “low” susceptibility line from amongst the White Leghorns themselves (progeny from hens giving high and low incidences of rumpless- ness were saved and inbred). After continued selection the susceptibility of the high line e Rumplessness x Micromelia 120 130 1407 150) 160! Hours 90 100 110 lime of Injection ot Tamsilain Fig. 247. Effect of time of injection of 2 units of insulin on the incidence of rumplessness and micromelia. All injections into yolk-sac of embryos of ages indicated. The continued small incidence of rumplessness beyond 72 hours is the same as the sporadic occurrence in controls. (Data from Landauer and Bliss, ’46; Landauer, *47a.) increased to such an extent that, with 2 units after 4 hours of incubation, the inci- dence of rumplessness was 51.9 per cent while in the low line it was 18.2 per cent (Landauer, unpublished). This increased susceptibility to injection of insulin at early stages 1s not retained at later stages. When insulin is administered at 5 days there is a high incidence of micromelia. This incidence is no greater in the high line (i.e., high in regard to production of rumplessness) than in the low line. STAGE OF TREATMENT It has been recognized for some time that the stage during which an embryo is sub- jected to a particular type of treatment is of utmost importance in determining the type of abnormality which the treatment will cause. In general the younger an em- TERATOGENESIS bryo is when subjected to some foreign environment the more profound will be the alterations produced. This is frequently ex- pressed by a higher incidence of very early mortality or a cessation of development at an early stage. Among the survivors, how- ever, one may find a surprising regularity in the appearance of a particular defect. This has been taken to indicate that the cells EPIDERMIS=x NOTOCHORD«x 6 UT——: —__——_ LENS ——_—_—_—_—_———————": —_—— PRONE PH RO S x LIVER AND PANCREAS EAR AND LATERAL LINEx GERM CELLS MYOTOMES HEART: MID- AND HINDxBRAIN TAIL EMBRYONIC ORGANS OPERCULUM AND JAWS————__ x 713 is being most actively elaborated. This may, in general, be true. However, Solberg (738) has presented material which demonstrates that different structures vary considerably in regard to the relationship between their sensitivity to x-ray treatment and the time of actual elaboration of the organ. As in- dicated in Fig. 248, some organs in Fundulus react to x-rays both prior and subsequent to AIR BLADDER————__________________ RS A ee ee SS TS SPUN AL COR (a Oa rer re an FR ee ES PORE BRAWN ___. EYES AND NARES————$—$—$—$—$_—$_ << ———_____ to) 12 24 36 48 60 72 84 96 108 AGE IN HOURS AT RADIATION Fig. 248. The effects of x-rays on the differentiation of the embryonic organs of Fundulus heteroclitus. The zx indicates the stage in development when the anlage of an organ appears. The line to the right of each organ’s name indicates over which period that organ may be affected by the irradiation (see text). (Courtesy of Dr. A. N. Solberg and the editors of the Journal of Experimental Zoology.) which are the precursors of the affected structures are particularly susceptible to the treatments at that stage. Moreover, when the same treatment is administered to em- bryos at different stages not only will there be a difference in incidence of a particular anomaly but new anomalies may be pro- duced; a new set of structures has become susceptible. The data from Landauer and Bliss (46) and Landauer (’47a) in regard to injection of insulin into chick embryos illustrate this point (Fig. 247): when these injections are made from 0 to 64 hours of incubation the most frequently encountered anomaly is rumplessness. From 72 to 168 hours, however, the insulin produces a high incidence of micromelia. There is a widespread belief that the sensitive or critical period in the develop- ment of a structure occurs chiefly during its early formative stages, when the cells are as yet undifferentiated or when the structure their formation, while others are affected only before the structure appears. By draw- ing a line perpendicular to the time axis one can determine which structure will be affected by the irradiation at any given stage. This does not indicate the degree to which the structures are affected. In order to obtain good anterior duplica- tions by means of physical or chemical in- terferences one must treat very early em- bryonic stages. Only poorly organized or posterior duplications have resulted from centrifugation of amphibian blastulae and gastrulae (Torrey and Breneman, ’41; Pas- teels, °47). In Stockard’s (’21) experiments with Fundulus anterior duplications were obtained most successfully when early cleav- age stages were treated. This was confirmed by Hinrichs and Genther (31), who demon- strated that U-V irradiation 30 minutes prior to first cleavage was most favorable for production of anterior duplications. The pre- 714 cleavage stage is also most sensitive to centri- fugation (Forsthoefel, ’51). There may, thus, be a considerable delay before visible effects of some treatments are evident. Ancel (45a) concluded that the micro- melic effect of sulfanilamide resulted from its interference with pre-cartilaginous mes- enchyme. He applied the drug at 48 hours. The writer’s earlier discussion of the differ- ences in susceptible periods to various micromelia-inducing chemicals (p. 705) in- dicates that stages of active chondrification are also affected. In addition, the application of sulfanilamide at 120 hours (Zwilling and DeBell, *50) will also produce micromelia. Moreover, if sulfanilamide is injected at 48 hours and nicotinamide at 120 hours (exposing the limb to the drug during the pre-cartilage stage) the micromelic effects are entirely abolished. All of this indicates that sulfanilamide acts over a considerable period of time. This probably is true of other teratogenic agents. It is likely that some of them may have relatively little effect until development of a particular structure has proceeded sufficiently for it to reach its sensitive stage. Experiments must be care- fully planned for accurate determination of sensitive periods and not based solely on the results of treating one stage. In like manner, in view of the fact that indistinguishable end results may be mediated through more than one developmental pathway (i.e., limb- lessness or taillessness, either by degeneration of normally formed structure or failure of the structure to be elaborated), only the most limited conclusions about sensitive pe- riods can be obtained from the appearance alone of a fully developed organism. NATURE OF TERATOGENIC AGENCY Instances where the same terata have been produced by a variety of treatments have led to the emphasis on stage of treatment and neglect of specific action of the agent (vide supra). However, there is evidence which indicates that the specific nature of the substance used is of considerable impor- tance. Most impressive, in recent years, has been the research of Ancel (45a, °45b, 50) and his collaborators. These investigators have applied some 90 or more substances to chick embryos of the same age in the same manner. They have found considerable dif- ferences in regard to the terata produced. For instance, eserine sulfate and a number of sulfa compounds produce micromelia (frequently accompanied by parrot beak and TERATOGENESIS occasionally by syndactyly) but no rump- lessness or coelosomy (a condition in which the viscera protrude from and are outside the body cavity). Another group of sub- stances (ricin, trypaflavine, etc.) produced coelosomy but no micromelia or rumpless- ness. Still another group (sodium cacodylate, sodium methyl arsonate, saponin) resulted in rumplessness, which was frequently asso- ciated with coelosomy and ectro- or hemi- melia (a condition in which all or part of a limb is absent), but no micromelia. More- over, these substances retain their specific action even when applied simultaneously (Ancel, °47). Trypaflavine alone resulted in the non-development of the amnion but no rumplessness, while methyl arsonate pro- duced rumplessness and no abnormalities of the amnion. Together they produced, in addi- tion to embryos with either of the two anomalies, some embryos which were both rumpless and anamniotic. Ancel and his group have not overlooked or minimized the im- portance of the stage when these substances are applied. They have demonstrated that the same substance may have different effects on different stages. All specific effects must, therefore, be defined in terms of both the stage treated (or affected) and the nature of the agent. The investigations of Hamburger and Habel (747) offer additional evidence for specific action. These authors have shown that two different viruses (influenza-A and mumps), both of which penetrate embryonic tissues and multiply therein, may have dif- ferent teratological effects. Both of the viruses, applied to 48-hour embryos, were toxic but only the influenza-A was terato- genic. It caused a specific syndrome: micro- cephaly, micrencephaly, twisting of the axis and impairment of growth of the amnion. It must be concluded that these differences in effect are due to specific differences in the action of the two viruses. Along with the specific action of many substances one frequently finds “general toxic effects” or “non-specific mortality.” As implied, these effects are noted in those in- dividuals which succumb to the treatment at a relatively early stage and usually show a uniformity regardless of the nature of the treatment. The more specific effects are found in the survivors. In early chick embryos, for example, one encounters a typical syndrome preceding death due to a variety of causes: the extraembryonic blood vessels break down and the entire embryo is retarded, with the head especially showing the retardation. TERATOGENESIS While the substances which produce such general effects may alter embryo metab- olism in a specific manner, the visible effects show no specificities in these moribund em- bryos. In determining whether a given effect is specific one must very carefully consider the general toxic effects and use adequate controls. The non-specific symptoms of mori- bundity must be recognized so that they may not be confused with more specific effects. EFFECTS OF DOSE In general the greater the dose of a given treatment the greater will be its effect. This may be due to an increase in intensity of action at the time of application or to an increased duration of its action. Both of these points must be borne in mind in any dis- cussion of critical periods. It is possible to subject an embryo to a treatment prior to the critical period for a structure, yet if the effects of the treatment persist long enough, it may include the critical stage for this structure. A more intense treatment may produce anomalies in a stage which does not respond to less intense treatment. Thus Job, Leibold and Fitzmaurice (’35) produced hydrocephalus and defects of the eyes and jaws in rats which had been exposed to x-rays during the ninth, tenth and eleventh days of gestation. Exposure of older embryos to the same dose did not produce any abnor- malities. Warkany and _ Schraffenberger (47), on the other hand, obtained some anomalies after irradiating with higher doses as late as the sixteenth day of gestation. There was a very high incidence as late as the fourteenth day after the intense treat- ment. While it is true that many teratogenic agencies are most effective in doses which are toxic to a fairly high percentage of the treated embryos, one must not confuse the toxic effects with the teratogenic ones. There is ample evidence to show that the two are frequently separable. Ancel (’45c) has shown that the toxicity of some substances may in- crease without concomitant increase in tera- togenicity. On the other hand Landauer and Bliss (46) have shown that an increase in dose of insulin causes a relatively greater increase in the teratogenic effect than in toxicity. If one injects both sulfanilamide and nicotinamide into an egg, there is a much greater mortality than with the former alone (Zwilling and DeBell, 50). However, the micromelic effects of the sulfanilamide are eliminated. gA\S) RECOVERY One of the factors involved in differences of response to the same treatment in a rela- tively homogeneous group of embryos is the ability of some of them to recover from the effects of the treatment sufficiently early for the affected structures to resume some semblance of normality. In some types of experiments recovery phenomena may be demonstrated by removing the embryos from the altered environment and placing them in a normal one. In such cases the degree of recovery is usually related to the duration of the treatment. Under certain conditions some of the effects of treatment may be alleviated by slowing down the metabolic rate of the exposed individuals shortly after the treatment. This was done by Cook (739), who kept Ascaris eggs at low temperatures following a lethal x-ray treatment. On re- turn to higher temperatures the lethal effects of the irradiation were not noted, although other effects (delayed cleavage) were still present. Child (41) cites numerous instances in which recovery seems to be related to an acclimatization (differential acclimation or tolerance) to a toxic environment. It has been noted (Zwilling, 48) that chick em- bryos may have quite different patterns of hypoglycemia after treatment with the same dose of insulin. Some remain hypoglycemic for as long as 8 days, while others recover normal blood sugar levels at various inter- vals prior to this. The recovery is associated with a decrease in the micromelic effect of the insulin; embryos which recover early are less micromelic. Such differences in tol- erance and ability to recover are probably related to genetic differences between in- dividuals (see above). RESUME In this chapter an endeavor has been made to show that aberrant development results from a distortion of normal developmental processes; that both genetic and environ- mental factors may be responsible for pro- ducing anomalies; and that the same de- velopmental principles are involved in both. Terata may follow treatment of embryos with highly toxic foreign substances but also may result from dietary imbalance, blood factor incompatability, irradiations and ac- tion of disease-causing agents. The reaction of the embryo may depend not only on the stage at which it is subjected to the abnormal influence, but on the nature of the tera- 716 togenic agency, the length of time it is ex- posed, the physiological condition of the embryo and its resistance or susceptibility as determined by its genetic constitution. Similar anomalies may result from more than one developmental aberration. Teratology is still at the stage where it is most concerned with the visible effects which are the expression of prior physio- logical disturbances. The next big step in studies of the etiology of anomalous develop- ment must concern itself with the details of these physiological processes. REFERENCES Ancel, P. 1945a L’achondroplasie. Sa réalisation expérimentale — sa pathogénie. Ann. d’Endo- crinol., 6:1—24. 1945b Sur l’action tératogéne élective de certaines substances chimiques. Compt. Rend. Soc. Biol., 139:983-984. 1945c Les variations individuelles dans les expériences de tératogenes. Rev. Sci., §3:99- 106. 1947 Sur la mise en évidence de différ- ences individuelles dans la constitution des em- bryons par l’action associés de deux substances chimiques tératogénes. Compt. Rend. Soc. Biol., 141:208-209. 1950 Ia chimiotératogenése chez _ les Vertébrés. G. Doin et Cie, Paris. Arey, L. B. 1954 Developmental Anatomy, 6th ed. W. B. Saunders Co., Philadelphia. Asmundson, V. S. 1939 Some factors affecting hatchability of eggs. Poul. Sci., 78:399. 1942 An inherited micromelia in the do- mestic fowl. J. Hered., 33:328-330. Atlas, M. 1935 The effects of temperature on the development of Rana pipiens. Physiol. Zool., 8: 290-310. Bagg, H. J. 1929 MHereditary abnormalities of the limbs, their origin and transmission. II. A morphological study with special reference to the etiology of club-feet, syndactylism, hypodac- tylism, and congenital amputation in descendants of X-rayed mice. Am. J. Anat., 43:167-219. Ballantyne, J. W. 1904 Manual of Antenatal Pathology and Hygiene. William Green & Sons, Edinburgh. Baltzer, F. 1930 Uber die Entwicklung des Tri- tonmerogons Triton taeniatus Q X cristatus 7. Rey. suisse Zool., 37:325—332. Barth, L. G. 1946 Studies on the metabolism of development. J. Exp. Zool., 103:463-486. , and Jaeger, L. 1947 Phosphorylation in the frog’s egg. Physiol. Zool., 20:133-146. Bodenstein, D. 1948 The effect of nitrogen mus- tard on embryonic amphibian development. IT. Effects on eye development. J. Exp. Zool., 108: 93-126. , and Kondritzer, A. A. 1948 The effect TERATOGENESIS of nitrogen mustard on nucleic acids during em- bryonic amphibian development. J. Exp. Zool., 107:109-122. Bonnevie, K. 1934 Embryological analysis of gene manifestation in Little and Bagg’s abnormal mouse tribe. J. Exp. Zool., 67:443-520. 1936 Abortive differentiation of the ear vesicles following a hereditary brain-anomaly in the “Short-tailed Waltzing Mice.” Genetica, 18:105-125. 1943 Hereditary hydrocephalus in the house mouse. I. Manifestations of the hy-muta- tion after birth and in embryos 12 days old or more. Skr. norske Vidensk Akad. Oslo, Mat.-Nat. Kl. No. 4, pp. 3-32. Boyden, E. A. 1922 The development of the clo- aca in birds with special reference to the origin of the bursa of Fabricius, the formation of the urodeal sinus, and the regular occurrence of a cloacal fenestra. Am. J. Anat., 30:163-202. 1927 Experimental obstruction of the mesonephric ducts. Proc. Soc. Exp. Biol. & Med., 24:572-576. Brachet, J. 1947a Embryologie Chimique. Mas- son et Cie, Paris. 1947b Biochemical and physiological in- terrelations between nucleus and cytoplasm dur- ing early development. Growth, 77:309-324-. Briggs, R. W. 1941 The development of abnor- mal growths in Rana pipiens embryos following delayed fertilization. Anat. Rec., 87:121-135. , and Berrill, N. J. 1941 Transplantation experiments with an ectodermal growth of frog embryos. Growth, 5:273-284. Byerly, T. C., Titus, H. W., Ellis, N. R., and Lan- dauer, W. 1935 hdl —_— Jai anh ale ~~ Pepe ved dag agen As mn, =. * 7 mii — 4 iy P a : aa é a ys. i r wie: a. ; sons Sel hell , ee wae rr o a 1 : ar, Ye ee > 0, ae an A 46 5 ay mak a ' ba 7 ‘ ' ear ys i@ Ore : Ges re im Pt initl il ‘ re et , y A) : a ( ~“o27 .< vee enetinhte rks i) : ¥ ' - ’ ; : sud a i+ - TE ni wz ‘i aie : > : »< 7 j ' a 102 a ce eR rt A : ‘ ion a . *) rT ie ( Fi ¥ 7 9 nN Oh A, is : ve an ; i . v rr ea ‘ rea +h ee , : i ¢ rr ty : ’ i We \ ’ 4 2 ’ ; i ; n ’ ‘ —_ : : 7 Fa INDEX ABDUCENS nucleus, 389 Accessory genital structures. See Genital structures, accessory. Accessory sex structures, 595, 596, 597, 598. See also Genital struc- tures, accessory. Acetylcholinesterase, 376, 383 Acoustic ganglion, 383 Activation of egg, 202-203 Adenylpyrophosphatase, 376 Adrenal, fetal cortex of, 586, 587 morphogenesis of, 586 Adrenal cortex, ACTH effects on, 585 ascorbic acid accumulation in, 588, 589 effects on anterior pituitary, 585 lipid development in, 588 onset of activity, 577, 587, 588, 599 vascular pattern of, 587 Adrenal medulla, differentiation of, 590, 591, 592 sympathetic nervous system and, 592, 593 vascular development of, 587, 592 Adrenaline, initial formation of, 577, 590, 591 pituitary-adrenal system and, 590-593 selective action of, 593 Afferent influx, 392 Affinity in neuron development, 356, 363, 371 between Schwann cells axons, 367 Alkaline phosphatase, 375 in amphibian limb regenera- tion, 683 Allometry. See Relative growth. Ameloblasts and enamel forma- tion, 494, 496 Amitosis, 91, 119 Amphibian egg, twinning of, 127-— 128 Amphibians, behavioral develop- ment in 546, 547 energy sources in, 535-538 exogastrulation in, 707 eye regeneration in, 690, 693 fields in regeneration in, 677, 678 and Amphibians, head structures, re- generation of, 685 hybrids, abnormal development of, 708, 709 lens regeneration in, 690-693 limb-girdle development in, 429-436 limb regeneration in, 677-686 alkaline phosphatase in, 683 blastema, origin of, 679, 680, 681-683 determination of blastema in, 679 endocrine glands in, 686 epidermis in, 679-684 nervous system in, 685 skeleton, origin of, 678, 679 metamorphosis of, 647-648, 649, 650, 651, 652, 654655 anterior pituitary and, 574— 575, 635-637 cessation of, 656 endocrine control of, 631-637 iodine and, 634 neoteny and, 633, 634, 655 thyroid and, 631-635 notochord abnormalities in, 707, 708 nucleus and cytoplasm in devel- opment of, 127, 128, 130-138, 140, 141, 142, 143, 144, 145, 146 overripeness of eggs as terato- genic factor in, 709, 711 progressive differentiation in, 230-296 regeneration in, 677-686 regulation in embryos in, 677 species differences in regenera- tion of, 677 Amphioxus, regeneration of, 675 Anaerobiosis in development, 523 Androgenesis, 143-146 Aneuploidy and development, 137-138 ““Aneurogenic” limb, 361 Angioblast as source of blood ves- sels, 450, 451 Animalization in animal halves, 327, 328, 329 duration in, 329 of entire eggs, 329-330 graded intensity in, 329 gps! Animalization, hypotheses to ac- count for, 330-331 in sea urchins, 327-331 time required for, 328 Annelids, nervous system in re- generation of, 665 polarity in regeneration of, 668 progressive differentiation in, 318-321 regeneration of, 665, 668 Anode, 356 Anomalies. See Teratogenesis. Anophthalmia, 383, 711 Anterior pituitary in dwarf mice, function of, 159 effects of adrenal cortex on, 585, 586 growth and, 600-606 histogenesis of, 580, 605 hormonal deficiency of, 159 metamorphosis and, 574-575, 635-637 onset of ACTH activity in, 579, 589 histogenesis and, 590 of gonadotrophic activity in, 595-600 histogenesis and, 599 quantitative increase, 596, 597 of thyrotrophic secretion in, 578, 582 histogenesis and, 582, 583 quantitative increase, 583, 584 thyrotrophic function of, 574— 575, 635, 637, 654-655 trophic effects on adrenal cor- tex, 585 See also Pituitary. Anteroposterior axis, reversal of, 378, 387 Antibody, 370 cross reaction of, 557, 558, 559 kinds of molecues in, 558 natural, 564 placental transfer of, 564 Antibody-forming capacity, devel- opment of, 564-566 Antifertilizin, chemical nature of, 190 of eggs, 191 T2z Antifertilizin, function of, 193- 194 of sperm, 189-191 Antigen-antibody reactions, 185, 191 Antigens of blood cells, 563-564 of brain, heart, and spleen, 559, 560 in development, 556-564 differentiation of, 165 hapten as determinant of speci- ficity of, 557, 558 kinds of determining groups, 558, 559 of lens, 559, 560 neural, 370 “non-vitelloid,” 562 with serum specificity, 559, 560, 561, 562 Arrest, developmental, as terato- genic factor, 703, 705, 706 Ascidians, polarity of regenerates of, 665 regeneration of, 664, 665, 666, 667, 668, 671, 672 apyrase in, 672 dedifferentiation in, 667, 668 inhibition of, 671 stimulus for, 666, 667 temperature effect on, 666, 671 Asymmetry of heart rudiments, 457, 458 of “organizer,” 457, 458 “Attractions,” 351, 354, 356, 357, 362, 366 Auditory vesicle. See Ear vesicle. Auto-antibody concept, 566-567 adhesion of cells and, 567, 568 immune antibodies and, 567 relation to gene action, 568 to induction, 568-570 Autotomy, tail, 686, 687, 688 Axial differentiation of CNS, 306, 307, 308 Axis cylinder, 346, 349 elongation of, 350 outgrowth of, 350-361 Axis determination, 230-232 of ear, 417-420 Axis formation in birds, 302-303 provisional, 297 in teleosts, 299, 300, 301, 302, 304: Axis of polarity, cleavage planes and, 315 constituents along, 317 cortex and, 318 differentiation and, 318 establishment of, 317 ground substance and, 317 structural basis of, 318 visible differences along, 315- 316 Axon, 348 branches of, 350 caliber of, 347, 363-365, 367 elongation of, 350-351, 356, 360, 363, 367, 374 of medullary cells, 370 growth of, 363-365 tissue culture of, 349 Axoplasm, 356 catabolic degradation of, 364 damming of, 364 movement, rate of, 364 synthesis of, 364, 365 BaLrour rule in cleavage, 215 Behavior, development of, 390- 392, 546, 547 Bilaterality, establishment of, 331-333 sperm entrance point and, 331, 332 Birds, axis formation in, 302-303 beak regeneration in, 689 chemical teratogenic agents in, 105,712, 713. (14a 5 chondrodystrophy in, 700, 704, 706 Creeper factor in, 700, 706 feather regeneration in, 689 insulin as teratogenic factor in, 703; 712, 71327142) 715 lethal mutants im, 700 limb-girdle development in, 437438 micromelia in, 704, 705, 711, 714, 715 rumplessness in, 700, 701, 702, 703, 710, 712 virus as teratogenic factor in, 702, 714 See also Regeneration; Teleosts and Birds; Teratogenesis. Birefringence in neural tube, 359 Blastema in amphibian limb re- generation, 679, 680, 681-683 Blastomeres, interaction of, 324— 331 specification of, 318-323 egg cytoplasm and, 318-323 egg polar lobe and, 319-320 egg pole plasm and, 320-321 Blastulation, 221-223 mechanisms of, 223 Blood, 452-453 erythrocyte source, 452 Blood cells, antigens of, 563-564 isoagglutinogens of, 563 Blood islands as_ erythrocyte source, 452 potency of primitive cells of, 452, 453 Blood vessels, 361, 367, 375 angioblast as source of, 450, 451 blood flow effects on, 451 INDEX Blood vessels, blood pressure ef- fects on, 451 embryogenesis of, 450-452 Body size, 128-129 in dwarfs, 623 eye and, 624 individual cells and, 620 limitation of, 622 racial differences in embryo and, 622, 623 as reflection of cell number, 620 Bone, regeneration of, 690 Brain, 361, 370, 378, 380, 383, 388, 389, 392 antigens of, 559, 560, 561, 562 compression of, 372 cortex of, 360, 372, 376 macrolocalization in, 388 microlocalization in, 388 early morphogenesis of, 371- By/D herniation of, 372 potentials of, 376 size of, 376 structure of, 359 ventricle of, 371, 373 ciliary beat in, 372 Brain hormone, action on pro- thoracic gland, 639, 640 of neurosecretory cells, 640 Branches, nerve, 346, 347, 349, 360-361, 368, 381 abortive, 368 collateral, 347, 360, 361, 365, 368, 391 cutaneous, 362 incidence of, 360 muscular, 362 overproduction of, 368 terminal, 360 Branchial skeleton, 372 “Buengner’s cords,” 349 CaLIBER of axon, 347, 363-365, 367 of nerve fiber, 365 Carbohydrate as energy source, 533, 534, 535, 536 Carbon marking, value of, 32 Castration (embryo), 482-483 Causal analysis, 347 Causalnexus of events, 20, 27, 31 Cell(s), adhesion of, 567, 568 basophilic components of, 44 blood, 452-453 centrioles and kinetosomes in, 44 constitution of, 39-64 cortical gel in, 55-56 cytochemistry of, 40 degeneration of, 378, 381 in teratogenesis, 702, 709, 710 diameter of, 621 INDEX Cell(s), differentiation of, cell division in, 109 without cleavage, 528 endomitosis in, 117, 118 division of, 91-119 mitogenesis in, 111-115 mitotic time and length of in- terphase in, 114-115 proliferation control im, 113 respiratory metabolism and, 528-530 enzyme activity differences in, 81 enzyme concentration differ- ences in, 80-81 enzymes in, structural orienta- tion of, 82-83 fate of, determination of, 379 fibrogenesis in, 51-52 fibrous systems in, 46 fixation of, 41 glycolytic and oxidative mech- anisms in, 71-77 Golgi system in, 45 granular fractions of, 43-45 growth of, 375, 381 indifferent, 380 interaction of, 61-63 lineage of, 213, 214, 379 membrane of, 56-59 enzymes of, 58 molecular ecology of, 58-59 surface charge of, 59 metabolism of, 70-88 control of, 79-80 microsomes in, 44 migration of, 381 mitochondria in, 43 number at end of cleavage, 621 particulates in, 41-45 proliferation of, 378, 382 secretion granules in, 43 size changes in metamorphosis, 620, 621 sizes of, 620 in differentiated tissues, 621 in eggs, 621 at end of cleavage, 621 strains of, 380 character of, 379 structural analysis of, 39, 40 submicroscopic fibrous struc- tures in, 49-51 substrate and cofactors in, sup- ply of, 82 surfaces of, 350 complementariness of, 568, 569 interaction on cell shape, 62 unstabilization of, 381 synthesis of, 77-79 Cell complex, determination of, 379 Cell-lethals, 152, 159 220-221, “Central action systems,” 386, 391 Central discharge, generation of, 392 Central functions, “total pattern” of, 391 Central nervous system, 347-348 axial differentiation of, 306, 307, 308 development of, 370-389 group relations in, 366- 370 Centrifugation, 233, 234, 243, 263, 286 Cerebrospinal liquor, turgor of, 372 Cervical segments, 382 Chemotropism in nerve orienta- tion, 356 Chimeric eye, ments of, 408 Chondrodystrophy, 700, 704, 706 Choroid fissure, 403, 411 Chromatin diminution in Ascaris, 170-172 Chromosome distribution, effects of, 139-141 Chromosome mosaics and develop- ment, 138-139 Chromosomes, 172-174 See also Mitosis. Cleavage, 213-221 adaptation in, 215 Balfour rule in, 215 bilateral, 213. cell number and, 621 determinate and indeterminate, 216, 218, 220 energy release during, 529 in non-nucleated egg frag- ments, 216-218 without nuclei, 216-218 ooplasmic segregation and, 218, 219, 220 organizational plasticity and, 300 radial, 213 growth adjust- elimination — of, respiratory metabolism and, 528-530 respiratory rhythms during, 529, 530 size inequality in, 215-216 spiral, 213, 214 in teleosts, 298-300 Cleavage nuclei, totipotency of, 339 Cleavage pattern, cytoplasmic streaming and, 218 of mosaic vs. regulative eggs, 220 surface tension and, 218 Cleavage substances, 218 Cloaca and urinogenital sinus, re- actions to sex hormones, 481-— 482 723 Coelenterates, regeneration of. See Tubularia. Collateral branching of nerves, 347, 360, 361, 365, 368, 391 Colloids, 359, 360, 367 Color pattern. See Pigmentation pattern. Commissural neurons, 377, 380 Competence, 257-259, 276, 285 abnormal, in teratogenesis, 707, 708 of gastrula areas, 241, 242 genetic basis of, 263, 264 regional pattern of, 258-259 time pattern of, 257-258 Competition, 352 in regeneration, 669-671 Complement, appearance of, 564: Complementariness of cell sur- faces, 568, 569 in induction, 570 Conduction, nerve, 363, 367, 376, 385 velocity of, 347 Connections, nerve, intracentral, 359 peripheral, 361-363 terminal, 361, 362-363 Constriction of egg, 233 Contact affinities, 366, 372 “Contact guidance” in nerve ori- entation, 352, 354, 356, 360, 366, 384 Coordinated function, localization of, 391 Coordinated limb activities within cord segments, 386 “Coordinated” movement, 384 Coordinating mechanisms, 386 Coordination, 378, 384, 386, 388, 392 central origin of, 391-392 Coordination patterns, source of, 392 Cornea, induction of, 407 regeneration of, 512, 689 transplantation of, 385 Corpora allata, 343, 637 function of, 638 hormone of, 641 Corpora cardiaca, 637, 638 Corpus allatum hormone and pro- thoracic hormone, 641 Correlations in teratogenesis, 703, 706 See also Induction. Cranial ganglia, 381 Cranial nerves, 383, 386 Creeper fowl, 700, 706 Critical periods. See Sensitive periods. Cross-agglutination, 193 Cross-fertilization, 193 Cross reactions, 556, 557, 558, 559, 560, 561, 562 724 Cross reactions with haptens, 557, 558 interpretation of, 557, 558 Ctenophores, progressive differ- entiation in, 321 Cutaneous fibers, 385 Cyclopia, and local induction of, 307 Cyclostomes, regeneration of, 675 Cytochrome c, 376 Cytokinesis, 107, 109 factors involved in, 108-109 Cytoplasm and body size, 128-129 Cytoplasmic areas, chemical char- acterization of, 324 Cytoplasmic particles in differen- tiation, 284-286 Cytoplasmic streaming, cleavage pattern and, 218 DEDIFFERENTIATION in regenera- tion, 667, 679, 680 Degeneration, 382, 383, 388 cellular, in normal develop- ment, 378, 381, 709, 710 in teratogenesis, 702, 709, 710 “Demyelinization,” 388 “Dendrites,” 351, 376 Dendritic fields, 356, 382 Denervated field, 365 Denervated tissue, from, 368 Dermis, 499 dermatome and, 500, 501 neural crest and, 500, 501 somatopleural origin of, 500, 501 Desoxyribonucleic acids, 365 Determination of axes, 230-232 of cell complex, 379 of choroid fissure, 403 of dorsalization, 232-234, 271, 283 of ectodermal derivatives, 255, 256 of heart, 443, 444 of imaginal discs, 340, 341 of individual cell fate, 379 by induction, 338 initial steps of, 232-234 in insect development, 337-338 of lens regeneration blastema, 692 of limb regeneration blastema, 679 of medullary plate, 250-254 metabolic patterns and, 282-283 of mouth implements, 263-264 of otocyst, 258 progressiveness of, 338, 340 of retinal cup, 402-404 symmetry in, 231-232 Development, activity metabolism in, 522 anaerobiosis in, 523 emanations Development of androgenetic hy- brids, 143-146 aneuploidy effects on, 137-138 of antibody-forming capacity, 564-566 antigens in, 556-564 asymmetry determination in, 161-163 of blood vessels, 450-452 Boveri, 19 of brain, 371-372 causal analysis, beginning of, 17-18 cell-lethals in, 152, 159 chemical modifications of, 329- 331 chromosome mosaics on, 138— 139 of complement, 564 cytoplasmic factors of, 126-128 differentiation in, 521 of diploid hybrids, 141-143 Driesch, 19, 20, 21 energy release in, 522-523 energy requirements for, 520- 522, 528-533 energy sources during, 533, 538 enzyme activation in, 87 enzyme activity in, 84 enzyme concentration, 84 of enzymes, 543-548 functional differentiation and, 545-548 of forelimb, 429-431, 437-438 genic material for, 151-152 of girdle, 434435, 436, 437 growth in, 520-521 of haploid hybrids, 143-146 haploidy effects on, 130-134 Harrison, 21 His, 17, 18, 19 homeotic mutants in, 161 of immunological properties, 556-573 maintenance and, 531-532, 533 maintenance energy vs. energy for, 528-533 maintenance metabolism dur- ing, 520 mechanical explanations of, 17, 18 metabolic rate in, 524-525 metabolic regulation in, 83-88 of natural antibodies, 564 nuclear differentiation in, 146 without nucleus, 129 nucleus and cytoplasm in, 126- 150 oxygen consumption during, 523-528 oxygen supply during, 522-523 of polyploid hybrids, 146 polyploidy effects on, 134-137 of red cell antigens, 563, 564 respiration during, 523-533 INDEX Development, Roux, 18, 19, 20, 21 Spemann, 19, 21 yolk utilization in, 527 Developmental block in insects, 5307531 metabolism during, 530-531 Developmental interactions in in- sects, 341, 342 Developmental processes, depend- ent and autonomous, 159-161 Diapause, 645 hormonal breaking of, 639, 640 in insects, 530-531 Differential growth, 353, 360, 365, 374 Differentiation of antigens, 165 autonomous and dependent con- trol of, 159-161 of cells, 109, 117, 118, 220-291, 528 without cleavage, 323-324 cytoplasmic particles in, 284- 286 in development, 521 differential genic effects in, 164-165 diffusible substances in, 160- 161 of ear axes, 417-420 of ear ectoderm, 415, 417, 418, 419, 420 energetics of, 521 energy requirements in, 521, 522 enzymes and metabolism in, 331 of gastrula areas, 238-241 genic endowment and, 163-166 histochemistry of, 376 hormones in, 342, 343 of medullary plate, 252-254 of neurula mesoderm, 247- 249 non-autonomous, 160-161 of nucleus, 146 potencies, in gastrula ectoderm, 238 in gastrula entoderm, 240 in marginal zone, 238-239 progressive. See Progressive dif- ferentiation. of secondary embryos, 243-245 of tail bud, 249-250 See also Self-organization. Disaffinity in neuron develop- ment, 371 Dissociation sponges, 668 Dorsal root, 356, 362, 381 deflection of, 360 Dorsalization, 232, 233, 234, 271, 283 Dorso-ventral inversion of limb cord segments, 387 experiment in INDEX Dorsoventrality, establishment of, 331-333 reversal of, 332 Duplications, 707, 713 Dwarf mouse, anterior pituitary of, 159, 605 endocrine deficiency in, 586 growth cessation in, 604 Dwarfs, body size in, 623 Ear, organogenesis of, 415-423 plate (placode) of, 415 prospective areas of, 415, 416 vesicle (otocyst) of, 415 Ear capsule, formation of, 422- 423 induction of, 422 Ear ectoderm, 415 axial determination of, 417- 420 ciliary beat of, 420 equipotentiality of, 415, 419 polarization of, 417-420 regenerative capacity of, 415 Ear plate (placode), normal fate of, 415 Ear vesicle, capsule formation of, 422-423 induction of, 420-422 mesodermal activation of, 420, 421, 422 morphogenetic influence on, 416 as mosaic, 415 neural activation of, 420, 421, 422 Echinoderms, gradients in egg of, 665, 666 progressive differentiation in, 327-331 Ectoderm, ear. See Ear ectoderm. primitive organization of, 306- 309 Ectodermal derivatives, mination of, 255, 256 Effector, 351 Egg, activation of, 202-203 animalization of, 329-330 antifertilizin of, 191 axial polarity of, 315-318 centrifugation of, 233 constriction of, 233 cytoplasmic areas of, 324 cytoplasmic specification of, 316-317 diameter of, 621 fertilizin of, 185-189 fusion of, 233, 298 inductive capacity of, 233, 265 marginal plasma of, 230, 233, 234 organization of, cortex and, 318 genotype control of, 317 stratified constituents of, 317 polar lobe of, 319-320 deter- Egg, polar organization of, 317, 318 pole plasm of, 320-321 reaction of, 194-201 rotation of, 230, 231 senescence of, 184 spermatozoon approach to, 181-— 182 structure of, 230-231 Electric effects, 356, 357 Electric fields, 350, 356 Electric guidance, 356 Electric potentials in hydroid re- generation, 668, 669 Electronmicroscopic demonstra- tion of fiber collaterals, 368 Eleocytes, 665, 667 Embryo, interference with, 31- 37 normality of, criteria of, 27, 28 Embryogenesis, energy exchange and enzymes during, 520- 555 in teleosts and birds, 297-314 See also Development. Embryological problems, methods and techniques in, 25-37 molecular level, view of Roux, 27 observation vs. interference as approach to, 26-31 Roux’s analysis of, 31, 32, 33, 35, 36 Embryology, evolution and, 13- 22 history of, 1-22 Naturphilosophie in, 8-13 Embryonic axis and cleavage (tel- eost), 298, 299, 300 Embryonic induction between germ layers, 338-339 in insects, 338-339 Embryonic organization in initia- tion of insect egg, 337 Embryonic shield in teleosts, 302, 304, 305, 307 Enamel organ, ameloblasts of, 494: organizing influence of, 494 Endocrine correlation, anterior pituitary—adrenal, 585-593 anterior pituitary—gonad, 593-600 anterior pituitary and growth, 600-606 anterior pituitary—thyroid, 574-585 ontogeny of, 574-619 Endocrine glands in amphibian limb regeneration, 686. See also Hormones, and specific glands. Endocrine receptor, 583, 597, 605 Endomitosis in cell differentiation, 117, 118 Endoneurial fluid, 346 End-organ, 347 125 Energy release, alternate path- ways of, 541-542 during cleavage, 529 cyanide-insensitive tion and, 541-542 during development, 520-523, 528-533 non-phosphorylating glycoly- sis and, 542-543 phosphogluconic acid shunt and, 542 Energy sources in amphibian em- bryo, 535-538 from carbohydrate, 533, 534, 535, 536 in chick embryo, 533 during development, 533-538 from fat, 533, 534, 535, 538 in fish embryo, 534 in grasshopper embryo, 534 ontogenetic sequence of, 533- respira- 538 from protein, 533, 534, 535, 536, 537 in sea urchin embryo, 534— 535 utilization changes in, 538 Entelechy, 20 Entoderm, potency localization in, 310-311 Entodermal derivatives, organo- genesis of, 453-458 Environmental factors in terato- genesis, 701, 702, 711 Enzymes, activation in develop- ment, 87 activity in embryos, 84 adaptation of, 165 adaptive, 85 of cell membrane, 58 concentration in embryos, 84 in differentiation, 331 gene control of, 165 in ontogenesis, 543-548 relation to functional differen- tiation, 545-548 cholinesterase and behavior in, 546, 547 respiratory, synthesis of, 543 Epidermal ridges, mesodermal initiation of, 510-512 regeneration of, 512 Epidermis, 499 dermal papilla action on, 509 differentiation-dependency of, 308 in limb regeneration, 679-684 mesodermal role in specializa- tion of, 507-512 physiological regeneration of, 689-690 regeneration of, 512 See also Skin. Epigenesis, 2, 3, 4, 5 Aristotle and, 3, 4, 5 726 Epigenesis, Harvey, views on, 4, 5,6, 12 influence of microscope on, 7, 8 Wolff and, 7, 8, 11, 21 Equipotentiality of ear ectoderm, 415, 419 Erythroblastosis, 702 Erythrocytes, isoagglutinogens of, 563 mosaicism of, 566 nature of, 452, 453 source of, 452 Evocation, 259, 270, 280 Evolution, biogenetic law and, 15, 16 Darwin and, 13, 14, 15, 16, 17, 18 embryology and, 13-22 Haeckel and, 15-17, 18, 21 Excitability, thresholds of, 363 Excitation, conduction of, 350 Exogastrulation, 224, 226, 235- 237, 707 Explants, 350 from M-cell region, 379 External genitalia, castration ef- fects on, 483 sex hormone action on, 482 Eye, 380, 383, 386 growth of, 407-409 growth adjustments of, 408, 409 initial size of, 407, 408 muscle size of, 409 orbit size and, 409 organogenesis of, 402-414 regeneration of, 409-412, 690, 693 size relation of, 628 transplantation of, 358 See also Cornea, Lens, Retina. Eye-forming area, 307 FAscICULATION, 361, 366-367, 368, 380, 382, 384 Fat as energy source, 533, 534, 535, 538 Fate maps of gastrula, 234, 235, 301, 302, 303, 305 of heart, 441 of neurula, 247 of wing bud, 437 Feather, dermal papilla as induc- tor of, 509 development of, 507, 508, 509 regeneration of, 689 Feather germ. See Feather pa- pilla. Feather papilla, dermal compon- ent of, 504 development of, 507, 508, 509 induction of, 509 melanoblast invasion of, 504 mutational change of, 515 reaction gradients of, 515 Feather papilla, response to hor- mones, 507, 514, 515 time-space order of, 506 Feather tract, 506, 507, 514, 515 Fertilization, 181-201 antifertilizin role in, 193-194 block to polyspermy in, 194: cones of, 195 cortical change in, 194 electrical changes upon, 197 fertilizin-antifertilizm action in, 193-194 fertilizin role in, 193-194 lytic agents in, 194: membrane elevation in, 195— 196 metabolic changes upon, 197— 199 permeability changes upon, 197 pronuclei union upon, 199-201 protein solubility change upon, 196 shape-volume change upon, 196 specificity of, 193-194 viscosity change upon, 196 Fertilizin, 185-189 chemical nature of, 188, 189 function of, 193-194 Fertilizin-antifertilizin 194 Fetal cortex (adrenal), 586, 587 Fibers, cutaneous, 385 nerve. See Nerve Fibers. Fibrin, 352, 353, 356 Fibrinolytic agents, 353 Fibrogenesis, 51-52 Fibrous systems in cells, 46 Fibrous units in nerve orientation, 352 Field of thyrotrophic substance, 583, 654 Fields, 370, 378 in amphibian regeneration, 677, 678 definitive position and restric- tion of, 312 dendritic, 356, 382 denervated, 365 of ear, 415 electric, 350, 356 gene action on, 165-166 heart-forming, 310, 444 of lower germ layers, 309-311 mosaic of, 370, 375 nephros-forming, 310 for notochord-somite formation, 309 See also Morphogenetic fields. Forelimb, disc of, 431, 432 duplication of, 435 location of, 432 normal development of, 429- 431, 437 as_ self-differentiation system, 431 reaction, INDEX Forelimb. See also Wing bud. Forelimb and girdle, organogen- esis of, 429-439 “Formative turgor,” 364 Freemartin, 475, 477, 483, 593, 594: Functional differentiation and en- zyme development, 545-548 Functional localization, 387 in brain cortex, 388 Fusion of amphibian eggs, 233 of teleost eggs, 298 GALVANOTROPISM, 356 Gametes as teratogenic factors, 711 life span of, 182-185 Ganglia, 347, 354, 373, 381, 383 Gastrula, fate maps of, 234, 235, 301, 302, 303, 305 isolation experiments on, 238— 241 organization of, 238-243 Gastrulation, 223-227, 234-235 cell movements in, 225 cell number in, 224, 225 cell shape in, 224, 226 forces of, 226-227 mechanisms of, 224-227 partial, 236, 237 in teleosts and birds, 300-306 types of, 223 Gene action. See Genic action. Generation, Aristotle’s epigenesis and, 3 contribution of male and fe- male in, 2 Fabricius, influence of, 4 Harvey’s views on, 4, 5, 6 part formed first in, 2, 3, 4 theories of the Greeks on, 1-3 Genes, differential response of, 164 Genetic asymmetries, 161-163 Genetics and embryology, 25, 28, 32 Genic action, 151-166 auto-antibody concept and, 568 in developmental processes, 159-161 embryonic fields and, 165- 166 site of, 159-161 with substrate, 164 time of, 155-157 type of, 157-159 Genic content, differentiation and, 163-166 qualitative variability of, 153-155 quantitative variability of, 152-153 Genital ridge, formation of, 473 sterility of, 473 INDEX Genital structures, accessory, 477— 483 castration effects on, 482— 483 gonad grafting effects on, 478-479 hormone administration ef- fects on, 479-482 See also Sex structures, ac- cessory. Genital system, 470-486 Germ cells, determination of, 174 gonad formation and, 175-— Leia origin of, 170-177 in insects, 339 site of origin, 175, 176, 177 Germ ring (teleost), 299, 300, 302, 305 axial structures and, 305 Gill formation, entodermal role in, 455 Girdle, development of, 434-435, 436-437 Glands, regeneration of, 690 Glia cells, pulsations of, 364 Golgi system in cells, 45 Gonad, cortex and medulla of, 470 formation of, 473 germ cells of, 470, 471-473 localization of structural ele- ments in, 470-471 onset of hormone activity in, 578, 593-595 origin of, germ cells and, 175- 177 reciprocal action on anterior pituitary, 599 responsiveness to gonadotroph- ins, 597-598 sex differentiation of, 475-477 sex reversal of, 475 sex-specific organization of, 474: sexually indifferent, 470 sterile, 473 Gonad-forming area, 471, 474 map of, 464 Gonaducts. See Sex ducts. Gradient fields in archenteron in- ductivity, 255 of brain-eye, 252, 253 in lens inductivity, 278 in mesodermal mantle of neu- rula, 248 Gradients, 234, 282-284, 356 in annelid regeneration, 665 axial, 282, 665 concentration, 356, 358 double theory of, 234, 283 inside-outside, 283-284: in pH, 350 of reaction in feather papilla, 515 in sea urchin egg, 330-331, 665, 666 Graft incompatibility, 565 Graft tolerance, induction of, 566 Gray crescent, 230-232, 233 Gray matter, 363, 367 “Ground substances,” 352, 360 Group relations (CNS), develop- ment of, 366-370 Growth, antibody stimulation of, 569 cell division and enlargement to, 624 of cells, 375, 381 of chimeric eye, 408 complexity of factors in, 602, 605, 606 curves of, 623, 624 definition of, 520, 521 differential, 353, 360, 365, 374 energy requirements in, 521, 522 of eye, 407-409 of eye muscles, 409 formal treatment of, 30 glutathione concentration and, 623 gradient of, 625, 627 hormonal activation of, 343 hormonal interaction in, 600, 601, 602, 603, 604 hormones in, 342, 343 inhibition of, 623 internal force and, 624 intrinsic capacity for, 603, 605, 606 moulting and, 342, 343 of oocyte, 177-180 pattern, 373-376 potential, 350 pressure, 368 rates of, 375-376, 622, 623-625 during embryonic period, 622, 623 of regenerating organs (or parts), 627 regulation in eye and limb, 628 relative. See Relative growth. respiratory increase and, 525, 526 self-duplicating entities in, 567 temperature and, 628, 629 of tooth, 492-497 Growth hormone, history of con- cept, 600 onset of activity, 580, 604, 605 synergistic action of, 601, 602, 603, 604, 605 Guide structures, 353, 354, 359 Gut formation, mesodermal role bak Sill Hair, dermal papilla of, 510 as inductor of, 510 primordia of, 509 stages in formation of, 509, 510, 511 P=] Hair papilla, components of, 510 Haploidy and development, 130- 134 Haptens, cross reactions with, 557, 558 Harmonious morphological sys- tem of limb and girdle, 434 Head structures, amphibian, re- generation of, 685 Heart, determination of, 443, 444: duplication of, 445, 446 induction of, 442, 443 localization of, 440, 441, 442 organogenesis of, 440-450 primordia of, 440, 443, 444 axial determination of, 445 polarity of, 445 self-differentiation of, 444 totipotency of, 445, 446 self-differentiation of, 443, 444 Heart (embryonic), beat rate of, 443, 448 electrocardiogram of, 448, 449 initial contractions of, 446, 447, 448 initial striations of, 448 innervation of, 450 intrinsic rhythms of, 448, 449 myogenic contractions of, 449, 450 nervous control of, 449-450 pacemaker of, 447, 450 “tubular” formation of, 446, 447 Heart-forming area, 310, 441, 442, 443 map of, 443 organizing role of, 311 Hereditary factors, phenocopies of, 702-703 in teratogenesis, 700-701 Hindbrain, 389 Histogenesis, 576-581 of anterior pituitary, 578, 579, 580, 582, 583, 590, 599 in dwarf mouse, 605 Homeotic mutants, 161 Homoeosis in insects, 341 Hormone(s), action in imsects, 638-644: of brain, 640 of corpus allatum, 641 in development, 347, 388, 574 619 in differentiation, 342, 343 feather papilla response to, 507 functional interlocking of, 602, 603, 604, 605, 610 growth and, 343 in insects, 342, 343 local action of, 389, 583, 594 metamorphic mechanism and, 653-656 728 Hormone (s), onset of chemical in- tegration of, 610, 611 pigment cell response to, 513, 514, 515 of prothoracic gland, 641 sex. See Sex hormones. skin response to, 505, 506 specific action of, 343 Humoral activity, correlations in development of, 610 Humoral milieu, 388-389 Hybridization, 141-146 Hybrids, androgenetic, 143-146 cell degeneration in, 709 development of, 141-146 developmental abnormalities in, 708 organs of, 388 Hydrocephalus, 372 Hydroids. See Tubularia. Hyperdactyly, 383 ““Hyperneurotization,” 369 Hyperplasia, 375, 381 Hypophysis. See Anterior pitui- tary. Hypoplasia, 381 Hypothalamus, 365 ImMaGINAL discs, 340, 341 differentiation of, 643-644 hormonal control of, 342, 343 Immigration of nerve cells, 375 Immunological properties, devel- opment of, 556-573 Immunology, cross reactions in, 556, 557, 558 Incompatibility of tissues, 371 in grafts, 565, 566 in vitro, 566 Individuation, 259, 280 neural basis of, 391 Induction, 235-238, 243-247, 250- 252, 254-256, 265-282, 385, 387, 568-570 archencephalic, 250-252, 269, 971 assimilative, 244, 248 of auditory vesicle, 420-422 auto-antibody concept and, 568— 570 by cell injury, 272-273 chains of, 256-257 chemical aspects of, 267, 270- 972, 278-279 complementary, 244, 570 of cornea, 407 of cyclopia, 307 cytoplasmic particles in, 285- 286 by dead tissue, 265-269 determination by, 338 deuterencephalic, 250-252, 269 of ear capsule, 422 evolutionary aspects of, 264-265 of feather papilla, 509 Induction, field characteristics of, 280-281 of heart, 442, 443 homodynamic, 264-265 homoiogenetic, 256 kinetic cell properties, effects on, 273-276 of lens, 404-407 of middle ear, 423 of nasal placode, 423-424 of neural crest, 254-255 of non-neural ectodermal struc- tures, 255-256 by nucleic acids, 271-272, 285, 569 by organizer, 243-247 of otocyst, 258 physiological mechanism of, 265-279 potency release in, 304 by primitive streak, 306 qualitative aspects of, 279 quantitative aspects of, 278 in regeneration, 667, 668, 678, 693 by relay, 270-273 spino-caudal (trunk-tail), 247, 250-252, 259, 269, 271 in teratogenesis, 706, 707, 708 xenoplastic, 246, 263-265 Induction fields, 250, 252, 255, 278, 307, 312 Inductor systems, 255-257, 276 Inductors, chemical identification of, 270-272 distribution in animal kingdom of, 267-268 general properties of, 275-276, 285-286 nucleic acids of, 271-272, 285 regional specificity of, 245-247, 250-252, 259- 261, 269 secondary, 256-257, 279-280 See also Induction. Inheritance of acquired charac- ters, early views on, 2 Innervation, 361 compensatory increase of, 368 density of, 347 Insect egg, activation center of, 337 cortical plasm of, 339 cytoplasmic determiners in, 278- 339 determination in, 337, 338 developmental imteractions in, 341, 342 differentiation center of, 337 germ band of, 337 imaginal discs of, 340, 344 mosaic vs. regulative types, 338 organ fields in, 340, 344 pole plasm of, 339 INDEX Insect egg, regulative capacity of, 337, 338 Insect metamorphosis, 645, 648, 650, 652-653, 655-656 cessation of, 656-657 controlling factors in, 637- 644 endocrine relations in, 638, 639, 640, 641 larval molting in, 641-642 pupation molt in, 642-644: Insects, diapause in, 530-531 energy source in embryo of, 534 homoeosis in, 344 hormones in differentiation of, 342, 343 progressive differentiation in, 337-345 Insulin, dosage effects of, 715 hereditary differences in sus- ceptibility to, 712 sensitive periods to, 713 as teratogenic agent, 703, 705, 712, 715 Insulin secretion, unfolding of regulatory mechanisms of, 607— 611 Integrated movement, 386 Integumentary patterns, 512- 515 genes and morphology of, 512, 513 See also Skin. Interaction of blastomeres, 324— 331 in echinoderm eggs, 327-331 gene-substrate, 164 in mosaic eggs, 324-327 nucleo-cytoplasmic, 146 Interfaces, 352, 356 Interference techniques, appraisal of, 31-37 Internodes, 367 length of, 347 Internuncial cells, 380, 383 Intestine, coiling of, 456 regeneration of, 690 Intracentral nerve connections, expansion of, 391 Intracentral nerve tracts, 359 Intracentral neurons, 363 Inversion experiment, amphibian egg, 232, 234 Iodine in amphibian metamorpho- sis, 634 Islets of Langerhans, 606-611 alpha and beta cells in, 580, 581, 606, 607 secretory activity and, 607 insulin secretion of, 607 morphogenesis of, 606 Isoagglutinogens of blood cells, 563 INDEX Isolation of blastomeres, 318, 319, 320, 321, 323, 324, 325, 326, 327 of egg fragments, 319, 320, 321, 323 LaByRINTH (membranous), axial modification of, 417, 418, 419, 420 capsule formation of, 422-423 Laminated structure of myelin sheath, 367 Larval molting, hormonal control of, 342, 343, 638, 641 Lateral line, 362 Learning, 384, 392 Lens, antigens of, 559, 560 dependent differentiation of, 404, 405 fiber formation in, 407 from “foreign” epidermis, 404, 405 induction of, 308, 4044.07 factors responsible for, 406 initial size of, 408 from iris border, 308 from iris epithelium, 411 regeneration of, 410-412, 690- 693 reinforcing inductors of, 405 specific proteins of, 407 suture lines of, 407 Wolffian regeneration of, 410- 412, 690-692 Lens-inducing stimulus, nature of, 406 Limb, 361, 362, 374, 381, 382, 392 abnormalities of. See Chondro- dystrophy, Creeper fowl. “aneurogenic,” 361 control by spinal cord, 387, 388 exchange of, 388 formation of, mesoderm in, 433 function of, defective, 387 movements of, 378, 384, 391 normal development of, 429- 431, 436-437 regeneration of, 369 in amphibians. See Amphib- ians, limb regeneration in. blastema in, 679, 680, 681- 683 endocrines in, 686 epidermis in, 679-684 field of, 677 mesoderm in, 680-684: nervous system in, 685 phases in, 681, 683 in reptiles, 686, 687 size regulation of, 628 suppression of, 381 transplantation of, 368, 380, 382 to head, 386 Limb area (disc), 431 equipotentiality of, 433, 434 fate of, 431 location of, 432 potency of, 432 Limb bud, 358, 361, 380, 431, 432, 433 Limb centers, disarrangement of, 387 Limb muscle, transplantation of, 384 Limb plexus, 358, 384 Lipid development in adrenal cor- tex, 588 Liver, differentiation of, heart in- ductor and, 456 extirpation effects of, 456 onset of function, 457 regeneration of, 456 Liver-pancreas, labile determina- tion of, 456 self-differentiation of primor- dia, 455 Localization of cytoplasmic areas, 316, 321, 322, 324 of heart, 440, 441, 442 Lytic agents, chemical nature of, 192, 193 function of, 194 of sperm, 192-193 MAINTENANCE and development, mechanisms of, 531, 532, 533 Malformations by chemical-phys- ical agents, 242-243 origin of, 237-238, 242-243 Mammals, chemical teratogenesis in, 702 erythroblastosis in, 702 hereditary malformations in, 700, 701, 702, 703, 704, 708, 710 limb-girdle 437-438 vitamin deficiencies as terato- genic agents in, 701, 712 x-ray induced malformations in, 701, 710, 715 Marginal plasma, 230, 233, 234 Marginal zone, 233, 234 potencies of, 238-242 Matrix, 353, 358, 361 colloidal, 352, 359 Maturation, 381, 382 of nerve centers, 383 of ventral horn cells, 382 Maturation divisions, 180-181 Mauthner’s neuron, 377, 378, 379, 389 fibers of, 362, 363 M-cell region, explants from, 379 Medulla oblongata, 363, 375, 378, 379, 387, 392 Medullary cells ment of, 373 development in, (CNS), move- 729 Medullary plate, induction of, 250-254: organization of, 252-254 Melanoblasts. See Pigment cells. Melanophores. See Pigment cells. Membrane, cell, 56-59 Mesoderm, neurula, 247-252 Mesonephric areas, position of, 463, 464 Mesonephric potency, 464 Mesonephros, development of, 468 forming area of, 464 nephric duct effects on, 468 Metabolic enzymes and mental derangements, 388 Metabolic patterns, induction of, 282-283 Metabolism of cell division, 110- 111 of cells, 70-88 in development, 524-525 in differentiation, 331 invertebrate regeneration and, 672 metamorphosis and, 646-647 regulation of, 83-88 Metamorphic mechanism, activa- tion of, 653-656 in amphibians, 654-655, 657 in insects, 655-656, 657 Metamorphic stimulus, action of, 644-653 Metamorphic transformation, ces- sation of, 656-657 Metamorphosis, 381, 391, 631-663 in amphibians, 631-637, 647- 648, 649, 650, 651, 652, 654-655 anterior pituitary and, 574— 575, 635-637 of brain, 389 cell size change in, 620, 621 hormonal mechanism of, 644, 646, 653-656 in insects, 637-644 metabolism and, 646-647 pattern of, 650 in amphibians, 650-652 in insects, 652-653 tissue response in, 647-648 sensitivity of, 649-650 specificity of, 649-650 Metanephros, correlations in de- velopment of, 469, 706 Methods and techniques, appraisal of, 25-37 Microcephaly, 706, 714 Micromelia. See Chondrodystro- phy, Creeper fowl. Microphthalmia, 383, 711 Microsomes, 44, 521 Midbrain, 361, 374, 379, 380, 383, 385, 387 83-88, 520, 3855580, 730 Middle ear, formation of, 423 Mitochondria, 43, 521 cytochrome oxidase and, 540 Mitogenesis, 111-115 biochemical changes in, 112- 113 generation time of, 114-115 “latent period” in, 114 Mitosis, centrosome, 100 chromosomal fibers in, 104-105 birefringence of, 102, 104 chromosome chemical changes in, 95-96 chromosome cycle, 94-98 chromosome movements in, 98— 107 chromosome reproduction in, 96-97 chromosome structural changes in, 94—95 chromosomes in interphase nu- cleus, 97-98 cytoplasmic changes in, 109- 110 description of, 91-94 differential, 109 hormonal effects on, 113 inhibition of, 116-117 kinetochore in, 103-104 metabolism of, 110-111 metakinesis in, 105-106 modification of, 117 nucleolus in, 98 spindle in, 101-103 and astral fibrilization in, 52 See also Cell division. Mitotic activity in CNS, 373 dorsoventral differential of, 375 Modulation, 383-386, 387, 391 of neurons, 385 Molecular ecology, 27 Molecular population, 376 Mollusks, progressive differentia- tion in, 318-320 Molting, humoral control of, 342, 343, 638, 644 Morphogenesis, physical-chemical considerations of, 59-61 See also Development. Morphogenetic field, 279-282 chorda-mesoderm (“‘organ- izer”), in, 239, 240, 242, 279-282 of eye-forebrain, 252-253 gradient of egg, 234 induction of, 250, 280 in neurula mesoderm, 248 in regeneration, 677-678 regulation of, 63-64: self-organization of, 279-281, 284. size related to patterning, 281-282 Morphogenetic movements in gastrulation, 234-235 as organizing factors, 235- 238, 241, 281 Morphogenetic potential, 243, 248, 278, 283 Morphogenetic processes, 59-61 Morphology, new form of, 29, 30 Motor cells, 377, 389 selective excitation of, 384 Motor coordination, 392 Motor fiber, predestination of, 384 Motor neurons, 381, 382, 384, 385, 391 specification of, 384 Motor performance, 391 Motor roots, 362, 368, 382 Motor unit, 360 Mouse, fusion of vertebrae in, 701 hereditary bleb formation in, 710 hereditary tail abnormalities in, 700, 702 shaker-short mutant of, 708 x-ray as teratogenic factor in, 701 See also Dwarf mouse. Mouth, determination of, 263-264 induction of, 454, 455 Movement, integrated, 386 Muscles, 361, 362, 363, 369, 381, 383, 384, 385 denervation of, 361. 368 extra set of, 384, 386 fibers of, 360 phylogenetic relations of, 351 regeneration of, 690 transplantation of, 384 Myelin, 347, 349, 367, 375, 388 Myelin sheath, 350 laminated structure of, 367 Myelinization, 367 Myotomes, segmental arrange- ment of, 372 “Myotopic” function, 384 Nasa cartilages, formation of, 424: Nasal passageways, shaping of, 424. Nasal placode, dependency in ori- gin of, 308 differentiation of, 423 induction of, 423-424 Natural antibodies, development of, 564: Natural auto-antibodies, concept of, 566-567 of egg and sperm, 566, 567 Naturphilosophie, 8-13 Goethe, 8, 9, 10, 11, 13, 21 von Baer, 11, 12, 13, 14. 16, 91 Neoblasts, 667 Neoteny and amphibian meta- morphosis, 633, 634, 655 INDEX Nephric duct, caudal growth of, 4.67 origin of, 466-467 territory influence on, 467 Nephric fields, 310 differentiation effects of, 464, 465 Nephric system, developmental plan of, 463 units of, 462 Nephrogenic areas, topographic localization of, 462-464: Nerve, cranial, 361, 383, 386 cross unions of, 363 deflection toward growing or. gans, 358-359 degeneration of, 358 regeneration of, 349-350 fiber numbers in, 369 repetition of, 364 stereotypism of, 361 transection of, 349 Nerve branches. See Branches, nerve. Nerve cell body, size of, 365 Nerve cells, 370 immigration of, 375 Nerve conduction, 363, 376, 385 salutatory theory of, 367 velocity of, 347 Nerve connections, intracentral. 359, 391 peripheral, 361-363 terminal, 361, 362-363 Nerve cord, diameter of, 375 early morphogenesis of, 371- 372 Nerve courses, deflection of, 359 Nerve fibers, 352 atrophy of, 365 bundles of, 367 caliber of, 365 cell body of, 350 constriction of, 363 deviation of, 356 diameter of, 363, 367 filopodia of, 352 internuncial, 359 intracentral, 350 invasion of, 381 orientation of, 351 outgrowth of, 349, 356 “pioneer,” 351 recurrent, 356 resorption of, 361 sympathetic, 362 tracts of, 347, 357 longitudinal, 359, 366 “Nerve modulation,” 384 Nerve orientation, 352 Nerve patterns, 358-361 Nerve roots, segmental arrange- ment of, 372 Nerve stump, 349, 350, 352, 353, 357, 358, 362 INDEX Nerve tracts, intracentral, 359 Nervous system, cholinesterase ac- tivity of, 546, 547 development of, 346—401 in amphibian limb regenera- tion, 685 Neural antigens, 370 Neural arches, segmental arrange- ment of, 372 Neural axis and primitive streak, 307 Neural crest, 370, 371 dermis and, 500, 501 induction of, 254-255 morphogenesis of, 372 oral ectoderm and, 455 skin pigment cells and, 502, 503, 513 visceral arch and, 455 Neural epithelium, defects im, 379 diversity within, 380 precursor cells in, 379 Neural plate, 370, 371, 374, 387 folding of, 370, 379 localization of, 307, 308 Neural tube, 359, 370 Neuroblasts, 348, 379, 381, 382 cytoplasm of, 350 Neurofibrils, 376 Neurogenesis, 346-401 nutrient requirements in, 388 nutritional deficiencies in, 366, 372, 388 “Neuromas,” 350 Neuromuscular activity and cho- linesterase development, 546, 547 Neuron growth, 389 biochemistry of, 365-366 factors controlling, 365 Neuronal connections, precise pat- terns of, 387 Neurons, 350-370, 381-383, 385 cathodal site of, 356 commissural, 377, 380 intracentral, 363 “overloaded,” 365 pools of, 392 size of, 365 size classes of, 365 specificity of, 347 Neuropil, 347, 352, 360 density of, 370 Neurosecretions, 347 Neurosecretory cells, brain hor- mone of, 640 of insects, 637 Neurotropism, 356-358 Neurula, heterotropic grafts of, 378 inductive capacity of mesoderm in, 250-252 organization of, 247-250, 252- 254 Neurulation, 370-371 Node of Ranvier, 361, 367, 368 Noradrenaline, 590-591 Nose, 362 cartilage formation of, 424 nerve fiber origin of, 424 organogenesis of, 423-424 prospective areas of, 415, 416 sensory epithelium of, 424 Notochord, 309, 371, 373 abnormalities in amphibian metamorphosis, 707, 708 Nucleic acids, in induction, 569 Nucleus, 364, 365, 367, 376 development and, 129-147 differentiation of, 146 enlargement of, 365 transplantation of, 146 OBSERVATION vs. interference, 26— 31 limitations to interpretation from, 26-28 Boveri on, 32 Odontoblasts and dentin forma- tion, 494, 496 “One-center effect,’ 354 Ontogenesis, enzymes in, 543-548 differentiation and, 545-548 Ontogeny of immunological prop- erties, 556-573 Oocyte, growth of, 177-180 nucleoprotein absorption of, 179-180 nurse cells of, 179-180 source yolk constituents of, 177— 179 Optic centers, 380 Optic chiasma, 360 Optic nerve, 362 aberrant, 383 fibers of, 383, 385 deflection of, 360 Organ, supernumerary, 381 Organ fields im insects, 340, 341 Organ growth, self-regulation of, 113 Organ-forming area, 309 Organization, colloidal state of, 44 fibrous pattern of, 59-60 of gastrula, 238-243 of insect egg, 337 molecular level of, 27 new approach to, 30 problem of, 29, 30 segmental, 378 Organizer, 243-247, 279-282 extent of, 246 field properties of, 279-281 in regeneration, 668 regional differences of, 245-247, 250-252 Organogenesis of blood vessels, 450452 : of ear, 415-423 viel Organogenesis of entodermal de- rivatives, 453-458 of eye, 402-414: of heart, 440-450 of limb and girdle, 429-439 of nervous system, 346-401 of nose, 423-424 of skin and derivatives, 499-519 of teeth, 492-498 of urinogenital system, 462-491 Orientation, 359 fibrillar, 354: mechanism of, 351—356 of nerve, 351, 353 Otocyst, determination of, 258 induction of, 308 Oviduct, origin of, 469 Oxygen in development, 522-528 Pancreas. See Liver-pancreas. Pangenesis, early theories of, 2 Parasympathetic neurons, 382 Parathyroid glands, 611-614 bone development and, 613 chief cell origin in, 611 fetal regulation of calcium level and, 612 indices of functional activity of, 611, 612 Parthenogenesis (artificial), 201, 202, 203 activation of, 202-203 chromosome numbers in, 201 of non-nucleated egg frag- ments, 201 origin of cleavage amphiaster in, 201 sex and, 202 Parthenogenetic merogony, 217, 218 Pathway structures, 362 Pathway systems, 354, 356, 359, 360 Pathways, nerve, 356, 358, 360, 361 preneural, 361-362 specificity of, 361 Patterns, nerve, 359 structural, 354 Pectoral girdle, chondrogenesis of, 434: limb disc removal on, 434 localization of parts of, 434, 435 mosaic character of, 434, 435 Pelvic girdle, 435-436 Perikaryon, 346, 347, 349 Peripheral innervation, density of, 367 Pharynx, 362 Phase-specific changes, 388 Phenocopies of hereditary abnor- malities, 702-703, 706 Phenocritical periods in _terato- genesis, 702 T32 Pigment cells, 372 differential sensitivity to hor- mones, 513, 514, 515 genotypic constitution of, 513, 514 “infective” transformation of, 515 invasion into feather germ, 504: migration into skin, 503, 504 neural crest origin of, 502, 5035513 rhythmic pattern formation and, 515 tissue environment and ex- pression of, 514-515 Pigmentation pattern, feather papilla locus and specificity of, 514 formation by pigment cells, 513-514 genotype of pigment cells and, 513, 514 reactions between pigment cells in, 515 sex-hormone effects on, 514— 515 tissue environment effects on, 514-515 Pioneering nerve fibers, 359, 362, 382 Pituitary in regeneration, 686 in teratogenesis, 710 See also Anterior pituitary. Pituitary-adrenal system, adren- aline in, 590-593 Placenta, insulin transportation and, 607 parathyroid hormone and, 612, 613 transfer of antibodies by, 564 transfer of erythrocytes by, 566 Planaria, head as organizer in re- generation of, 668 nervous system as organizer in regeneration of, 668 regeneration of, 664, 665, 668 Plasma, marginal, 230, 233, 234 Plexus formation, 346, 359-360, 384 Polarity in annelid regeneration, 665 in ascidian regeneration, 665 in ‘planarian regeneration, 664 of sea urchin egg, 665, 666 in Tubularia regeneration, 665 Polarizing factors, 350 Polyploidy and development, 134— 137 Polyspermy, block to, 194-195 Preformation, 3, 6, 7 influence of microscope on, 6, 7 Malpighi and, 6, 7, 8 Weismann’s contribution to, 19 Preganglionic autonomic cells, 378 Primitive streak, blastopore and, 301 de-epithelization of, 304 Hensen’s node and, 302 induction capacity and, 306 mesoderm formation and, 302, 304 Primordial germ cells, 175, 176, 177, 471-473 bipotentiality of, 474 crescent of, 471, 472 gonad formation and, 473 migration of, 472 source of, 472 See also Germ cells. Progressive differentiation in am- phibians, 230-296 in annelids and mollusks, 318-321 in birds and teleosts, 297-314 in ctenophores, 321 in echinoderms, 327-331 in insects, 337-345 in selected invertebrates, 315— 336 in teleosts and birds, 297- 314: in tunicates, 321-323 Proliferation, 353, 354, 358, 381, 383, 389 rate of, 378, 381 regional patterns of, 374-375 Proliferative patterns, 373 Pronephric area, position of, 463 Pronephric duct, 362 Pronephros, differentiation _ of, 4.65466 dorsoventral axis of, 465 Proprioceptive fibers, 385 Prospective significance of gas- trula areas, 235 of neurula areas, 247 of tail area, 249 Prostate, castration effects on, 483 reaction to sex hormones, 481 Protein as energy source, 533, 534, 535, 536, 537 Protein fibers, chemical identifi- cation of, 50-51 classification of, 47 configuration of, 46 hierarchy of fiber sizes, 47— 49 repeat period in, 47 Prothoracic gland, 638 hormone of, 644 action on tissues, 641 Protoplasm, colloidal organiza- tion of, 41 paracrystalline state of, 53-54 particulate systems of, 42 sol-gel transformations of, 54— 55 INDEX Protoplasm, tactoids and long range forces in, 53-54 water and dissolved substances of, 42 Pupation, hormonal control of, 639, 640, 642-644 Rassit, brachydactyly in, 709 Rat, chemical teratogenesis in, 702 gray-lethal mutant in, 703- 704, 708 vitamin deficiency as terato- genic agent in, 701, 712 x-rays as teratogenic agent in, 701, 715 Reaction potency. See Compe- tence. Receptor, 351 Regeneration, amino acids in, 672 in amphibians, 677-686, 690-— 693 in Amphioxus, 675 in annelids, 665, 668 in ascidians, 664, 665, 666, 667, 668, 671, 672 in birds, 689, 690 blastema of, 369, 679, 680, 681- 683, 692 of bone, 690 compensatory, 828, 674-688 competition hypothesis of, 669- 671 control by inhibition, 670 of cornea, 512, 689. in cyclostomes, 675 dedifferentiation in, 667, 679, 680 definition of, 674 of ear, 415 electric potentials in, 668 endocrines as controlling factor in, 685, 686 environment as controlling fac- tor in, 671 of epidermal ridges, 512 of epidermis, 512, 689 of eye, 409-412, 690, 693 of feather, 689 fields, 677, 678 of glands, 690 growth of organs (or parts) and, 627 of head structures, 685 induction in, 667, 668, 678, 693 inhibition of, 669, 670, 671 in insects, 341 of intestine, 690 in invertebrates, 664-673 of lens, 410-412, 690-693 of limb, 677-687 limits of, 627-629 of liver, 456 in mammals, 689-690 of muscle, 690 INDEX Regeneration of nasal placode, 423 organizers in, 668 physiological, 674, 688-690 polarity in, 664: population density as factor in, 671 potency of cells in, 667, 679- 685 reconstitutive, 674 in reptiles, 686-688 reserve cells in, 667, 680 respiratory rate in, 672 salt balance as factor in, 672 of scales, 688 in sponges, 668 stage of development as factor in, 676, 678 stimulus for, 666, 667 of tail, 675, 686-688 in Tubularia, 664, 665, 666, 668, 669, 670, 671 in vertebrates, 674-698 Regional host influence, 259-263 Regional specificity, competence of ectoderm, 258-259 host influence on induction, 259-263, 264 inductivity of adult tissues, 265, 269 inductivity of archenteron roof, 250-252, 254-255 organizer, 245-247 Regulation, 379, 383, 388 in embryos in amphibian meta- morphosis, 677 in insect eggs, 337, 338, 340 Relative growth, 625-627 concept of transformations in, 625, 627 power equation in, 625-627 Repertory of behavioral perform- ances, 392 Reptiles, limb-girdle development in, 436-437 regeneration of epidermis in, 689 of limb in, 686, 687 of scale in, 687 of tail in, 686-688 Resonance, 383-386 Resorption of cells, 375 Respiration in absence of cleavage, 528, 529 of amphibian embryos, 526-528 in chick development, 523, 524, 525 cyanide-insensitive, 541-542 in development, 523-533 during diapause of insects, 530, 531 enzyme synthesis, 543-545 during gastrulation, 527-528 growth and, 525 in hybrid development, 530 Respiration in maintenance and development, 531, 532, 533 mechanisms of, 538-543 citric acid cycle in, 539, 540, 541 in echinoderm eggs, 539-541 Embden-Meyerhof scheme in, 538, 539, 540, 541 schematic representation of, 539 in various species of embryos, 541 Warburg-Keilin 539 in metabolism and cell division, 528-530 in “no-X” insect eggs, 530 rate of, 524, 525, 526 in sea urchin development, 524, 529 Respiratory quotient, 535 Retina, 360, 385 determination of, 407, 408 functional localization in, 403 from pigment layer, 410 regeneration of, 410 Retinal cup, choroid fissure of, 403 determination of, 402-404 environmental influences on, 403 fixation of polarity of, 403 regulatory capacity of, 402, 403 Rhythmic automatisms of nerve centers, 392 Ring gland, 342, 638 function of, 643, 644 Roentgen irradiation. See X-ir- radiation. Rumpless fowl, 700, 701, 702, 703, 704, 710, 711, 712, 714 degeneration processes in, 702, 710 embryology of, 702, 703, 710, 714 phenocopies in, 703 seasonal changes of incidence of, 711 sensitive periods for insulin- produced, 713 system in, 539,097 Sarcoma, 370, 381 “Saturation,” 366, 367-370 “Scar,” 350, 352, 356, 360, 367 Schwann cells, 347, 350, 356, 358, 366, 367, 372 Schwann cords, 349, 362 Sea urchin egg, animal half re- sponse of, 328 animalizing influence on, 327-331 developmental modifica- tion of, 327-331 133 Sea urchin egg, dispermy in, 139, 140, 151 gradients in, 665, 666 double system of, 330 single system of, 330- 331 respiration of, 524, 529 self-differentiation in, 328, 329 vegetalizing influence on, 327-331 Selective contact affinities, 362 “Selective fasciculation,” 366, 380, 384 Selectivity, 356, 363, 366 Self-differentiation, 378, 387 Self-organization, 281, 304, 305 in brain-eye field, 253, 254 in chorda-mesoderm field, 240, 243-244, 281 in explants, 261, 273, 282 role of gradients in, 284 Sensitive periods in teratogenesis, 702, 705, 712-714, 715 Sensory nervous system, suppres- sion of, 392 Sensory neurons, 385 qualitative diversity among, 380 Sex differentiation, dominant gonad component in, 475 environmental agencies in 475 hormonal control of, 475— 477 patterns of, 483-484 hormonal theory of, 483 monhormonic theory of, 484 Sex ducts, castration effects on, 4824.83 hormone effects on, 480-481 Sex hormones, effects on cloaca and urinogenital — sinus, 4814.82 embryonic vs. adult, 485-486 inductor substances and, 486 mode of action of, 484485 “paradoxical” effects of, 476, 481 specificity of action of, 485 time of production and libera- tion of, 593-595 Sex reversal in freemartin, 475, 477 by grafting, 476, 478, 479 by hormone administration, 476-477, 479-482 Sex structures, accessory, 595, 596, 597, 598. See also Genital structures, accessory. Sexual behavior, hormonal effects on, 389 Sheath cells. See Schwann cells. Situs inversus of gut and heart, 457, 458 ¢} 734 Size of body, 128-129 of cells, 620-621 determination of, 620-630 of eye, 407, 408, 409, 628 Skin, 362, 363, 368, 385 beginnings as integrated sys- tem, 499 derivatives and, 499-519 epidermal ridges of, 510-512 functional differences of, 505- 506 morphological patterns of, 512- 513 origin of, 499-504: pigment cell migration into, 503-504 pigmentation patterns of, 513- 515 regional differences established early, 506, 507 regional hormonal response of, 505, 506 regional specialization of, 504— 512 source material of, 499-501 source of pigment cells in, 502— 503 structural differences of, 504 505 See also Epidermis. Skin ectoderm, local induction of, 4.99 polar organization of, 501- 506 primordial capacity of, 500, 501 specialization of, 501, 506 Specificity, 361-362, 384, 386 in regeneration, 362 Sperm, agglutination of, 182, 185, 186, 187, 188, 189, 190 antifertilizin of, 189-191 lytic agents of, 192-193 Spermatozoa, senescence of, 182~ 184 approach to egg, 181-182 Spina bifida, 372 Spinal action systems, 387 Spinal cord, 361, 362, 363, 367, 370, 374, 375, 378, 379, 380, 381-383, 387, 388, 392 central canal of, 371, 373 longitudinal pathways in, 360 regional differences in, 382 serial functional localization in, 387 Spinal ganglion, 362, 372, 377, 378, 380, 381, 383 Sponges, dissociation experiments in, 668 ionic balance as factor in regen- eration of, 672 Spongioblasts, 379 Stomach, 456 Stresses in medullary plate, 359 Stretch of longitudinal fiber tract, 359 of neural tube, 373 Stump. See Nerve stump. Substrata, 348, 352 orientation of, 352, 354, 359 Surface of axon, 361, 368 of neural plate, 369, 370, 373 contraction of, 370 Surface tension, cleavage and, 218 “Symbiotic” interdependence, 388 Symmetry, bilateral, 230, 231, 284 Sympathetic fibers, 370 Sympathetic ganglia, 372, 381 Sympathetic nervous system, adrenal medulla and, 592, 593 Sympathetic neurons, 382 Synaptic connections, 363 Synthesis of respiratory enzymes, 543-545 Tai, hereditary abnormalities of, 700, 701, 702 phenocopies of, 703 regeneration of, in cyclostomes, 675 in reptiles, 686, 687, 688 in teleosts, 675, 676 Tail bud, differentiation of, 249- 250 Teeth. See Tooth. Teleosts, cleavage in, 298-300 endocrines in regeneration of, 685-686 regeneration of nervous system, 676, 677 related to stages, 676 of tail, 675, 676 sensitive periods for x-ray in- duced abnormalities in, 713 twinning in, 711, 712, 713 Teleosts and birds, axis formation in, 299-304 cleavage period in, 298-300 gastrulation in, 300-306 pre-cleavage period in, 297-— 298 progressive differentiation in, 297-314 Temperature as factor in inverte- brate regeneration, 671 growth and, 628, 629 as teratogenetic agent, 701, 702 Tension, 352, 353 Teratogenesis, 237-238, 242-243, 699-719 abnormal initial stimulus in, 706, 707, 708 abnormal responsiveness of tis- sues in, 707, 708 abnormal tissue differentiation in, 708 carbohydrate metabolism in, 705, 706 INDEX Teratogenesis, causal agents of, 700-703 arrest of development, 703 chemicals as, 701, 702, 704, (055708712 713. 714. 715 dietary factors as, 701, 704, 705 environmental factors as, 701-702 gametes, physiological con- dition, 711 hereditary factors as, 700- (AO, 7slal, 7/52 insulin, 703, 705, 712, 713, 75 irradiation, 701, 709, 710 overripeness of egg, 709, 711 physiological as, 710 pituitary deficiency, 710 retardation of develop- ment, 704, 705 rubella, 702 storage condition of chick egg, 711 virus, 702, 714 vitamin deficiency, 705, 712 x-rays, 701, 709, 710, 715 classification of terata, 706 correlations in, 703, 706 degeneration (cellular) in, 702, 709, 710 dosage effects on, 715 environmental factors in, 701, 702 erythroblastosis and, 702 hereditary factors in, 700, 701, (lil, 7A persistence of normally degen- erating structures in, 710 phenocopies in, 702, 703 recovery from teratogenic agents, 715 responsiveness of tissues in, 707, 708 sensitive periods in, 702, 705, 712-714, 715 specificity of chemical agents in, 714-715 supernumerary organs in, 707 Terminology, difficulties of, 29 Roux and, 29 Terni, nucleus of, 378 Thyroid in amphibian metamor- phosis, 631-635 criteria of activity of, 633 initial response to thyrotrophic hormone, 582 iodine concentration in, 634 onset of hormonal activity of, 575, 576 deficiencies 701, INDEX Thyroid, progressive increase of iodine compounds in, 575, 581 reciprocal action on anterior pituitary, 584, 585 selective affinity for 575 thiouracil inhibition of, 633 vascularity of, 581 Thyroid hormone, action on neu- ral elements, 389 Tissue culture, 352, 356, 358, 360, 367, 381 axon outgrowth in, 350 Tissue interaction, 113 Tooth, abnormalities in, 495, 4.96, 497 appositional calcification of, 495 appositional growth of, 495-— 497 formative span of, 496-497 potential for, 496 rates and gradients of, 496 attrition of, 497 calcification of, 497 components of, 492, 494 developmental stages of, 492, 4.93 eruption of, 497 germ of, 494 growth of, 492-497 growth centers of, 496 incremental cones of, 496 initiation of, 492, 494 life cycle of, 492 morphodifferentiation in, 495 organogenesis of, 492-498 Tooth germ, 494 enamel organs of, 494: initiation of, 455 morphodifferentiation in, 495 odontoblasts of, 494 “Towing,” 351, 356, 361 Transcendentalism, 9, 14, 17 Transneuronal effects, 383 iodine, Transplantation of cornea, 385 of eyes, 358 heterotopic, 361 of limb buds, 358, 359, 361 of M-cell region, 379 of nasal placodes, 358 xenoplastic, 263-265 Tubularia regeneration, competi- tion hypothesis in, 670- 671 correlation in, 668 electrical potentials in, 668- 669 inhibition of, 666, 667, 668, 669, 670 polarity in, 665 respiratory rate in, 672 stimulus for, 666 temperature as 671 Tumor agent, 381 Tumor transplantation, 381 Tumors, 708, 709 innervation of, 362 Tunicates, progressive differenti- ation in, 321-323 Turgor, 371, 372, 374 Twinning, 127 ef amphibian egg, 127-128 in annelids, 325 in bird blastoderm, 306 1m insects, 338 of kidney, 469 in sea urchin, 327 in teleosts embryos, 304 species differences in, 711, (als “Two-center effect,” 354, 358, 359 Type specificity, 380 factor in, URINOGENITAL sinus and cloaca, reactions to sex hormones, 481— 482 Urinogenital system, organogene- sis of, 462-491 T35 VASCULARIZATION of neural tube, 375 Vegetalization of entire eggs, 329 graded intensity in, 329 hypotheses to account for, 330- 331 im sea urchins, 327-331 time required for, 328 in vegetative halves, 327, 328, 329 Virus as teratogenic agent, 702, 714 Visceral asymmetry. See Situs in- versus. Vital staming, 234 value of, 32 Vitamin deficiencies as terato- genic agents, 701, 712 WALLERIAN degeneration, 349 Wing bud, essentiality of ecto- derm for, 437, 438 fate maps of, 437 sequential order of parts of, 437 Wingless mutant, 706, 707 “Wink” reflex, 389 Wolffian lens regeneration, 410- 412, 690-692 Wound healing, epithelium in, 689-690 XENOPLASTIC inductions, 246, 280 Xenoplastic transplantations, 263— 265 X-irradiation, effect on developing CNS, 373 in teratogenesis, 715 sensitive periods to, 713 as teratogenic agent, 701, 709, 710 treatment of gametes by, 711 Youk utilization im development, 527 ve se a a mir ree a Pt me Be , a baba brs th i es eee te ey PASSA ‘Sy As bacnies elu fu: RoR sree es . ~ s - 7 . : me - a aS res = Le or 4 Pam 2 : posit » q : Boi a! ‘ . i ere ’ a ~ ‘ eae on sh od oo 3 e ff Sn , = € Se wee # ahs “ So t : " ieee, y Bg oad bs 9 P ” 1853 5 ' ¥ See i (ones STR % se . Bessey : ‘ fi ? A sR noe S a & 4 BS ot ‘ 4 . & : fs Bi » an fe 7 ne af > 6 ‘ . nis fons som > cy Vase i~ < = ie = — ne ee S . od is " ey songt . a ge ocd: ae ae ‘—— * Lhe ines “ » ‘ ’ 5 : os . 4 df ~ Y ae a ef ; eas a x 3 : y & a ae. e si Lowey os: ; ; : at fae ey 13 x be worn es) = - . ” “ ong Po x oar