CD- =O . CT r-^ D nj D m D MONOGRAPHS ON EXPERIMENTAL BIOLOGY . EDITED BY JACQUES LOEB, Rockefeller Institute T. H. MORGAN, Columbia University W. J. V. OSTERHOUT, Harvard University THE ELEMENTARY NERVOUS SYSTEM BY G. H. PARKER, Sc.D. MONOGRAPHS ON EXPERIMENTAL BIOLOGY PUBLISHED FORCED MOVEMENTS, TROPISMS, AND ANIMAL CONDUCT By JACQUES LOEB. Rockefeller Institute THE KLI Ml NTARY NERVOUS SYSTEM By G. H. PARKER. Harvard University IN PREPARATION THE CHROMOSOME THEORY OF HEREDITY By T. H. MORGAN. Columbia University INBREEDING AND OUTBREEDING: THEIR GENETIC AND SOCIOLOGICAL SIGNIFICANCE By E. M. EAST and D. F. JONES. Bussey Institution. Harvard University PURE LINE INHERITANCE By H. S. JENNINGS, Johns Hopkins University THE EXPERIMENTAL M< >I)IFICATION OF THE PROCESS OF INHERITANCE By R. PEARL, Johns Hopkins University LOCALIZATION OF MORPHOGENETIC SUBSTANCES IN THE EGG By E. G. CONKLIN, Princeton University TISSUE CULTURE By R. G. HARRISON. Yale University PERMEABILITY AND ELECTRICAL CONDUCTIVITY OF LIVING TISSUE By W. J. V. OSTERHOUT, Harvard University THE EQUILIBRIUM BK'l \\ I I X ACIDS AND BASES IN ORGANISM AND I NVIRONMENT By L. J. HENDERSON. Harvard University UIEMICA!. i:\SIS OF GROWTH By T. B. ROBERTSON, University of Toronto COOUDIN \TION IN LOCOMOTION liy A. R. MOORE. RutKi-rs College 'I III- NATURE ()!• ANIMAL LIFE Hv 1C. N. HAKVKY. Princeton University } OTHERS WILL FOLLOW MONOGRAPHS ON EXPERIMENTAL BIOLOGY THE ELEMENTARY NERVOUS SYSTEM BY G. H. PARKER, Sc.D. PROFESSOR OF ZOOLOGY, HARVARD UNIVERSITY 53 ILLUSTRATIONS PHILADELPHIA AND LONDON J. B. LIPPINCOTT COMPANY COPYRIGHT. IQ19. BV J. B. L1PPINCOTT COMPANY r.leclrotyped and Printed by J. B. L\p{nncott Company Tkt Washington Square Prea. Philadelphia. i/.i.X. EDITORS' ANNOUNCEMENT THE rapidly increasing specialization makes it im- possible for one author to cover satisfactorily the whole field of modern Biology. This situation, which exists in all the sciences, has induced English authors to issue series of monographs in Biochemistry, Physiology, and Physics. A number of American biologists have decided to provide the same opportunity for the study of Experimental Biology. Biology, which not long ago was purely descriptive and speculative, has begun to adopt the methods of the exact sciences, recognizing that for permanent progress not only experiments are required but that the experi- ments should be of a quantitative character. It will be the purpose of this series of monographs to emphasize and further as much as possible this development of Biology. Experimental Biology and General Physiology are one and the same science, by method as well as by contents, since both aim at explaining life from the physico-chemical constitution of living matter. The series of monographs on Experimental Biology will therefore include the field of traditional General Physiology. JACQUES LOEB, T. H. MORGAN, W. J. V. OSTERHOUT. 5 AUTHOR'S PREFACE THE dependence of human affairs upon the nervous system of man is so absolute that it was inevitable, as soon as this relation was understood, that the activities of the simpler animals should be interpreted as though these creatures were miniature human beings. That such interpretation was carried far beyond its legitimate bounds, even by the scientifically trained, is now admit- ted on almost all sides, but it is no easy or simple task to ascribe to this movement its proper bounds. That these bounds are vastly more restricted than has usually been supposed is certain. An approach to a clearer under- standing of what they are is assured through the appli- cation of experimental and quantitative methods to the questions concerned rather than by a continuation of the older more purely observational procedure. It is from this standpoint that an attempt has been made in this vol- ume to portray the elementary nervous system as it exists in the simpler animals and in the simpler parts of the more complex forms. It is believed that this treat- ment of the subject may help in the solution of the gen- eral problem by removing once and for all some of the old misunderstandings concerning the nervous system and by inviting the student's attention to new methods of attack. The subject matter of the volume is drawn almost en- tirely from the three simpler phyla of the multicellular animals, the sponges, the crelenterates, and the cteno- phores. This is intentionally done for the reason that the activities of such animals as the echinoderms, worms and 7 AUTHOR'S PREFACE so forth, will be treated in another monograph of this series, one by Dr. A. JJ. Moore. It is also to be remarked that in Chapter Xlll <>!' the present volume, which deals with the hydroids, the account of the activities of these animals is rather more detailed than that accorded to other forms. This method of treatment has been followed in consequence of the fact that the details of the re- sponses of hydroids are by no means so readily available in the literature as are those of other groups of ccelenterates. The writer is indebted to his wife for a full revision of the manuscript and to the editors of the series for many helpful suggestions. He is also under obligations to Mrs. Donald E. Woodbridge for the preparation of the illustrations. G. H. P. HARVARD UNIVERSITY, CAMBRIDGE, MASS. NOVEMBER, 1918. CONTENTS CHAPTER PAGE I. INTRODUCTION 13 SECTION I— EFFECTOR SYSTEMS II . SPONGES 25 III. SPONGES (Continued) 37 IV. INDEPENDENT EFFECTORS IN HIGHER ANIMALS 50 V. NEUROID TRANSMISSION IN HIGHER ANIMALS 64 SECTION II.— RECEPTOR-EFFECTOR SYSTEMS VI. THE NETJROMUSCULAR STRUCTURE OF SEA- ANEMONES 76 VII. NERVOUS TRANSMISSION IN SEA- ANEMONES 89 VIII. JELLYFISHES 102 IX. THE NERVE-NET 115 X. TRANSMISSION IN THE NERVE-NET 130 XI. APPROPRIATION OF FOOD AND THE NERVE-NET 138 XII. OTHER COMPLEX RESPONSES AND THE NERVE-NET 150 XIII. HYDROIDS 175 SECTION III.— CENTRAL NERVOUS ORGANS XIV. CONCLUSION . 199 LITERATURE 215 INDEX . . 225 9 ILLUSTRATIONS FIG. PAGE 1. Diagram of the Primary Sensory and Motor Neurones of the Spinal Cord of a Vertebrate 17 2. Diagram of the Primary Sensory and Motor Neurones of the Ventral Ganglionic Chain of the Earthworm 18 3. Ectodermic Layer from the Tentacle of a Sea-anemone 20 4. Diagram of the Canal System of a Calcareous Sponge 26 5. Radial Portion of a Transverse Section of Stylotella 30 6. Diagram of a Finger of Stylotella 32 7. Dermal Membrane Over a Subdermal Cavity in the Sponge Stylotella. 34 8. Three Dermal Pores in Stylotella 35 9. Fingers of Stylotella in a Strong Current of Seawater 40 10. A Finger of Stylotella Protected from a Gentle Current of Seawater . . 41 11. Diagram of the Apparatus for Measuring the Strength of a Current of Seawater Produced by Stylotella 44 12. Diagram of the Apparatus to Test the Strength of the Dermal Pores of Stylotella 47 13. Diagram of the Heart of a Fish 54 14. Diagrammatic View of a Longitudinal Section of Scalpa; Enlarged View of an Isolated Heart of Scalpa 57 15. Side View of the Ctenophore Mnemiopsis leidyi 67 16. Plan of a Temperature Box 69 17. Diagram of a Longitudinal Section of the Sea-anemone Metridium. . . 77 18. Transverse Section of the Sea-anemone Metridium 78 19. Diagram of the Ectoderm of a Sea-anemone 80 20. Diagram of a Partial Longitudinal Section of the Sea-anemone Metridium 85 21. Metridium With Its Column Completely Girdled by a Cut that Passes at all Points Entirely Through the Wall 91 22. Metridium in Whose Column Wall an Oblong Incision Has Been Made so that the Resulting Piece is Attached to the Animal Only by the Mesenteries 92 23. Metridium Cut Vertically in Two Except in the Region of the Lips . . 92 24. Metridium from Which an Equatorial Tongue Has Been Cut from the Column 93 25. Metridium from Which a Tongue Running from the Pedal Disc to the Equator of the Column Has Been Cut 94 26. Metridium from Which a Portion of the Pedal Edge of the Column Has Been Partly Cut Off 94 11 FIG. PAGE 27. Metridium Cut Through Vertically Except for Its Oral Disc 95 28. Metridium Cut Through Vertically Except for Its Aboral Disc 95 29. Melridium Cut Through Vertically Except for a Small Connecting Bridge on the Pedal Edge of the Column 96 30. Section at Right Angles to the Sphincter of the Bell of Rhizostoma. . 105 31. Diagram of the Jellyfish Aurelia from Which the Central Mass and Seven of the Eight Marginal Bodies Have Been Removed and the Outline of the Bell Further Complicated by a System of Interdigitated Cuts 107 32. Diagram of the Jellyfish Aurelia from Which the Central Mass and Seven of the Eight Marginal Bodies Have Been Removed and the Outline of the Bell Further Complicated by a Circular Incision 108 33. Diagram of a Portion of the Bell of Rhizostoma 110 34. Preparation of Two Muscle Fields from Rhizostoma Ill 35. Preparation of Two Muscle Fields from Rhizostoma, One With a Mar- ginal Bod}' and the Other Without Such a Body 113 36. Nerve Network from a Small Blood-vessel in the Palate of the Frog. 118 37. Diagram of the Nervous Organization in the Intestinal Wall of a Vertebrate 128 38. Diagram of the Reactions of a Tentacle of a Sea-anemone to Tran- section 131 39. Sense Cells With Their Attached Fibers in a Tentacle of the Sea- anemone Cerianthus 132 40. Diagram to Illustrate the Preparation of the Intestine in Mall's Experiment on Intestinal Polarization 134 41. Outline and Transverse Section of the Sea-anemone Meiridium 136 42. Diagram of a Jellyfish Aurelia Whose Bell Has Been Deeply Incised in a Radial Direction at Eight Places 166 43. Side View of the Hydroid Corymorpha 176 44. A Polyp of Corymorpha Split Lengthwise Except at Its Aboral End . 188 45. Diagram of an Independent Effector 200 46. Diagram of a Simple Type of Receptor-effector System 201 47. Diagram of a Complex Type of Receptor-effector System 202 48. Transverse Section of the Ventral Nerve-chain of the Marine Worm Sigalion 203 49. Transverse Section of the Ventral Nerve-chain of the Earthworm Allolobophora 205 50. Diagram of a Transverse Section of the Spinal Cord of a Vertebrate (Salamander) 206 51. Stages in the Differentiation of Sense Cells 210 52. Stages in the Differentiation of Nerve Cells 211 53. The Appropriation of Secondary Sensory Cells by Primary Sensory Neurones in the Vertebrates 213 12 THE ELEMENTARY NERVOUS SYSTEM CHAPTER I INTRODUCTION THE elementary nervous system is that type of nerv- ous system in which the structural and functional ele- ments present themselves in their simplest states. It is most extensively represented in the lower multicellular animals, but it is found locally in as highly differentiated forms as the vertebrates. It probably represents the primitive nervous system in that it reproduces much that must have been characteristic of nervous organs in their earliest stages of evolution. It has been an object of in- vestigation and interest only within comparatively recent years. Its derivative, that complex aggregate of organs known as the central nervous system of the higher ani- mals, attracted the attention of anatomists and physiol- ogists at a much earlier period. This was doubtless due to the unusual development of the central nervous organs in man and other higher animals. The human brain is a structure whose size and position naturally compels at- tention, whereas the chief nervous organs in many of the lower animals are often so insignificant as to be easily overlooked. Although the first attempts at elucidating the struc- ture and functions of the nervous system were made on 13 14 THE ELEMENTARY NERVOUS SYSTEM the higher animals, it was a long time before this system was recognized as the exclusive seat of the most striking characteristic of man, his mental life. Only gradually was it discovered that his conscious states represent the action of a single system of organs, the nervous system, as contrasted with the rest of his body. To the ancients this conscious life, whose chief at- tribute is personality, seemed to permeate the whole human frame. In fact, Aristotle, who was such an ac- curate observer and profound thinker in so many fields of biology, denied positively that the brain was in any direct way concerned with sensation and declared the heart to be the sensorium commune for the whole body. To Galen is ascribed the belief that the brain is the seat of the ra- tional soul, the heart the location of courage and fear, and the liver that of love. The distribution of the elements of personality over the physical body finds its expression in the common speech of to-day, particularly in relation to the heart, which is widely accepted by the popular mind as the seat of the more tender emotions. Although this opinion may commonly imply a certain amount of poetic license, it is quite certain that many an untrained person holds even at the present time to a literal interpretation of the ancient view of the location of sensations. The pain of a pin nrick is commonlv believed by many persons to be where the pin enters the skin. To thorn nothing seems more obvious and certain than that the punctured spot is the seat of the pain, and any attempt to change their views on this point will usually be regarded by them with suspicion and mistrust, for it seems contrary to common sense. Nevertheless, it is well known that if a nerve distrib- uted to a given area of skin is cut at some distance from INTRODUCTION 15 that area, the spot, though unaffected by the operation in any direct way, will give rise to no further sensations even when it is severely injured. Hence, it is clear that the pain does not reside simply in the skin. But not only may pain thus be absent from a given area of skin ; it may be present when the skin with which it is supposed to be associated is absent. Persons who by accident or other- wise have lost an arm or a leg often experience long after the loss vivid and intense sensations from definite parts of the missing member. So precise and sharp are these sensations and so certainly do they seem to be associated with the lost part that some of the less knowing of these unfortunates have attempted to exhume or otherwise get possession of the lost member in an endeavor to alleviate their unpleasant sensations. These misunderstandings, for such they are, can be swept away at once and the matter put in its true light when we recognize that our sensations are not located in the peripheral parts affected, but in the central nervous system, and within that portion of it known as the cere- bral cortex. As long as this organ is intact, sensations may arise, and, though these are usually due to nervous impulses from the sensory surfaces, they may be called forth by an internal stimulus as well. Thus it is that a missing arm may be represented by sensations years after it has been severed from the body. With a loss of an appropriate part of the cerebral cortex, however, comes a loss of sensation that is absolute and final. From this there is no recovery (Parker, 1916 b). This modern view of the relation of sensation to the nervous system was initiated by the anatomists and physiologists of the Renaissance. Thus, Vesalius taught, in the sixteenth century, that the chief soul was engen- 16 THE ELEMENTARY NERVOUS SYSTEM dered in the brain by virtue of the powers of the proper material and form of that organ. And, although Stahl attempted, a century and a half later, to revive the belief that the soul and the sensorium commune were diffused over the whole body, that is, were resident as well in the tip of the finger as in the brain, the idea of the localization of these active properties in the nervous system became so well grounded through the investigations of the physi- cians of that time, particularly Haller, as to assume the form of permanency. This growth of knowledge led di- rectly to the modern view that personality is, strictly speaking, a property of the nervous system and is in no true sense the direct result of any other system of organs. (Foster, 1901.) The nervous system, to be sure, is embedded among the other organs of the body, and the environment thus provided influences profoundly its condition and action; but what is meant by individual personality, acuteness or dulness of sense, quickness or slowness of action, tem- peramental traits, such as a gloomy or bright disposition, incapacity, shiftlessness, honesty, thriftiness, or sweet- ness, are all in the strictest sense functions of the nervous system. Hence, it is a matter of no small biological inter- est to ascertain from the conditions presented in the sim- pler animals which of these various states are elemental and what has been the probable line of evolution of that system of organs with which our personality is so indis- solubly connected (Parker, 1914 a}. The nervous system of the higher animals, though enormously complex in its organization, is composed of relatively simple cellular elements, the neurones, arranged upon a comparatively uniform plan. This plan is well exemplified in the spinal nerves and spinal cord of the INTBODUCTION 17 vertebrates (Fig. 1). In this complex the sensory neu- rones, whose cell-bodies lie in the dorsal ganglia, extend from the integument through the dorsal roots to the gray matter of the cord. Motor neurones, whose cell-bodies are situated within the gray matter of the cord, reach from this region to the muscle-fibers which they control. These two classes of neurones would seem to be sufficient for all ordinary reflex operations, but the cord contains within its limits other neurones which serve to connect Fio. 1. — Diagram of the primary sensory s and motor m neurones of the spinal cord of a vertebrate showing their connections with the integument i and with muscle fibers mf. one part of its structure with another. These neurones, which have been called internuncial neurones, are inter- polated between the sensory and motor elements just de- scribed and must thereby lengthen and extend the courses of the reflex impulses. Such neurones make up a large part of the substance of the cord and doubtless increase enormously its internal connections. In the brain they not only add to the nervous interrelations, but they af- ford in the region of the cerebral cortex the material basis for all intellectual operations. The plan of neuronic arrangement as exemplified in the vertebrates also obtains in animals as lowly organ- ized as the earthworm (Fig. 2). In this form the sensory 2 18 THE ELEMENTARY NERVOUS SYSTEM neurones, whose cell-bodies are situated in the integu- ment instead of being gathered into special ganglia, ex- tend, as in the vertebrates, from the skin to the central nervous organs, the brain or the ventral ganglionic chain. The motor neurones are essentially duplicates of those in the vertebrates in that their cell-bodies lie within the cen- tral organ whence their fibers extend to the appropriate FIG. 2. — Diagram of the primary sensory s and motor m neurones of the ventral gan- glionic chain of the earthworm showing their connections with the integument i and with the muscles ms. (Modified from Retzius.) musculature. Intemuncial neurones are also abundantly present in the earthworm, though their function here, in contrast with that in the higher vertebrates, is pure nerv- ous intercommunication, for it is very unlikely that the earthworm possesses what in any strict sense of the word can be called intelligence. Thus from a morphological standpoint, the nervous systems of the higher annuals, even including such forms as the earthworm, have much in common, their three sets of interrelated neurones, sensory, motor, and internuncial, being arranged upon what is in the main a uniform plan. Considered from a physiological standpoint, the nerv- ous system with its appended parts as just sketched falls in the higher animals into three well-marked categories. On the exterior of these animals are to be found sense or receptors such as the free-nerve terminations INTRODUCTION 19 of the sensory neurones in the vertebrates or the sensory cells in the integument of the earthworm. These organs have for their function the reception of the external stim- uli and the production of the sensory impulses. The re- ceptors are connected by nerve-fibers with the central nervous organ or adjuster composed of the central ends of the sensory and the motor neurones and of the inter- nuncial neurones. Here the impulses arriving from the receptors are directed toward the appropriate groups of muscles by which the animal may respond to the stimulus and, if the animal is highly organized, impressions are made upon the adjuster which, as memories, may become more or less permanent parts of the animal's nervous equipment. Finally the adjusters are connected by nerve- fibers with the third set of elements, the effectors, which as muscles, electric organs, glands, etc., enable the animal to react on the environment. Thus three physiological categories are to be distinguished which in the order of their sequence in action are sense organs or receptors, central nervous organs or adjusters, and muscles or other effectors. It is to be noted in passing, that the physiological scheme just outlined includes a wider range of parts than is generally admitted under the head of the nervous sys- tem. The additional parts are the effectors, which, as will be shown later, form as truly a part of the whole sys- tem as do the sense organs or the central nervous organs. Since the term nervous system does not ordinarily include the effectors, it is perhaps best to designate the whole chain of related parts, receptors, adjusters, and effectors, as the neuromuscular mechanism, and in dealing with the elementary nervous system it will be found important to keep this relation in mind, for in such an inquiry, the 20 THE ELEMENTARY NERVOUS SYSTEM V.rr-'^-.^ };(#fr;.v.A real question that must be confronted concerns the inter- relations within the neuromuscular mechanism rather than those that are simply within the nervous system itself. The type of neuromuscular mechanism described in the preceding paragraphs in which a group of receptors is connected with a well centralized ad- justor that in turn controls a complex sys- tem of effectors, is found only in the more differentiated animals. Certainly in the simple forms, like the jellyfishes, corals, sea-anemones and so forth, only the slight- est evidence of this type of nervous organ- ization can be discovered. Nevertheless, these animals possess a neuromuscular mechanism, but on so simple a plan that in- vestigators have. long been inclined to re- gard it as representing the first step in the differentiation of neuromuscular organs. This plan of structure is well represented in the sea-anemones. Each of the two lay- ers of cells that make up the living sub- stance of the sea-anemone's body consists ordinarily of three sublayers ; a superficial or epithelial layer, a middle or nervous layer, and a deep or muscular layer (Fig. 3). The epithelial layer contains, besides many other kinds of cells, large numbers of sensory cells which terminate peripherally in bristle-like receptive ends and centrally in fine nervous branches. These fine branches constitute collectively the middle or nervous layer in which occasionally large branching cells, the so- called ganglionic cells, occur. Immediately under the nervous layer is the deep layer of elongated muscle-cells. Fio. 3. — E c t o- dermio layer from the tentacle of a sea-anemone show- ing the three sub- layers, epithelial «, nervous n, and muscular m. INTEODUCTION 21 The condition thus briefly described is present over much of the sea-anemone's body, and though the nervous layer may be somewhat emphasized in some regions, it cannot be said to be really centralized in any part. Hence this type of nervous system has been designated as diffuse in contrast with the centralized type found in the higher animals. What is really present in the neuromuscular portion of the sea-anemone's body is a large number of peripheral sensory cells whose deep branching ends connect more or less directly with the muscles, that is, without the inter- vention of a true central organ. This neuromuscular sys- tem, if described in the terms already used, could be said to be composed of receptors and effectors without an ad- justor or at least with this member present in only a most undeveloped state. Hence the adjuster or central organ is in all probability an acquisition that represents a later stage in the evolution of the neuromuscular mechanism than that seen in the crelenterates. If the coelenterates represent a stage in the develop- ment of- the neuromuscular mechanism in which sensory cells and muscles are the only important parts present, it is natural to ask if there is not a still more primitive state from which the ccelenterate condition has arisen. On this question several hypotheses have already been ad- vanced. Glaus (1878) and, subsequently, Chun (1880) maintained that originally the nervous system and the muscles were differentiated independently and that they became associated only secondarily. This view has de- servedly received very little attention, for it is extremely difficult to conceive of an animal that would develop re- ceptive ability without at the same time acquiring the power to react. Such an animal would have a certain 22 THE ELEMENTARY NERVOUS SYSTEM resemblance to a person suffering from complete motor paralysis who might still be able to receive impressions from the exterior and even to reflect on them, and yet would be incapable, in consequence of the complete sep- aration of nerve and muscle, of carrying out activities that would maintain a harmonious adjustment with the exterior. Not only does this view of the independent origin of nerve and muscle meet with the inherent and serious difficulty just alluded to, but among the lower forms not a single animal is known in which nerve is un- associated with muscle. Hence the hypothesis of Claus and of Chun has received very little serious consideration. Much less subject to criticism than the hypothesis of the independent origin of nerve and muscle is Kleinen- berg's theory of the neuromuscular cell. In 1872 Klein- enberg announced the discovery in the fresh-water hydra of what he designated as neuromuscular cells. The per- ipheral ends of these cells were situated on the exposed surface of the epithelium, of wrhich they wrere a part and were believed to act as nervous receptors ; the deep ends were drawn out into muscular processes and served as effectors to which transmission was supposed to be ac- complished through the bodies of the cells. Each such cell was regarded as a complete and independent neuromus- cular mechanism, and the movements of an animal pro- vided with these cells was believed to depend upon the simultaneous stimulation of many such elements. It was Kleinenberg's opinion that these neuromuscular cells di- vided and thus gave rise to the nerve-cells and muscle- cells of the higher animals. In fact, In- declared that the nervous and muscular systems of these animals wrere thus to be traced back to the single type of cell, the neuromus- cular cell, which morphologically and physiologically rep- INTRODUCTION 23 resented the beginnings of both. But Kleinenberg 's neu- romuscular cells were subsequently shown by the Hert- wigs to be merely epitheliomuscular cells and no inter- mediate stage between them and the differentiated neuro- muscular mechanism of higher forms was ever discov- ered. Hence this hypothesis, too, has been largely aban- doned (Parker, 1911). Some years later, in 1878, the Hertwigs published an account of the neuromuscular mechanism in co?lenterates. In this account they described the sensory cells, the gan- glionic cells, and the muscular cells of the ccelenterates, and maintained that these elements arose not by the di- vision of single cells, as stated by Kleinenberg, but that each element was differentiated from a separate epithe- lial cell and yet in such a way that during differentiation all these elements were physiologically interdependent. This hypothesis of the simultaneous differentiation of nerve and muscle, which has been the current opinion among biologists for more than a generation, is not with- out its serious difficulties, for it appears that in the sponges, which are more primitive animals than the ccelenterates, there are muscles of a very simple type but without any associated nerve (Parker, 1910 a). In fact, no nervous tissue of any kind has been definitely identified in sponges. It, therefore, appears that of the two ele- ments, nerve and muscle, the latter may exist indepen- dently of the former and in such a way as to indicate its more primitive character. From this standpoint it seems that the receptor-effector system of the ccelenter- ates was preceded by a simpler state in which only the effector element, muscle, was present and that this element may, therefore, be regarded as the original one in the evolution of the neuromuscular mechanism. Muscle once 24 THE ELEMENTARY NERVOUS SYSTEM developed as an effector gave occasion for the addition of receptors or sensory elements after which the adjuster or central nervous mechanism was differentiated. Hence in dealing with the elementary nervous system it will not be inappropriate to consider first such independent effec- tors as are found in sponges after which the receptor- effector system of the coelenterates may be dealt with, in part as a system in itself and in part as the source of the differentiated receptor-adjustor-effector systems of the more complex invertebrates and the vertebrates. Heretofore the approach to this subject has been chiefly from the side of anatomy and histology; but important as this line of research has been, the more recent ex- perimental results have shown that the physiological side of the question is quite as illuminating as the morpholog- ical. In the following pages some of the more important problems of the elementary nervous system will be dis- cussed. While these problems will be taken up chiefly from a functional standpoint, the structural aspect of the questions concerned will not be omitted, for in discus- sions of this kind the double viewpoint is more likely to lead to sound conclusions than if only one aspect is kept in sight (Parker, 1909, 1910 6, 1914 &). SECTION I. EFFECTOR SYSTEMS CHAPTER II SPONGES SPONGES are such inert organisms that their member- ship in the animal series was for a long time unsuspected. Their methods of development, however, settled this question beyond dispute. In form they are either single, more or less goblet-shaped animals (Fig. 4) or somewhat amorphous colonies whose chief activity is exhibited in the very considerable currents that they produce in the surrounding water. These currents enter the substance of the sponge through innumerable pores scattered over the surface of its body, pass through its more solid parts by a system of canals that converge on a central cavity, the cloaca, from which they emerge by a large opening, the osculum. This opening is situated at the apex of the individual sponge or, in the case of a colony, each osculum is ordinarily on a somewhat conical elevation rising above the general surface of the colony. The currents that emerge from sponges are often so considerable as to deform the surface of the water above them much as is done by a vigorous spring. They are produced by the action of the choanocytes or flagellated cells that line the canals in the body of the sponge or that invest more specialized chambers interpolated on the course of these canals. So far as can be judged by the examination of the inner parts of sponges, the choano- cytes are incessantly active, and the obliteration and re- vival of currents as seen in many of these animals are 25 26 THE ELEMENTARY NERVOUS SYSTEM due not to changes in the activity of the choanocytes but to other causes. Some sponges, such as the fingered form Stylotella, appear, when out of water, to be more or less shrivelled or contracted and under other circumstances to be plump and well rounded-out. The differences, which, for rea- sons to be mentioned pres- ently, are known not to be due to the simple physical loss of fluid, are apparently depen- dent upon a general contractil- ity of the whole flesh of the sponge which, though slight, may nevertheless enable the sponge to change its form somewhat. Aristotle in the fourteenth chapter of his fifth book on the history of animals makes the interesting state- ment that the sponge is sup- posed to possess sensation because it contracts if it per- ceives anv movement to tear • it up and it does the same when the winds and waves are so violent that they might loosen it from its attachment. He further adds in his characteristic way that the natives of Torona dispute this. The idea that the common flesh of the sponge is con- tractile is not without modem support. Merejkowsky (1878) stated that if the sponge Suberites is so placed that it is partly out of water, it will curve the body until it is under water as much as possible, and if the body is then Fio. 4. — Diagram of the canal sys- tem of a calcareous sponge (modified Haeckel). The innumerable superfi- cial pores receive water from the exte- rior, as shown by the arrows on the sides; the osculum at the apex dis- charges water to the exterior. SPONGES 27 covered with water, it will return to its former position. It must be evident, from what has already been stated, that much of the common flesh of Stylotella is contractile. As already noted, specimens out of water quickly assume a shrivelled and rugose appearance as though the flesh had contracted on a resistant skeleton, a condition that it also quickly assumes in quiet seawater. Moreover, if a sponge is placed partly in running seawater and partly in the air, the portion in the seawater remains smooth and that in the air becomes rugose. Specimens made rugose either in the air or in quiet water soon recover their smooth ap- pearance on being placed in running seawater. Air or quiet water may then cause a contraction of the common flesh of Stylotella, a condition counteracted by running water. The contraction of the common flesh can also be well seen around some of the larger cavities, such as the clo- acal cavity. If a long finger of Stylotella, whose two ends have been cut off and whose cloacal cavity extends along one of its sides, is placed in quiet seawater, the cloacal cavity is soon indicated by an external groove due appar- ently to the partial collapse of the cavity. This groove, however, is caused not by collapse, but by the contraction of the common flesh which, as partial partitions or even trabeculaB is abundant about the sides of the cloaca. On returning the finger to running water the flesh relaxes and the groove mostly disappears. Although the common flesh of Stylotella is unques- tionably contractile, the body of this sponge has never been observed to move as a whole in consequence of this contractility. Thus in no instance did a finger of Stylo- tella when partly immersed in seawater, bend farther into the water, though fingers have been allowed to stand in 28 THE ELEMENTARY NERVOUS SYSTEM positions favorable for this form of response for over a day. Nor have fingers been observed to turn in con- formity to the direction of the general current of water in which the sponge was standing. In some instances the fingers of Stylotella are not directed straight upward, but their tips are turned to one side or the other so that their oscula open laterally. A number of these sponges were set, some with their oscula facing the general current, others with these openings away from the current, and still others sidewise to the current. After three days none of these had noticeably changed their directions, thus giving no evidence of a general movement of the body. Attempts were also made to get evidence of the gen- eral movement of Stylotella through geotropic stimula- tion. This sponge ordinarily grows with its fingers and oscula directed upward, as though it were negatively geotropic. A large colony was, therefore, kept inverted in an aquarium, of circulating seawater for about a week on the assumption that the fingers might be brought to turn from this unusual position, but at the end of this period there was no discoverable change of position. This observation, however, does not prove that Stylo- tella is not geotropic. Slight evidence of its geotropism is to be found in its method of regenerating oscula. When a moderately long finger of Stylotella is cut off and the Avhole of its oscular end removed, the cylindrical body thus resulting will under favorable conditions form a new osculum. Whether this regeneration will take place at the end nearer or farther from the former osculum seems to depend chiefly on the position of the piece of sponge with reference to gravity. If the end that was nearer the former osculum is uppermost, it always regenerates SPONGES 29 a new osculum; if it is down, the opposite end very gen- erally regenerates the new organ. Thus in the regenera- tion of the osculum, Stylotella shows some slight geo- tropic activity, and while it must be admitted that the common flesh of this sponge is contractile, this contrac- tility does not seem to result in movements of the body as a whole such as might be looked for in geotropic and other like responses. It is possible that in this sponge the skeleton, which is well developed, is too resistant to allow the body as a whole to be bent, and that, therefore, the contractility of the common flesh can make itself man- ifest only in the local ways already mentioned. But, as has already been intimated, the chief activity of sponges is not shown in the general movements of their bodies, but in the currents of water that they produce. These currents are due to the incessant activity of the choanocytes in the passages within the body of the sponge. Access to these passages is gained through the innumerable pores on the general external surface of the sponge and the exit from them is through the cloaca and the relatively large terminal opening of the cloaca, the osculum. The cessation or revival of these currents, as seen in many sponges, is not due to changes in the activ- ity of the choanocytes but is dependent upon the closing or opening of the pores and of the osculum, whereby the water current is checked or allowed to pass. The pores on the surface of Stylotella are of the kind designated as .dermal pores, or ostia, in that instead of leading directly to the inlet canals of the flagellated sys- tem they admit to relatively large subdermal spaces which in turn communicate through the incurrent canals proper with the flagellated chambers (Fig. 5). These chambers empty through the excurrent canals into the cloaca. 30 THE ELEMENTARY NERVOUS SYSTEM The opening and closing of the dermal pores in Stylo- teUa has been observed directly in living preparations by Wilson (1910), whose account will be referred to pres- ently. In the following experimental tests the presence or absence of currents was used as an indication of the state of the pores. The demonstration of these currents has been accomplished from the earliest times by the ad- Fio. 5. — Radial portion of a transverse section of Stylotella; the flesh of the sponge is dotted, the cavities are undotted; on the extreme left is the dermal membrane pierced by two pores p that lead into a large subdermal space s, from which incurrent canals i lead to the flagellated chambers /, which in turn open by excurrent canals e into the cloaca c. ditiou to the water of some such substances as carmine, starch, or indigo, whose particles could then be followed as they were carried in the moving water. Latterly this method has been severely criticized by von Lendenfeld (1889), who claims that even these small suspended par- ticles mechanically stimulate the sponge and cause it to close its pores. Von Lendenfeld used milk as an indi- cator and found no objection to it. With Stylo fell a it is easy to demonstrate the pore currents with can nine and the like, and, so far as could be discerned, this material could be used without causing partial closure of these apertures. In fact, as stated by Bidder (1890), carmine particles seemed to have no effect whatever on the dermal pores, but were swept into the interior of the sponge with SPONGES 31 great freedom for hours at a time. It must, however, be confessed that not only carmine but even milk is an un- natural substance for a sponge, and as Stylotella lives in water that ordinarily contains much fine suspended ma- terial, it was found necessary only to watch this material to gain all the information that was needed as to the direction or strength of the currents at the pores. In testing the pores, a finger of sponge was pinned under the microscope in a small glass aquarium so ar- ranged that a continuous current of seawater could be kept running through it, and by watching the suspended particles along the sides of such a preparation under a magnification of about ninety diameters, it was compara- tively easy to ascertain whether the pore currents were running or not. As a rule the objective of the microscope was necessarily plunged under the surface of the sea- water. In making observations it was, however, neces- sary for the time being to stop the current of seawater that was running through the small reservoir in which the sponge was, otherwise the movement of the suspended particles over the surface of the sponge was so rapid that it was impossible to tell whether they entered a pore or simply glided past it. When this current was shut off, the osculum often closed and under such circumstances, as might have been expected, the pore currents ceased. To be certain that the cessation of these currents was due to the closure of the outlet, the tip of the sponge finger including the closed osculum was cut off with the result that the pore currents almost immediately began again. Moreover, when the cut oscular end of a finger on which pore currents could be easily seen was ligated, these currents ceased at once and on the removal of the THE ELEMENTARY NERVOUS SYSTEM ligature they at once recommenced. From these obser- vations it is quite clear that the osculum controls in a purely mechanical way the current within the sponge. When the osculum is open, this current may run ; when it is closed, the current ceases even though the dermal pores are open and the choanocytes continue to beat. In view of these facts the oscular ends of the fingers of Stylotella were regularly removed. Although the presence of a pore current is conclusive evidence of the open condition of the pores, its absence FIG. 6. — Diagram of a finger of Stylotella from which the tip has been cut showing cur- rents entering the subdermal spaces and emerging from the cloaca. is not proof that the pores are closed even supposing that the oscular end is cut off, for it is conceivable that the choanocytes may cease to beat, in which case the cessa- tion of the currents would be misleading as to the condi- tion of the pores. To meet this difficulty a simple pro- cedure was adopted. If the oscular end of a finger of Stylotella is cut off at some distance from the osculum itself, the cut face includes not only the cloaca and some of the flagellated chambers, but also the subdermal spaces (Fig. 6). An examination of the currents from such a cut end will show a large, slow, central current emerging from the cloaca, and a considerable number of smaller more rapid currents entering the surrounding subdermal SPONGES 33 cavities. These cavities form a set of intercommunicat- ing spaces over the whole surface of the sponge, and the currents that set into them at the cut end depend largely upon the action of the choanocytes. If, now, no inward currents can be detected at the dermal pores but currents can still be seen to enter the subdermal cavities at their cut ends, it is clear that the absence of the currents at the pore entrances is due to the closure of the pores, and not to the cessation of the choanocytes. In this way the entrance of currents into the dermal pores can be used to indicate the open condition of the pores, and their ab- sence, when coupled with currents into the subdermal spaces, to indicate the closed condition of the pores. The means of inducing the opening and closing of the dermal pores in Stylotella as tested on preparations such as those described in the last paragraph were found to be numerous. The pores closed on injury inflicted in the neighborhood of the surface under inspection. They also closed when the seawater passing through them contained a small amount of ether (1/200), chloroform (1/200), strychnine (1/15,000), or cocaine (1/1000). They were apparently unaffected by mechanical stimulation, except injury, by temperatures as low as 9 degrees centigrade and by light. They opened in solution of atropin (1/1000), of weak cocaine (1/10,000), in dilute seawater, deoxygenated seawater, and warm seawater, 35 degrees centigrade (Parker, 1910 a). The mechanism of opening and closing the dermal pores is more complex than has generally been assumed and has been very adequately described by Wilson (1910). In Stylotella, according to this author, the membrane cov- ering each subdermal cavity and containing the pores, 3 34 THE ELEMENTARY NERVOUS SYSTEM the so-called dermal membrane (Fig. 7), is composed of an external epidermis which is believed to be syncytical and not cellular as heretofore assumed, of an intermedi- ate mesenchyme containing spicules, and of an inner epithelioid membrane forming the lining of the subder- mal cavity. Each pore consists of two parts, the pore or actual opening at the surface and the pore canal, a FIG. 7. — Dermal membrane over a subdermal cavity in the sponge Stylotella seen from the exterior as a somewhat transparent preparation. The outline of the subdermal cavity is indicated by the line of limiting cells I, within which is the dermal membrane pierced by six pores, three of which are partly closed by pore membranes. (After Wilson, 1910.) very short canal that leads through the thickness of the dermal membrane to the subdermal cavity below. In Stylotella the pores can be seen to contract and to close by the activity of the epidermal syncytium (Fig. 8). With the closure of the pore this layer forms over the external end of the pore canal an extremely thin sheet, the pore membrane, near the middle of which the pore has disappeared. In this state the pore membrane is some- what like the head of a drum, the pore canal representing the body of the drum. In other sponges studied by Wil- SPONGES 35 son, Reniera and Lissodendoryx, not only may the pores close by the formation of a pore membrane, but this process may be supplemented by a closure of the pore canal itself. This is probably due, according to Wilson, to a contraction of the epithelial lining in the pore canal acting after the fashion of a sphincter. Thus there ap- pear to be two somewhat independent devices for the closure of the pores, the pore membrane and the pore c FIG. 8. — Three dermal pores in Stylotella showing steps in their closure by the pore mem- brane; A, partly closed; B, more nearly closed; C, completely closed. (Modified from Wilson, 1910.) canal sphincter. All three sponges studied by Wilson were found to close their pores by means of pore mem- branes. Contraction of the pore canals was observed in Reniera and in Lissodendoryx, but not in Stylotella, though this sponge was not so favorable for such observations. The closure of the pore canals, according to Wilson, is quite obviously dependent upon the sphincter-like band of cells on the wall of the canal. These cells are in every way comparable to a primitive form of smooth muscle- fiber. Their superficial position places them in contact with the water passing through the canal and, as they respond to the differences in this water, they are without doubt capable of direct stimulation. The pore membrane is less muscle'-like in its action than the wall of the pore canal is and yet its movement is hardly to be described as 36 THE ELEMENTARY NERVOUS SYSTEM purely amoeboid. It seems to represent a stage of differ- entiation between amoeboid motion and simple muscle con- traction which may well indicate the kind of contractility that the common flesh of the sponge possesses. But wherever these two types of tissue belong, they certainly both exhibit nothing in their activities that suggest a coordinating system. They act with great independence and as though they were stimulated directly, and in this respect they exhibit all the characteristics of indepen- dent effectors. CHAPTER III SPONGES (Continued) THE oscula of sponges are, in comparison with their pores, relatively large openings and hence they admit of experimental treatment such as is not possible with the pores. In Stylotella each finger ordinarily carries at its free end a single large osculum. The opening and clos- ing of this osculum is the most obvious response of this sponge. If a colony under ordinary conditions is exam- ined, some of the oscula will almost certainly be found closed, though the majority will be widely open. If a small colony is closely inspected under a low power of the microscope, the open oscula will be seen to emit a large number of minute particles indicating that a current is setting out through these openings. In what seem to be closed oscula a minute but otherwise similar current can often be detected, showing that they are really not closed. Some oscula, however, show absolutely no current, though it has been invariably found that when in such cases the oscular tip is cut off, the current can be seen almost instantly and it is, therefore, concluded that at times the oscula do close completely and thus check ab- solutely the current that ordinarily passes through them. In order to get some idea of the natural movements of the oscula, a vigorous colony of Stylotella was isolated and three of its oscula were kept under approximately hourly observation for three days. The results of these observations are summarized in Table 1. 37 38 THE ELEMENTARY NERVOUS SYSTEM TABLE 1 PERIODS OF TIME IN HOURS AND MINUTES DURING WHICH IN THE COURSE op THREE DAYS OSCULA 1, 2 AND 3 WERE OPEN OR CLOSED Number of the Osculum Time in hours and minutes of each successive period of open or closed state Total time in 72 hours Open Closed Open Closed Open Closed Open Closed Open Closed Osculurn 1 . 0.45 21.50 0.45 2.00 3.20 0.25 19.05 24.20 21.40 3.20 1.30 2.35 20.15 21.00 23.50 7.50 2.35 16.10 42.40 67.10 68.45 29.20 4.50 3.15 Osculum 2 Osculum 3 . 0.15 23.10 Since the three oscula whose conditions are recorded in Table 1 were on the same colony and near together and were exposed to almost identical surroundings, the fact that osculum 1 was closed on the average one hour in every two and a half, while oscula 2 and 3 were closed only one hour in eveiy eighteen, must be attributed to the difference in constitution of osculum 1 as contrasted with that of the other two. The condition of general openness as exemplified by oscula 2 and 3 is doubtless typical for these organs. At least in any vigorous sponge under normal conditions, the majority of the oscula will be found open much of the time. The closure of the oscula can be brought about in a variety of ways. When a colony of Stylotella is trans- ferred from running water in the exterior to a collecting bucket of quiet water, the oscula commonly close. They reopen on being brought again into running water. This and many other similar experiments pointed to the im- portance of currents in keeping the oscula open, but this form of experiment did not show what particular aspect of the current caused the osculum to open or to remain open. Did the sponge give out excretions that in quiet water gathered to such an extent in its immediate neigh- borhood as to cause its oscula to close and only on the SPONGES 39 removal of these by a current of water would the oscula open, or did the current carry oxygen to the sponge or act in a purely mechanical way to induce the opening of the oscula? To test these matters the following simple experiment was tried. Three cylindrical glass aquaria of considerable size were placed at three levels so that the water from the uppermost aquarium could be siphoned freely into the intermediate one, from which the water overflowed into the third. Having filled the apparatus with seawater, it was possible to keep it running continu- ously with the same seawater by returning that which collected in the third or lowest aquarium to the upper- most one. If, now, the current of seawater carried away excretions from the sponge or brought oxygen to it and these operations had anything to do with the opening of the oscula, the use of the same water over and over again ought soon to bring on a condition that would no longer cause the oscula to open. But sponges placed in the cur- rent of the middle aquarium remained with their oscula open for hours in seawater that had been used many times over. Moreover, the oscula closed quickly when the cur- rent was cut off and reopened soon after it was started again. It, therefore, appears that the mechanical stimu- lation of a current of water is an effective means of open- ing or keeping open the oscula of Stylotella. These first experiments were made on whole colonies of Stylotella and only the general condition of the oscula was recorded. It was desirable, however, to ascertain what part of the sponge must be exposed to the current to induce an opening of the osculum or the reverse. To test this question a colony of Stylotella was placed in a strong current of seawater and, when the oscula were well opened, a glass tube was lowered over a vertical finger 40 THE ELEMENTARY NERVOUS SYSTEM of the sponge so that the tube protected the length of the finger from the laterally impinging current but was at no place in contact with the finger (Fig. 9). The water in this tube on examination was found to be for the most part quiet ; its condition, however, did not interfere with the slight currents produced by the sponge itself. Al- though the osculum of the finger under examination was fully open when the tube was lowered over the finger, it closed in seven minutes after the tube was in position and remained so for a quarter of an hour. A small tube was now inserted fV \J "V into the upper end of the large tube fjT~j I I \ surrounding the finger and a gentle current of seawater was run down Fio. 9. — Fingers of Stylo- . ,. tella in a strong current of sea- into the Water SUrrOUndlUg the water. One finger is protected from the current by haying a linger. In tourteen minutes the glass tube lowered over it and t closed; the oscula of the other OSCUlum WES again fully Open. Oil feT/L^sho^nVy^hra^rows: cutting off the current the osculum closed in six minutes. It is note- worthy that during the time of these experiments the oscula in the immediate neighborhood of the one tested showed no changes in reference to those observed in the individual within the tube, but they remained for the most part continuously open in the general cur- rent of seawater. The next question that naturally suggested itself was how much of a finger must be exposed to a current of sea- water to induce the opening of its osculum. To test this, ;i iclass tube was made to cover the distal half of a finger, leaving the proximal half exposed to the general current. To check the eddying of the current up into the tube a SPONGES 41 small ring of cotton-wool was inserted between the free end of the tube and the sponge. Under these conditions the osculum closed in eight minutes even though the lower half of the finger carrying the osculum was in a strong current of seawater. This form of experiment was re- peated with only the distal fourth of the finger protected from the current, and again the osculum closed in seven minutes. Thus it is only neces- sary to have quiet water around the outermost fourth of a finger to cause its osculum to close, and a strong current on the proximal three-fourths of the finger will not induce the osculum to open. These experiments were next reversed and attempts were made to ascertain how much of the dis- tal tip of a finger must be exposed to a current to induce the opening of its osculum. In making these trials, a piece of light-weight brass-tubing was cut to such a length that when it was slipped down over a vertical finger of the sponge, it covered the finger all but the tip (Fig. 10). The space between the oscular tip and the tube was filled with cotton-wool and the whole allowed to stand in quiet seawater. After the osculum had been closed for about a quarter of an hour, a gentle current was started across the end of the tube so that it impinged on only the oscular membrane. In three minutes the oscu- lum showed signs of opening and in eight minutes it was fully open. This form of experiment was many times repeated with essentially similar results. The closing of Fio. 10.— A finger of Stylotella protected from a gentle current of seawater by having apiece of brass- tubing lowered over it and the in- terspace between it and the tubing filled with cotton-wool. The cur- rent impinges only on the tip of the finger, whose osculum never- theless has opened and emits a current of water, as indicated by the arrow. 42 THE ELEMENTARY NERVOUS SYSTEM the osculum in quiet water and its opening in a current of water are both very local reactions and cannot be induced from points half a centimeter distant on the finger. If the oscula of Stylotella close simply because the water in contact with them ceases to move and not in con- sequence of the accumulation of waste or lack of oxygen, they probably close in the air on a falling tide because of the same mechanical conditions. If, in the laboratory, an inverted test-tube full of air is lowered over a finger whose osculum is open till the oscular membrane just comes in contact with the air, the osculum closes in about three minutes. The same result can be obtained when the test-tube contains washed hydrogen in place of air. Hence this reaction is not due to the oxygen of the air, but is very probably induced by a purely mechanical con- dition of quiescence into which the tip of the finger passes in going from the water into the gas. If a finger of Stylotella is cut off about a centimeter from the osculum, that aperture even in a current of sea- water is likely to close within a short time and to remain closed for an hour or more. If the finger is cut off at two centimeters from the osculum, there is less likelihood of the closure of the osculum than when the finger is cut at one centimeter. If a pin is stuck into a finger at one and a half centimeters from the osculum, this opening will close in about ten minutes. Thus injury to an adjacent part will bring about a closure of the osculum. The nature of the stimulus produced by injuring the flesh of a sponge seems to be rather mechanical than otherwise. Such an injury besides disrupting the tissues mechanically does little more than liberate their juices. These juices, how- ever, when collected and discharged on a normal sponge with open oscula do not cause the oscula to close. Hence SPONGES 43 the closure of the oscula on injury to the sponge is probably due to a mechanical rather than to a chemical stimulation. Oscula close in running seawater containing small amounts of ether (1/200), chloroform (1/200), strychnine (1/15,000), cocaine (1/1000), and in deoxygenated sea- water. They contract, but do not close, in dilute seawater and at temperatures higher than normal, 35 to 45 degrees centigrade. They remain open in currents of seawater after which their closure is inhibited by cocaine (1/10,000) and by atropine (1/1000), as well as by fresh water. Their activities are apparently not controllable by low temperatures, 25 to 9 degrees centigrade, by light, nor by weak solutions of cocaine (1/50,000) and of atro- pine (1/10,000). On the inner face of the oscular collar of Stylotella, there is a conspicuous sphincter whose contractile cells or myocytes are in many cases so close to the cavity of the osculum as to be in direct contact with the water pass- ing through it. It is probable that the epithelial lining of this region of the osculum is made up of elongated encir- cling and contractile epithelial cells such as Wilson has shown to occur in the main efferent canals of this sponge. This sphincter is doubtless responsible for the closure of the osculum which may open in consequence of the con- tractility or of the elasticity of the surrounding flesh. The fact that the osculum closes in from about three to eight minutes after the application of the stimulus and that it opens in from about seven to fourteen minutes when placed in running water shows that the sphincter is a more rapid and efficient mechanism than that concerned with the process of opening which after all may depend upon simple elasticity. Thus the closing and opening of 44 THE ELEMENTARY NERVOUS SYSTEM 10 the osculuin is accomplished by one set of contractile cells, in the nature of primitive muscle, working against an- other set of contractile cells or even against simple tis- sue elasticity. In this respect the motor mechanism of the osculum is more uniform than that of the dermal pores, for in the osculum there is no pore membrane. In the osculum as in the der- mal pore, the active mechan- ism is apparently directly stimulated and exhibits, therefore, another instance of independent effectors normally called into action by direct stimulation. The water currents pro- duced by sponges are depen- dent upon the activity of their flagellate cells or cho- anocytes, and these cells ap- parently beat on unremit- tingly and incessantly. There is no reason to suppose that they ever stand still or re- verse their direction of stroke. The only control over these currents seems to be the opening and closing mechanisms of the dermal pores, oscula, and other like devices. It has, howrever, been suggested that the currents are too considerable to be checked thus and that other means of control must be present. The strength of the water current from Stylo- Fio. 11. — Diagram of the apparatus for measuring the strength of a current of sea- water produced by Stylotella. A finger of this sponge, whose currents were in full activity, was tied off at the base and firmly attached at the tip to a vertical glass tube into which it discharged seawater. The seawater in the tube rose to a higher level than that of the outside water in conse- quence of the strength of current produced by the sponge. This strength of current could be measured by an attached scale. SPONGES 45 tella can be easily tested by tying into its osculum a glass tube of appropriate caliber and by allowing the sponge to work till the column of water in the tube conies to a con- stant level (Fig. 11). By attaching a millimeter scale to the tube this level can then be read, after which the sponge, without disturbance to the apparatus, can be cut off from the tube and the level to which the water in it drops can again be read. The difference between these two readings gives the water pressure generated by the sponge. In Stylotella this proved to be between 3.5 and 4.0 milli- meters. The current produced by Stylotella, then, has a maximum pressure equivalent to a column of water 3.5 to 4.0 millimeters in height (Parker, 1910 a). Determinations of the current strength on seven spe- cies of Bermuda sponges gave somewhat lower results as shown in Table 2. In these sponges the average pressures TABLE 2 CURRENT PRESSURE IN MILLIMETERS OF SEAWATER AS EXHIBITED BY SEVEN SPECIES OF SPONGES FROM THE BERMUDA ISLANDS Name of Sponge Height of Column of Seawater in Millimeters for five determinations Average Height Tethya 2.5 2.5 1.0 3.0 2.0 2.5 3.0 2.0 2.0 2.0 2.5 2.0 2.0 3.0 2.5 2.0 1.5 3.0 2.5 2.0 2.0 2.0 2.5 1.0 3.0 2.0 2.0 2.5 2.0 2.0 1.0 3.0 2.0 2.0 3.0 2.2 2.2 1.3 2.9 2.1 2.1 2.7 S'p'ircistrcllct Pachychalina fipinosclla Tedanid Stelospongia HiTcinio, of the currents varied from 1.3 to 2.9 millimeters of water. It is therefore clear that the pressure of the currents produced in sponges is very inconsiderable. The volume of these currents, however, is relatively great and has been determined roughly for Spinosella by the following method. A glass tube of known caliber 46 THE ELEMENTARY NERVOUS SYSTEM was tied into the osculum of this sponge, and the sponge and tube submerged in an aquarium. The flow of water through the tube was then determined by measuring the velocity of floating particles, such as grains of carmine, that were carried up the tube by the water current. This proved to be very nearly 4 millimeters a second. As the diameter of the glass tube was 17 millimeters, the osculum must have discharged a little over 0.9 of a cubic centimeter of water a second. At this rate the discharge would amount to some 78 liters a day. As an ordinary Spino- sella may have as many as twenty such oscula, a colony such as this would have to pass through its substance in a day about 1575 liters of seawater, or over 45 gallons. Sponge currents, therefore, are considerable in volume but low in pressure (Parker, 1914 c). This lowness in pressure, which is rather in contrast with the former belief as to the character of these cur- rents, is favorable for their easy control by the closing and opening devices at the dermal pores and oscula. The adequacy of these devices has been tested in only one in- stance, the dermal pores of Stylotella. A finger of this sponge, in which the dermal pores were closed, was tied to the small end of a glass tube which was bent in the form of a siphon and was so placed that the end carrying the sponge was in one vessel of water and the other end, quite free, was in another vessel of water (Fig. 12). The water in these two vessels was kept at the same level. After the whole apparatus was set up, the water in which the sponge rested was deeply colored with methyl green. The ves- sel with uncolored water was then lowered till the differ- ence in level between the water contained in it and in tli » ^ the hemisphere containing the iris still continued to ex- hibit this contraction. He concluded, though on insuf- ficient grounds, that the sphincter of the pupil was acted on directly by the light. The belief that this reaction was due to an intraocular reflex and not to the direct action of the light on the muscle concerned was shown to be extremely improbable by Steinach ( 1892 ) , who worked on a number of lower vertebrates, but especially on the frog and eel. Steinach found that when the half of the bulb containing the iris was cut down to the extreme by removing the edges of the retina and the ciliary body, the reaction still took place. Moreover, if a very small in- tense point of light was focussed on the iris, the portion of this organ thus illuminated was the first part to begin to contract and this activity spread from the region thus stimulated to the adjacent regions. A histological in- vestigation of the contracted iris showed that the ele- ments concerned in this contraction were the pigmented smooth muscle cells of the sphincter pupillae, whose pig- ment absorbed precisely those rays, the shorter wave lengths, that had been found to be especially stimulating. Steinach, therefore, concluded that in fishes and amphibi- ans the smooth muscle elements of the sphincter pupillae may be directly stimulated by light. The objection to this conclusion raised by Magnus (1899) to the effect that the reaction could be prevented by atropin, though the sphincter would still respond to electrical stimulation, was shown by Guth (1901) not to be well founded. Guth, moreover, was able to demon- strate a contraction in the sphincter nearly two weeks after the eye had been cut out, a period much too long for the persistence of an intraocular reflex mechanism in an organ thus removed from the body. He also was able to 52 THE ELEMENTARY NERVOUS SYSTEM produce contraction by the strong illumination of small groups of muscle fibers isolated from the sphincter. The application of Steinach's conclusions to the iris of the cephalopod eye, to that of birds, in which the sphincter is composed of cross-striped muscle fibers, and even to mam- mals was made by Nepveu (1907). Steinach's investigations, however, received their full- est confirmation from the work of Hertel (1906), who studied not only the eyes of the lower vertebrates but those of the higher forms including even man. As a means of stimulation Hertel used light from an electric arc, from a gas flame, and diffuse daylight. After the optic nerve had been cut, the iris in the eye of the eel and of the frog contracted to all three lights, whereas that in the eye of the cat and the rabbit contracted only to the arc light, a condition also observed in human beings who had suffered a degeneration of the optic nerve. In the three mammals thus tested a pencil of strong light could be thrown through the pupil into the fundus of the eye without producing any contraction, though as soon as this pencil was brought to bear on the edge of the iris a local contraction was observed. Hertel, therefore, con- cluded that the contraction of the pupil in blinded eyes was not due to an intraocular reflex but to direct stimu- lation and that this occurred in the eyes of mammals as well as in those of the lower vertebrates. The mammal iris, however, appeared to be less easily stimulated in this way than that of the fish or the amphibian, for it did not react to the weaker lights, gas light and daylight, to which the others were freely responsive. Thus Hertel 's ob- servations completely confirmed as well as extended those of Steinach. The sphincter pupilltf of the vertebrate eye, and prob- INDEPENDENT EFFECTORS 53 ably also that of the cephalopod eye, may, therefore, be regarded as muscles normally subject to direct stimula- tion by light, notwithstanding the fact that they are also under nervous control. In so far as this type of muscle is open to direct stimulation, it reproduces the condition found in sponges where a superficial sphincter is also the form assumed by the muscle in question. Whether in addition to stimulation by light the sphincter pupillaB in vertebrates is open to direct chemical stimulation, as might be inferred from the work of Auer and Meltzer (1909) on calcium, remains still to be definitely decided. Although the sphincter in the iris of the eye appears, therefore, to be a muscle normally subject to direct stim- ulation, and in this sense illustrates the principle of an independent effector, the muscle that has been most dis- cussed from this standpoint is that of the vertebrate heart. In general two opinions have been held respecting the cause of contraction in this muscle. According to one view, the myogenic theory, the beat of the heart is brought about through changes that are initiated in the heart muscle itself. According to the other view, the neuro- genic theory, the impulses to rhythmic beating originate in nervous tissue and are delivered secondarily to the heart. The myogenic theory implies that this muscle is essentially an independent effector not unlike the muscu- lar organs of sponges. The neurogenic theory places the heart muscle under nervous control, as most of the muscles in the higher animals are. These two theories of the nature of the heart-beat have had a long and complicated history, which has been well reviewed by Engelmann (1904) and by Howell (1906). It will, therefore, not be necessary in the present connec- tion to enter into this side of the subject beyond stating 54 THE ELEMENTARY NERVOUS SYSTEM that of recent years the myogenic theory, which took its origin in the very early work of Harvey, has on the whole gained precedence (Lewis, 1917) and that this is chiefly due to the work of Gaskell, Engelmann, and their fol- lowers. According to these investigators the heart in the lower vertebrates is a muscular tube bent in the form of a letter S and dilated into a series of four chambers (Fig. 13). The heart becomes complicated in the higher forms Fid. 13. — Diagram of the heart of a fish showing in lateral view its four chambers, the venous sinus s, the auricle a, the ventricle u and the bulb 6. The s-shaped axis is indicated by the arrow. chiefly through the appearance of a set of partitions whereby this organ is divided into right and left halves. Contraction normally begins in the muscle of the most posterior chamber, the venous sinus, or when this cham- ber is incorporated in the next chamber in advance, the auricle, in the muscle of the posterior portion of this chamber of the heart. The spot in which contraction originates is called in the hearts of the higher vertebrates the sino-auricular node. From this node the wave of contraction spreads over the auricle and across the nar- row bridge of muscle, to the ventricle. Here it is rapidly propagated by the specialized muscular tissue of the in- ner face of the ventricle, the so-called Purkinye tissue, over the whole of this part of the cardiac muscle, which is thus brought into a unified contraction. According to this view contraction arises in the muscle itself and is INDEPENDENT EFFECTOBS 55 propagated as a wave from the venous to the arterial end of this organ. The discovery by Eemak in 1844 that the heart muscle of vertebrates contains an abundance of nerve cells has been used as a strong argument in favor of the neuro- genic theory. But it is now generally believed that these cells are merely concerned with modifying a heart-beat which in its origin is myogenic. From this standpoint, therefore, the presence of these nerve cells affords no serious obstacle to the acceptance of the myogenic inter- pretation of the heart-beat, but it does leave the adult vertebrate heart a less clear example of an independent effector than might be desired. Although in this respect the adult heart falls short of all that might be wished, the embryonic heart of ver- tebrates is quite otherwise. In the developing chick the heart is the first complicated organ to assume functional activity. It arises after about twenty-three hours of in- cubation and it begins to pulsate about six hours later. At this step in the development of the chick, the stage of ten somites, the neural crests are not yet formed and neuroblasts are not yet differentiated. Hence there is every reason to believe that the heart is absolutely free from possible nervous influence and that its beat must be purely myogenic. Not until some time after the heart has been in action, in the chick on the sixth day of incu- bation, is this organ first invaded by nerve cells, a condi- tion also to be observed in a number of other vertebrates (His, 1893). Hooker (1911) has likewise shown that in frog embryos from which the developing nervous sys- tem has been removed, the heart not only differentiates but eventually beats, a condition that confirms the older results on the chick. Hence the early embryonic heart in 56 THE ELEMENTARY NERVOUS SYSTEM vertebrates must be admitted to have a purely myogenic beat and at this stage to be an absolutely clear instance of an independent effector such as is shown in the muscles of sponges. If further grounds were needed to support the opinion that the embryonic vertebrate heart has a myogenic beat, they can be found in the work of Burrows (1912), who cul- tivated the heart muscle from chick embryos in vitro. Pieces of the beating heart of young embryos were trans- ferred to culture fluids where they became centers of out- growths of cells. In time many of these cells began rhythmic beating at a rate of 50 to 120 beats per minute, a rhythm typical for beating pieces of the ventricular muscle. As some of the beating cells were absolutely dis- connected from the rest of the tissue, they afforded most conclusive evidence of myogenic activity. These unicellu- lar mechanisms are models, so to speak, of the embryonic vertebrate heart, whose nature as an independent effector is thereby exemplified. The independent rhythmic con- traction of skeletal muscle has also been demonstrated through this method by Lewis (1915). The modes of action of the heart muscles of other ani- mals than the vertebrates present considerable diversity. Thus Carlson (1904) has shown that in the King crab, Limulus, the heart-beat is purely neurogenic and is as dependent on extraneous nerve centers as are most other arthropod muscles. On the other hand, the heart of the silkworm is said by Pigorini (1917) to continue to beat even after it has been separated from the central nerv- ous system and cut into isolated segments, a condition that is also claimed for the heart in tunicates. The tunicate heart, in consequence of the remarkable periodic reversal in the direction of its beat, has excited INDEPENDENT EFFECTOES 57 the attention of investigators for many years. The organ itself is in the form of a swollen tube, one end of which is continuous with the blood-vessels distributed to the vis- cera and the other with those distributed to the gills (Fig. 14). The action of the heart, the reversal of which B Fia. 14. — A. Diagrammatic view of a longitudinal section of Salpa (after Herdman) showing by arrows the inlet or branchial aperture and the outlet or atrial aperture as well as the position of the heart h. B. Enlarged view of an isolated heart of Salpa showing the abvis- ceral end g connecting with the gills and advisceral end v running to the viscera. was apparently first observed by Van Hasselt in 1821, consists in a series of advisceral beats, by which the blood is driven toward the viscera, followed by a series of ab- visceral beats, by which it is driven toward the gills. The relations of these sequences of beats have been exten- sively studied by Schultze (1901) in Salpa, and more re- cently by Hecht (1918) in Ascidia. It is commonly believed that in these pulsation series the advisceral equal the abvisceral beats, but a closer ex- amination of the facts shows that in general the ad- 58 THE ELEMENTARY NERVOUS SYSTEM visceral beats predominate. There seems to be no ques- tion but that in the heart of the tunicate there are two centers from which contractions may arise, one at the advisceral and the other at the abvisceral end. It is ap- parently the alternate control of the heart by first one and then the other of these centers that brings about the re- versal in the activity of this organ (Loeb, 1899, 1902). Whether these two centers are located in the muscle of the tunicate heart or in other tissue connected with this muscle is somewhat uncertain. The heart itself is com- monly described as made up of a single layer of muscle cells whose inner face is covered with a delicate endo- thelium. Nerve cells have been supposed by most investi- gators to be entirely absent, but Hunter (1902) claimed to have identified them in Molgula, where they are said to be especially abundant at the two poles of the heart, the regions from which the contraction waves start. Their presence is also mentioned by Alexandrowicz (1913). The observations that the isolated middle portion of the heart, where nerve cells are very sparse, if in fact they are present at all, will continue in rhythmic contraction (Schultze, 1901; Bancroft and Esterly, 1903; Hecht, 1918) even after it has been isolated and placed in sea- water suggest that these cells have no more to do with originating this form of heart-beat than the similar cells in the vertebrate heart have. Hecht (1918) has also pointed out that in perfectly normal specimens of Ascidia the contraction wave in the heart muscle may sometimes originate in this middle portion of the heart, showing that this part may normally act as a pacemaker for the whole organ and that its action in isolation is, therefore, nothing unusual. It has also been determined that the rate of transmission of the contraction wave over the INDEPENDENT EFFECTOES 59 heart of Ascidia is relatively slow, 1.76 to 2.12 centimeters per second, a rate more in accord with muscle transmission than with nerve transmission (Hecht, 1918). The heart of the tunicate, therefore, appears to be like that of the adult vertebrate, an independent effector complicated in a secondary way by the presence of nerve cells. According to the investigations of Lewis and Lewis (1917) the amnion of the embryo chick is an organ that exhibits all the characteristics of an independent effector. This membrane is composed of a single layer of mesen- chyme cells overlaid by a single layer of epithelial cells. The mesenchyme cells differentiate into smooth muscle cells that undergo definite contraction as early as the fourth or fifth day of incubation. So far as is known, there is no nerve supply to the amnion. Hence contrac- tions in the smooth muscle layer in this organ must be excited by the direct stimulation of its fibers. This view is supported by the fact that in tissue cultures made from the amnion there can be easily found in the outgrowing cells after a day or two groups of smooth muscle fibers or even single fibers that exhibit rhythmic contractions. These are absolutely devoid of nervous connections and must, therefore, be stimulated directly, a conclusion that is supported by the fact that after they have ceased to contract rhythmically they can be re-excited to this activ- ity by washing them with a drop of fresh culture medium to which an excess of calcium has been added. There is, therefore, every reason to believe that the smooth muscle layer of the amnion of the chick is an independent effector. The instances of this peculiar type of tissue thus far noticed in animals above the sponges have been taken from the higher animals and almost entirely from the vertebrates. This is due to the fact that since these 60 THE ELEMENTARY NERVOUS SYSTEM animals have been much more fully studied than the lower ones, examples of this rather novel condition are better known among these more differentiated types than among the simpler forms. It is probable that a close ex- amination of the lower invertebrates will disclose many such instances. Thus in the sea-anemone Metridium the circular muscle of the column is very much like that in the vertebrate iris in that it is both under nervous control and open to direct stimulation. This sea-anemone, like most other members of its group, is a somewhat cylindrical animal, one end of which is attached to some firm object in the sea and the other provided with a ring of tentacles that surrounds a centrally located mouth. The circular muscle of the column is a sheet of muscle whose fibers in- vest in a circular fashion the inner face of the cylindrical outer wall of this animal (Parker and Titus, 1916). If this muscle contracts locally it produces a ring-like con- striction around the animal's body. In sea-anemones that have been well fed recently, constrictions of this kind form near the oral end of the animal and pass down the cylindrical column of its body to the opposite end. They may recur every four or five minutes and resemble in a general way the peristaltic waves of the intestine. They are probably concerned with the movement within of the newly acquired food, for they occur almost invariably after extensive feeding. It is probable, though this is not definitely proved, that the peristalsis just described is coordinated by the nerve-net contained within the wall of the sea-anemone 's body. If an area on the outer surface of this wall is thor- oughly anesthetized by allowing crystals of magnesium sulphate to dissolve on it, it can be easily rendered so in- sensitive to touch that the typical contraction of the ani- INDEPENDENT EFFECTORS 61 mal as a whole can no longer be called forth by prodding it. Nevertheless, such stimulation will induce a sharply denned ring of constriction to arise from the spot stimu- lated and pass slowly round the column. As this experi- ment succeeds even after the sea-anemone is deeply and fully anesthetized, the contraction of the muscle must be ascribed to direct stimulation. Hence the circular muscle of the column in Metridium, though partly under nervous control, is certainly open to direct stimulation, and from this standpoint represents an independent effector (Parker, 1916 a). A fuller instance of this condition is seen in the acon- tial muscles of the same sea-anemone, Metridium. This sea-anemone, in common with many other closely related forms, carries within its body great numbers of delicate thread-like filaments, the acontia. These acontia, which may be as much as four or five centimeters long, are attached by one end to the mesenteries in the interior of the sea-anemone's body. Ordinarily the acontia are more or less coiled up within the digestive cavity of the animal. When for any reason the sea-anemone is made to con- tract vigorously and the water contained within its cen- tral cavity is thus driven out, the acontia stream with the current of water in great profusion out of the mouth and out of the lateral pores, or cinclides, by which the diges- tive cavity is put into direct communication with the sur- rounding seawater. Thus the contracting sea-anemone throws out over its body a protecting system of filaments, for these living threads are armed with the most vigor- ous nettling cells of any in this animal and will effectively sting an unwary intruder. As the sea-anemone gradu- ally expands again by taking water into its body, the acontia are slowly drawn back and eventually returned 62 THE ELEMENTARY NERVOUS SYSTEM to their original position in the interior of the animal. This withdrawal is accomplished chiefly by the action of the cilia on the acontia, which form an extensive band on each of these filaments and beat in such a direction as to carry the acontium through the seawater toward its at- tached base. When the acontia are fully extended in the external seawater, and perhaps also after they are withdrawn within the sea-anemone, they may exhibit more or less continuous, slow, serpentine movements. These move- ments are due partly to their cilia and partly to a strand of muscle fibers that extends lengthwise their axial re- gion and is known as the longitudinal muscle of the acon- tium. It has been claimed that this acontial muscle has parallel and close to it a delicate band of nervous tissue, but, for the following reasons, this does not seem to be true. Pieces of acontia four to five centimeters long can be easily obtained from a large sea-anemone and will con- tinue alive and active in seawater for many hours. When such filaments are mechanically stimulated by agitating them in seawater or by dropping seawater on them, or when they are flooded with dilute meat juice, they twist themselves into irregular coils. This response takes place slowly and only after a minute or two. If the stimulus is limited to one end of a long acontium that end and that end only responds by becoming coiled. This reaction will occur as well at the central end as at the peripheral end of a given acontium. When acontia have been kept for twenty minutes or so in seawater containing chloretone, a period long enough to anesthetize the tentacles of an intact sea-anemone, they will still become coiled when flooded with dilute meat juice exactly as unanesthetized acontia do. When acontia still attached to a sea-anemone, INDEPENDENT EFFECTOBS 63 but extending several centimeters away from it, are vari- ously stimulated at their free ends, not the least response has ever been observed in the sea-anemone itself, though the acontia react vigorously in the region to which the stimulus is applied. The stimulation of their free ends seems to have no more influence on the sea-anemone than the cutting of the free end of a long hair has on a human being. From these observations it seems fair to conclude that the acontia of sea-anemones are devoid of nervous structures and that their longitudinal muscle must, there- fore, be stimulated directly as an independent factor (Parker and Titus, 1916). From the instances thus far given it is evident that independent effectors in the form of muscles occur among the most differentiated as well as among the simplest of the multicellular animals. The capacity of these effectors to be stimulated directly is only another aspect of what physiologists have long recognized in respect to ordinary muscle, namely, the great ease with which such tissue can be directly stimulated by almost any agent. The number of these independent effectors will doubtless increase as the animal series is more fully investigated. Some of those already noted, like the acontial muscles of sea-anem- ones, may be survivals of that primitive state seen in the sponges; others, like the muscle of the embryonic heart in vertebrates, may be special adaptations newly brought into being by the exigencies of the particular sit- uation. But however this may be, these examples all point to the principle that of the three elemental constitu- ents of the neuromuscular mechanism, the sense organ, the central nervous organ, and the muscle, the only one that can be thought of as existing independently is the muscle, and that this, therefore, is the most primitive of the three. by G.H. P,rk«r CHAPTER V NEUROID TRANSMISSION IN HIGHER ANIMALS ALTHOUGH sponges give no evidence of possessing any true nervous tissue and have at most only independent effectors in the form of muscles, it would be a mistake to assume that they are devoid of everything that is in any sense nervous in nature. It has already been pointed out that if a finger of the sponge Stylotella is cut into within a centimeter and a half of the osculum, this aperture will usually close after some minutes. The sluggish transmis- sion upon which this reaction depends represents without doubt that elemental property of protoplasmic transmis- sion from which true nervous activity has been evolved. It may, therefore, not inappropriately be called neuroid transmission. This elemental type of transmission prob- ably occurs in many tissues of the higher animals, but it is by no means easily detected, for most tissues are inca- pable of those activities by which such transmission could be indicated. Favorable conditions for its study, how- ever, are found in one type of epithelial tissue, namely, ciliated epithelium. This tissue is not only freely open to stimulation, but it possesses in its cilia convenient parts of an effector kind by which its responses to stimulation can be shown. Extended ciliated fields are, therefore, favorable grounds for the study of neuroid transmission. The structure and function of ciliated cells and epi- thelia have been reviewed recently by Putter (1903), by du Bois-Reymond (1914), and especially by Prenant (1912-1914), and an extensive and original investigation 64 NEUROID TRANSMISSION 65 of these elements has been carried out by Saguchi (1917). Some of the earlier workers advanced the opinion that at least among certain, invertebrates cilia were under the control of the nervous system, and Apathy (1897) went so far as to claim that the intracellular fibrillar system of the ciliated cell was nervous, though he never succeeded in demonstrating a connection between this system and nerves. The fact, however, that no one has ever been able to control ciliary activity through nerves and that all the cases of ciliary coordination thus far brought for- ward can be explained on the basis of neuroid transmis- sion renders the belief in the nervous control of cilia ex- tremely improbable. In fact, it may be stated that at present there is not the least ground for the assumption that true nervous activity is in any direct way involved in ordinary ciliary reactions. Yet notwithstanding this independence of ciliated and nervous tissues, the coor- dination in the activities of a ciliated field is one of its most striking features. If the ciliated epithelium from the palate of the frog is placed under a microscope, the various foreign par- ticles lying on its surface will be seen to be swept along in a definite and constant direction, in this particular in- stance toward the oasophageal end of the tissue. The ciliated field, moreover, will reproduce in its appear- ance almost exactly the aspect of a field of grain over which a gentle wind is blowing. The direction of the undulations in this field agrees with that in which the particles are borne along. When these appearances are further examined, they are found to depend upon two factors, the direction of the effective stroke of the cilia and the sequence of these strokes. The direction in which the particles move is 5 66 THE ELEMENTARY NERVOUS SYSTEM due to the direction of the effective stroke of the cilia. Each cilium moves back and forth on a fixed base. In its forward or effective stroke the cilium acts in such a way as to drive the supernatant fluid and its suspended par- ticles along with it. In its backward stroke or recovery the cilium returns to its former position, imparting as little motion to the surrounding fluid as possible. Thus each effective stroke moves the superimposed fluid for- ward and each recovery leaves this fluid in large part standing. Hence in general the fluid moves on in one direction only, namely, that of the effective stroke. The second factor in the ciliary activity of the frog's palate is shown in the sequence in which the successive cilia beat. In this particular example, as in most other ciliated membranes, the cilia do not beat all at the same moment or synchronously, but in regular order one after another, or metachronously. Thus after one cilium has begun its effective stroke the next cilium in the direction of this stroke takes up the activity, and so on till the wave has passed over the whole field. Thus the cilia of a given membrane do not act independently nor in unison but in sequence, and thus exhibit a high degree of coordina- tion. It is this metachronous coordination that gives to the ciliated membrane the appearance of a field of grain over which a wind is blowing. The two elements of ciliary action that have thus been pointed out, though of necessity intimately associated, are in truth quite independent of each other. Their in- dependence is perhaps best indicated by the fact that there are instances in which their directions are opposed instead of being in agreement, as in the epithelium of the frog's palate. One of the best of these is seen in the swimming plates of the ctenophores (Fig. 15). In these NEUROID TRANSMISSION 67 animals the swimming plates form eight well-defined rows that extend from the aboral pole toward, if not ac- tually to, the oral pole. The effective stroke of each plate is in the aboral direction, thus carrying the animal through the water with its mouth forward. The wave of ciliary action, however, sweeps over each row from its aboral to its oral end and thus takes a course the reverse of that indi- cated by the effective stroke. Hence it may be concluded that the effective stroke and the wave of ciliary action are inde- pendent factors; for though they usually agree in direction they may be directly opposed as in the example just given. The regularity with which one cilium beats after another, the coordinated metachronism of the series, is the feature of the ciliated epithelia that most recalls nervous control and that requires explana- tion. It might be supposed that this regularity was due to the mechanical influence of a given cilium on the one next following and so forth. Thus if cilium A begins to beat, it will strike toward cilium B, which on being struck will thus be called into action and by a similar process excite C and so on. This operation, at least so far as the effective stroke is concerned, is not unlike that seen in the successive toppling over of a row of bricks each on end where the fall of the first brick knocks over the second and so on. Although this explanation finds an easy application to the usual form of ciliary beat in which the effective Fio. 15. — Side view of the cteno- phore Mnemiopsis leidyi. Of the eight rows of swimming plates four are shown, two long ones and two short. All start from the aboral pole o and converge on the oral pole o. 68 THE ELEMENTARY NERVOUS SYSTEM stroke is in the same direction as the general wave, it is not so easily applied to cases where, as in the cteno- phores, the reverse is true, that is, the effective stroke is in a direction opposite to that of the wave of ciliary ac- tion. Here the effective stroke of a plate would not be directed toward the one next to act, but toward the one that had just previously acted. Hence the mechanical explanation offered above fails of satisfactory applica- tion. For reasons that will be given presently this mechanical explanation can also be shown to be entirely inapplicable also to those cases in which the direction of the effective stroke and that of the ciliary wave agree. If the wave of ciliary metachronism is not due to the direct mechanical action of one cilium on another, the form of coordination that this process exhibits must de- pend upon some part of the cell deeper than the ciliated zone. Such might well be the cytoplasm, even the super- ficial cytoplasm, of the cell itself. Evidence in favor of this opinion was advanced by Kraft as early as 1890. Kraft showed by two methods of experimentation that stimuli applied to a ciliated field on one side of a band of quiescent cilia could influence ciliary action on the other side of this band, thus demonstrating a transmis- sion over a region devoid for the time being of any form of mechanical activity. The first method of experimenta- tion involved the use of a small temperature box divided into three chambers and so arranged with inlets and out- lets that each chamber could be supplied with a current of water of fixed temperature (Fig. 16). These cham- bers all abutted on one face of the box and thus this face represented a surface on which there might be established three sharply defined areas each with its own tempera- ture. The whole apparatus was of such small size that NEUEOID TRANSMISSION 69 20° 0-2 20' a sheet of ciliated epithelium could cover all three areas, and the epithelium was always so placed that the axis cor- responding to the direction of its ciliary waves passed over all three chambers. When the two end chambers contained water at 20° C. and the intermediate chamber water at 0° to 2° C., the cilia beat continuously from one end of the membrane to the other. When, however, the temperature of the two end chambers was reduced to from 10° to 12° C., the cilia over these chambers beat at a very slow rate, one to two strokes per second, and those over the middle chamber were entirely qui- escent. If now the cilia over the first end chamber were stimulated to greater ac- tivity by stroking them gently with a fine brush, increased activity soon appeared in those over the other end chamber, though the cilia over the intermediate region remained entirely motionless. Thus a form of transmission other than that carried out by the mechanical activity of the cilia must be admitted to have taken place over the area between the two extremes. A second experimental test was carried out by Kraft with the apparatus just described, but with the use of a temperature not a mechanical stimulus. In this test the two end chambers were set at 10° C. and the intermedi- ate chamber was kept so cold that the cilia over it did not beat. Those over the end chambers beat only slightly. The temperature in the first chamber was then raised from its original point of 10° C. to 15° C. and the cilia over this chamber now began to beat more rapidly than Fio. 16. — Plan of a temperature box divid- ed into three chambers, each of which can be kept constant by a flow of water. The box is of such a size as to allow a piece of ciliated epithe- lium to be spread over the three temperature areas as in the experi- ment by Kraft. 70 THE ELEMENTARY NERVOUS SYSTEM before. Soon after this increase of activity was noted over the first chamber, a like increase appeared in the cilia over the third chamber, though the cilia in the inter- mediate region remained motionless. After a time, how- ever, these too began to beat. Kraft concluded, there- fore, that a region of quiescent cilia could transmit im- pulses to increased activity without showing any ciliary movement itself. The subsequent activity of the inter- mediate cilia he believed to be due to mechanical stimula- tion received from the especially excited cilia over the first chamber. It is thus clear, as Engelmann long ago maintained, that ciliated epithelia may transmit impulses to action without any associated mechanical disturbance. These impulses pass through the deeper protoplasmic parts of the tissue and call forth the successive activity of the cilia which thus gives evidence of this transmission wave. Such an activity is sufficiently nerve-like in char- acter to justify the designation neuroid. In ciliated epithelia this type of transmission exhibits a feature long ago recognized by Griitzner and especially by Kraft wherein it shows a remarkable resemblance to true nervous transmission in even the more specialized types of central nervous organs. Ciliated transmission is limited as to its spread and direction. If a spot in a relatively quiescent field of ciliated epithelium is mechan- ically stimulated the increased activity of the cilia thus produced does not spread in all directions over the field, but forms a band beginning at the spot stimulated and extending in the direction taken by the transmission wave. This band may become somewhat wider than the spot as one recedes from the region of stimulation, but it never spreads to any considerable extent. This condition justi- fies the conclusion that the individual cells in a ciliated NEUROID TRANSMISSION 71 epithelium are somewhat like the neurones in the differen- tiated central nervous organs in that transmission from element to element is easily accomplished in one direc- tion hut not in the reverse. Hence this type of transmis- sion is definitely restricted in its course. That this is a structural feature of ciliated epithelium is shown in the experiment of von Briicke (1916) who demonstrated that when a small piece of ciliated epithelium was cut out and reimplanted after having been turned through a half- circle, it retained after healing the original direction of its effective stroke and its transmission wave was in op- position to that of the surrounding field. Thus whatever it is that determines the polarization of ciliated epithelia, that feature is resident locally in the epithelium itself and is not impressed upon it from some external source. The facts that have thus far been brought out for cil- iated epithelium can also be demonstrated for the most part on the rows of swimming plates on ctenophores. In these animals each swimming plate is like a gigantic flat- tened cilium. It is compound in nature, for it arises from a group of cells instead of from a single cell ; otherwise, it is essentially like a cilium. These plates, as already in- dicated, form eight well-defined rows, extending over the body of the ctenophore from its aboral toward its oral pole. Transmission waves pass over these rows in an oral direction, but the effective stroke of each plate is in the aboral direction, thus driving the ctenophore through the water mouth foremost. The rows of plates, therefore, like rows of cilia, exhibit a well-developed metachronism. The fact that the plates are often 'normally quiescent and may become active only at intervals makes them very convenient for experimental work. If the longitudinal band of tissue to which the plates 72 THE ELEMENTARY NEEVOUS SYSTEM are attached in the common American ctenophore Mne- miopsis is cut across even in a very superficial way, the ordinary transmission waves can be seen to proceed from the aboral pole over the line of plates orally to the cut, where they cease. Beyond the cut the plates for some time after the operation are in incessant vibration, the waves beginning at the cut end of the series and running to the end near the mouth. The oral course of the trans- mission wave is so universally characteristic of the cten- ophores that the polarity of their rows of swimming plates may be said to be as pronounced as that of the rows of cilia in a ciliated epithelium. When a spot about midway the length of a row of plates in Mnemiopsis is touched, the region about it im- mediately becomes depressed and the edges of the de- pression fold over and cover in the plates. This reaction was observed long ago by Verworn (1890) on Beroe, and, though it has been questioned by Bauer (1910), the ob- servation seems to be abundantly confirmed by Lillie (1906) and by Kinoshita (1910), who demonstrated fur- ther that it could occur on a fragment of a Beroe over which the row of swimming plates extended. By this means in Mnemiopsis, and probably in a number of other ctenophores, half a dozen plates or so near the middle of a row may become so much restrained that they will not show the least motion. Nevertheless, transmission- waves that arrive at the aboral entrance to this depression emerge from its oral end with the greatest regularity. This may happen while the covered region is under the closest inspection through a lens and during which not the least sign of movement can be detected in the re- strained plates. Thus the mechanical movement of the swimming plates is no more necessary for the transmis- NEUBOID 'TRANSMISSION 73 sion wave in ctenophores than the movement of the cilia is for a like wave in ciliated epithelia. When a row of plates in a Mnemiopsis pinned out in seawater has passed under it a metal tube of small cal- iber and is chilled by running water at 4° to 5° C. through the tube, the plates in the region subjected to the cold cease to beat, though transmission waves may be seen to arrive at one edge of the cooled area and to emerge at the opposite edge with regularity. Again the swim- ming plates resemble cilia in that in an area in which the plates have been rendered quiescent by chilling transmis- sion is still possible. In handling specimens of Mnemiopsis in the experi- ments last described, it was noticed that when the rows of ' plates under which the metal tube passed were subjected to a little local stretching by the awkward manipulation of the tube, the plates often ceased to vibrate in the stretched region. On repeating this operation it was found that as a rule the slight stretching of the band of tissue to which the plates were attached would bring the plates of the stretched part to a standstill, though it did not interfere seriously with transmission. In such an oper- ation, however, much care was required not to overstrain the tissue, for otherwise a permanent cessation of action followed. Avoiding this difficulty, however, mechanical strain, like low temperature, may be made to check motion without interfering with transmission. Thus in several ways the swimming plates of the ctenophores exhibit all the peculiarities of rows of cilia (Parker, 1905 6). The bands of tissue to which the swimming plates are attached in ctenophores have been studied with much care, but in no instance have nerve cells been found as- 74 THE ELEMENTARY NERVOUS SYSTEM sociated with them in such a way that their ordinary powers of transmission and polarization could be attrib- uted to these elements. It seems likely that transmission along the bands of plates whereby their metachronism is maintained is accomplished through the epithelial cells that compose the bands as it is in the rows of cells in cili- ated epithelia. Although this view of the coordination of the swim- ming plates of ctenophores agrees well with the estab- lished facts in the case, it must not be forgotten that evi- dence has been brought forward to show that these plates, unlike ordinary cilia, are under some nervous control and that in this respect they represent a more complex state of affairs than that seen in ordinary ciliated epithelium. This evidence has been produced by Bauer (1910). If a swimming Beroe is gently touched in the region of the mouth, its swimming plates momentarily cease moving. If it is vigorously stimulated in the same region by being stuck or cut, for instance, its plates act for a short time more vigorously than usual. Thus a slight stimulus pro- duced inhibition, a considerable one excitation. If now the sense body at the aboral pole of the animal is cut out, thus removing what is usually assumed to be the chief coordinating center for the activity of the rows of plates, the same reactions in the plates recur on applying the appropriate stimuli to the region of the mouth. Bauer, therefore, believes that since these reactions can not be ascribed to the aboral sense body they must depend upon the action of the diffuse nervous system which has long been known as a subepithelial network and which, though chiefly concerned with the muscles of the ctenophore, probably also exerts a secondary influence on its rows of swimming plates. Thus the nervous control of the swim- NEUROID TRANSMISSION 75 ming plates is quite subordinate to that through which their normal activity is called forth and which in all re- spects agrees with the type of neuroid transmission al- ready described in ordinary ciliated epithelia. Prom these examples it appears that the ordinary tis- sues of animals, at least their ciliated epithelia, may ex- hibit sluggish forms of transmission that are so like those seen in sponges as to admit of being classed under the single head of neuroid transmission. Such a form of transmission is represented in sponges not only by the closure of their oscular sphincters when a more distant part of the animal is injured, but by their system of flag- ellated cells whose activity, like that of the cells in a cili- ated epithelium, must be coordinated by some such form of transmission. Although of the three identifiable ele- ments of the neuromuscular mechanisms of animals, sense organs, central nervous organs, and muscles, sponges possess only muscles, they nevertheless exhibit among their many activities neuroid transmission, a slug- gish form of transmission that may be considered the forerunner of nervous activities, and in this sense may represent the germ from which has sprung the real nerv- ous conduction of the more complex animals. SECTION II. RECEPTOR-EFFECTOR SYSTEMS CHAPTER VI THE NEUROMUSCULAR STRUCTURE OF SEA-ANEMONES SPONGES are animals in whose structure a very sim- ple type of muscle is the only part that represents the neurornuscular mechanism of the higher animals. These muscles, moreover, are so insignificant in amount and so slight in their action that a living sponge seems more like a plant than an animal in its inertness. Compared with such sluggish responses as those shown by sponges, the movements of hydroids, coral animals, sea-anemones, jellyfish and other coelenterates are quick, though the movements of these animals are in turn slow compared with those of vertebrates and especially of insects. This quickened rate of response, which distinguishes the coelenterates from the sponges, is associated with the fact that the crelenterates possess not only muscles but also nervous organs in the form of simple sensory surfaces by which their muscles are called into action more quickly than they would be by direct stimulation. Such a sys- tem includes two* of the three elements already pointed out as essential to a complete neuromuscular organiza- tion and may be designated from the particular ele- ments present a receptor-effector system. The receptor-effector system with some of its most important modifications is well shown in such animals as 76 THE NEUROMUSCULAE STRUCTURE 77 sea-anemones on which in fact some of the earliest studies in these directions were made. The more typical sea- anemones or actinians (Fig. 17) are cylindrical animals attached by one end, the pedal disc, to a rock or other firm support in the sea and carrying at the other end, the oral P FIG. 17. — Diagram of a longitudinal section of the sea-anemone Metridium; the area of attachment is the pedal disc p; in the middle of the oral disc o is the mouth leading into the oesophagus e which opens into the digestive cavity d. The oesophagus is held in place by the mesenteries m when complete c, the incomplete mesenteries i failing to reach this tube. disc, a cluster of tentacles in the center of which is the mouth. The mouth does not open directly into the single large internal space, the digestive cavity, but leads to a somewhat elongated oesophagus that extends downward into the interior of the actinian to the neighborhood of the pedal disc, where it opens out freely into the digestive cav- ity. The oesophagus, however, does not hang freely in the 78 THE ELEMENTAEY NERVOUS SYSTEM digestive cavity, but is held in place by membranes, the mesenteries, which extend in pairs from the inner face of the cylindrical wall of the actinian's body, the col- umn wall so-called, across the digestive space to the wall of the oesophagus. By means of these mesenteries the oesophagus thus comes to be held in an axial position in the actinian's body, where it serves as the one means of inlet and outlet for the digestive cavity. Although it is a relatively simple tube it is usually provided with a pair of longitudinal grooves, the siphono- glyphs (Fig. 18), by which water is continually pass- ing into the interior of the sea-anemone to escape in an outward current through the rest of the oesophagus. The mesenteries, as can best be seen in a transverse section of a sea-anemone (Fig. 18), are thin sheets of tissue which, as already mentioned, occur in pairs. The members of each pair are separated from each other only by a very narrow space, the entocele, which is really an extension of the digestive cavity. Each pair is separated from the pair on either side of it by a wider space, the exocele. The mesenteries that hold the oesophagus in place extend, as already stated, from that structure to the column wall and are known as complete mesenteries. Those that are attached to that portion of the oesophagus that forms a siphonoglyph have a peculiar arrangement Fio. 18. — Transverse section of the sea- anemone Metridium showing the oesophagus e with its two siphonoglyphs s and its support- ing mesenteries, the directives d and the com- plete non-directive c. Two series of incom- plete mesenteries i are shown. THE NEUEOMUSCULAE STEUCTUEE 79 of muscles and are called directive mesenteries, the others being designated as non-directives. Besides the pairs of complete mesenteries, there are many other pairs of in- complete ones. These are characterized by the fact that they extend only a short distance into the digestive cavity and thus fail to reach the oesophagus. The walls of a sea-anemone's body are in all places relatively thin. They are for the most part epithelial in nature and are composed of two layers of cells separated by a third layer of partly secreted, partly cellular ma- terial. Covering the whole exterior of the animal is the ectodermic layer, which at the mouth is reflected inward over the inner surface of the resophagus to the inner ter- mination of that tube. The whole interior of the sea- anemone is covered with the entodermic layer, which unites with the ectoderm at the inner end of the esoph- agus. Between the ectoderm, which thus covers the ex- terior and the entoderm that lines the interior, is a third layer, the supporting lamella. As has already been stated this is composed of secreted substance containing numerous cells. It is resistent enough to give a good deal of support to the sea-anemone; hence it partakes of the nature of a skeleton. The body wall of the sea-anemone, as already mentioned, is everywhere relatively thin and wherever it is punctured the three layers mentioned are cut through, the ectoderm first, then the supporting la- mella, and finally the entoderm, after which the digestive cavity is invaded. The resolution of these layers of tissues into their cellular elements was first successfully accomplished by 0. and E. Hertwig (1879-1880). Notwithstanding the dif- ferences in position and function of the ectoderm and en- toderm, there is much uniformity in their neuromuscular 80 THE ELEMENTARY NERVOUS SYSTEM structure. The superficial portion of each is epithelial in character (Fig. 19) and contains among its various types of cells a number of sensory cells that terminate superficially in free, bristle-like endings and that branch at their deep ends into delicate fibrils. This epithelial portion may be designated as the first sublayer. The fibrils from the deep ends of the sense cells constitute collectively a nervous sheet, the second sublayer, in which may be found not only the deep terminations of the sense cells but also special ele- ments, the so-called ganglion cells, whose branches add to the wealth of fine fibrils from the sense cells. Still deeper than the nervous layer is the muscular or third sublayer composed almost en- tirely of elongated muscle cells closely applied to the supporting lamella or even partly imbedded in it. These three sub- layers can commonly be identified in many parts of the ectoderm and the entoderm. According to the Hertwigs, when the sensory cells of a sea-anemone are stimulated, they transmit impulses to the nervous sublayer which in turn excites the muscles to action and thus the animal is brought to respond to an external change. If the stimulated sensory cells are in the ectoderm and the responding muscles are in the ento- derm, it was suppbsed by these investigators that the nervous impulses pass through the ectodermic nervous sublayer over the exterior of the animal to the resophagus, at whose inner end the impulses . are transferred from the ectoderm to the entoderm and thus gain access to Fio. 19. — Diagram of the ectoderm of a sea-anemone showing the epithelial e, the nervous n, and the muscular m sublayers. THE NEUROMUSCULAR STRUCTURE 81 the musculature of that layer. Although the Hertwigs believed that the nervous sublayer was rather uniformly developed all over the actinian, they maintained that a specially rich nervous region was to be found in the ecto- derm of the oral disc, where the beginning of a central nervous organ might be said to occur. This view that the actinian possessed an oral concentration of nervous tissue was accepted by Wolff (1904), and by Groselj (1909), who, however, placed the concentration in the wall of the oesophagus rather than in the oral disc. Havet (1901), who studied the nervous system of the sea-anemone Metridium by means of the Golgi method, was unable to confirm the statement that the nervous ele- ments were more abundant in the neighborhood of the mouth than elsewhere and declared that their arrange- ment was such as to justify the expression diffuse rather than centralized. Havet not only claimed a diffuse nerv- ous system for actinians, but he maintained that there were grounds for changing in certain important particu- lars the scheme of nervous interaction proposed by the Hertwigs. According to Havet the so-called ganglion cells described by these authors are really motor nerve cells which receive impulses from the sensory cells and transmit them to the muscles. Thus the elements in the neuromuscular organization of an actinian form a se- quence that reproduces in miniature that seen in the cen- tral nervous organs of the higher animals. Here, as, for instance, in the vertebrate spinal cord, a sensory neurone connects with a motor neurone which in turn leads to a muscle. The reflex arc thus outlined is reproduced in the actinian in that its sensory cell corresponds to the sensory neurone of the vertebrate and its motor nerve cell to the motor neurone. Thus the actinian and verte- 6 THE ELEMENTARY NERVOUS SYSTEM brate nervous system exhibit in one fundamental par- ticular a most striking similarity. Havet also claimed to have demonstrated a much closer relation between the ectodermic and entoderniic nervous layers than was suspected by the Hertwigs. He believed that he could show by means of the Golgi method that nervous fibrils pass from the ectoderm through the supporting lamella to the muscles of the entoderm and thus establish a direct union between structures that, ac- cording to the Hertwigs, were only indirectly united through the esophagus. This claim has been abundantly confirmed by Parker and Titus ( 1916) , who have shown by a special technique for nervous tissue that the supporting lamella of the actinian Metridium contains an abundant meshwork of branching neurofibrils that can be traced from the ectodermic side of this layer through its sub- stance to the more important systems of muscles in the mesenteries. Moreover, the supporting lamella can be seen to contain a great number of branching cells which have all the appearances of true nerve cells and which presumably form an essential part of the conducting sys- tem between the ectodermic sensory areas and the ento- dermic musculature. These observations revive in a way the opinion early advanced by von Heider (1877, 1895) that the supporting lamella in a number of actinians contains nervous ele- ments, a claim that, notwithstanding the opposition of Wolff (1904) and of Kassianow (1908), has been sup- ported by the conditions found in the soft corals by Hick- son (1895), by Ashworth (1899), and by Kiikenthal and Broch (1911). Thus, though the details of the nervous organization in the actinians is only just beginning to be THE NEQROMUSCULAR STRUCTURE 83 worked out, the presence of nervous elements in these animals is beyond dispute. The effector systems of sea-anemones, as might be expected, are more numerous and complicated than those in sponges. Sea-anemones possess at least four systems of effectors: mucous glands, ciliated epithelia, nemato- cysts, and muscles. There has never been any ground for the assumption that the mucous glands and the cilia in coelenterates are under nervous control. These effectors respond only to direct stimulation and are not open to influences from a distance. Even in the case of such coelenterate cilia as those of the lips of actinians where by appropriate stim- ulation a reversal of the effective stroke can be brought about (Parker, 1896, 1905 a; Vignon, 1901), the whole reaction is so strictly local that there is not the least reason to assume the intervention of nerves. The mucous glands and cilia, therefore, bear all the marks of indepen- dent effectors and hence free from nervous control. The nettle cells with their contained nematocysts, on the other hand, have often been regarded as subject to nervous influence. Schulze, who studied these cells in Cordylophora with great care in 1871, showed that each one was provided with a special bristle-like projection, a cnidocil, by which it could be stimulated directly, and argued from this that in their action they were indepen- dent of the nervous system. Nevertheless the discovery by the Hertwigs (1879-1880) that their basal processes branched as those of the sensory cells did, led these and many other investigators to believe that the nettle cells had nervous connections. This opinion has been ex- pressed even as recently as 1913 by Baglioni. There is, however, not the least experimental ground for assum- 84 THE ELEMENTARY NERVOUS SYSTEM ing that nettle cells are discharged through nervous ac- tion. In the sea-anemone Metridium these cells are abun- dantly present on the tentacles and especially on those delicate filaments from the interior of this animal, the acontia, Nevertheless, from both these structures the nettling threads can be discharged only under direct stimulation and this continues to be true even after the part in question has been thoroughly and completely an- esthetized with magnesium sulphate or chloretone. As these drugs temporarily abolish all traces of nervous ac- tivity and yet in no apparent way affect the activity of the nettled cells, it is most probable that these cells, like the mucous cells and the ciliated cells, are independent effectors and not under nervous influence (Parker, 1916 a). Of the nervous control of the fourth type of effector in actinians, the muscle, there is abundant evidence. The muscles of these animals were regarded by the earlier workers as more or less continuous sheets that gave to the animal as a whole some tiling of the character of a contractile sac. But after the publication of the impor- tant paper by the Hertwigs (1879-1880) on the structure of these animals, it became evident that their muscula- ture was more differentiated than had been previously supposed and that a considerable number of well denned muscles or groups of muscle could be distinguished. These muscles, as already stated, occur in the deep por- tions of the ectoderm and the entoderm and may in some instances even invade the supporting lamella. In Metridium it is possible to distinguish at least thir- teen such sets of muscles. Their positions are indicated in Fig. 20. Of the thirteen only two are found in the ecto- derm, the remaining eleven being in the entoderm. They THE NEUROMUSCULAR STRUCTURE 85 may be described briefly in the following order beginning with those in the ectoderm. 1. The longitudinal mnscle of the tentacle is an elon- gated conical sheet of ectodermic fibers that are in direct Fia. 20. — Diagram of a partial longitudinal section of the sea-anemone Metridium show- ing the thirteen classes of muscles: 1, longitudinal muscle of the tentacle; 2, radial muscle of the oral disc; 3, circular muscle of the ossophagus; 4< circular muscle of the oral disc; 5, circular muscle of the tentacle; 6, circular muscle of the column; 7, sphincter; 8, circular muscle of the pedal disc; 9, basilar muscle; 10, longitudinal muscle of the mesentery; 11, transverse muscle of the mesentery; 1%, parietal muscle; IS, longitudinal muscle of the acontium. contact with the supporting lamella. They course length- wise in the grooves and on the crest-like elevations that extend up and down the outer surface of the lamella, and 86 THE ELEMENTARY NERVOUS SYSTEM show no special grouping, being uniformly distributed around the whole tentacle. 2. The radial muscle of the oral disc is made up of ir- regular dense bundles of ectodermic fibers more or less imbedded in the supporting lamella of the disc. They radiate from the region of the mouth outward toward the periphery of the disc, making their way between the bases of the tentacles. 3. The circular muscle of the oesophagus ensheathes the oesophagus on its entodermic face. It is not very strongly developed and its fibers, which take a circular course, are more or less interrupted where the complete mesenteries are attached to the cesophageal wall. 4. The circular muscle of the oral disc is a flat cir- cular ring, whose fibers take a course concentric with the mouth and are often much involved in the supporting la- mella of the disc on its entodermic side. 5. The circular muscles of the tentacles are conical sheets of muscle on the entodermic side of the supporting lamella of those organs. In each muscle the fibers take a circular course and are fewer in number and finer than in the longitudinal muscle of the tentacle. They show no special differentiation except at the base of the tentacle, where there is a slight tendency to form a sphincter. 6. The circular muscle of the column is a well-devel- oped sheet of cylindrical fibers covering the entodermic face of the supporting lamella of the column from its attachment to the pedal disc to the region of its transition to the oral disc. The fibers in their circular course pass the lines of attachment for the mesenteries at right angles, but are not to any great extent interrupted at these lines. 7. The sphincter is a firm circular band of muscle fibers embedded in the supporting lamella of the column THE NEUROMUSCULAR STRUCTURE 87 but so distributed that they show at once that they have been derived from an entodermio source. In fact, the sphincter is merely a differentiated part of the circular muscle of the column. 8. The circular muscle of the pedal disc is a well-de- veloped, vigorous organ composed of a system of fibers concentric with the disc and more or less imbedded as circular bundles in the inner face of the supporting la- mella of the disc. 9. The basilar muscles are radial strands that extend along the mesenteries at the junction of these organs with the pedal disc. There is a pair of these muscles for each mesentery and they vary in length in accordance with the size of the mesenteries to which they are attached. These muscles cross the fibers of the circular muscle of the pedal disc at right angles and lie only a very short distance orally from them. 10. The longitudinal muscles of the mesenteries are sheets of muscle fibers on the exocele faces of the direc- tive mesenteries and on the endocele faces of the non- directives and of most of the incomplete mesenteries. They extend from the oral to the pedal disc. 11. The transverse muscles of the mesenteries are very thin uniform sheets of muscle that cover the endo- cele faces of the directives and the exocele faces of the other larger mesenteries. They are thus on faces op- posite those on which the longitudinal muscles are lo- cated. They are better developed on the complete mesen- teries than on the incomplete ones, from the smaller of which they may be entirely absent. 12. The parietal muscles of the mesenteries consist of longitudinal ridges on the exocele and endocele faces of almost all the mesenteries at their line of junction with 88 THE ELEMENTAEY NERVOUS SYSTEM the column wall. On the larger mesenteries these muscles are small and inconspicuous in comparison with the other musculature of these organs, but on the very small mesen- teries they are almost if not quite the only muscles present. 13. The longitudinal muscle of the acontium in Metrid- inm is a delicate double band of fibers that extend length- wise the filamentous acontium, one of which is attached to the free edge of each mesentery. These thirteen sets of muscles constitute the effectors through which the nervous system of the actinians finds its only means of reacting on the surroundings, for the other three systems of effectors, the mucous glands, the cilia, and the nettle cells, are entirely free from nervous control. CHAPTER VII NERVOUS TRANSMISSION IN SEA-ANEMONES A PLAN for nervous transmission in the body of a sea- anemone was long ago described by 0. and R. Hertwig (1879-1880). According to tins plan, it was believed that the stimulation of any given group of sensory cells on the surface of such an animal would excite activity in the sub- jacent nervous layer, which, in turn, would call forth contractions in the underlying muscles and thus originate a response. This is well exemplified in the tentacles of sea-anemones. When a strong stimulus is applied to the ectoderm of these organs they immediately respond by retracting due to a contraction of their longitudinal ecto- dermic muscles. If, however, as is often the case, the sensory cells stim- ulated are in the ectoderm of the animal and the respond- ing muscles are in its entoderm, the course of transmis- sion as advocated by the Hertwigs was believed to be much less direct than in the former instance. In this case the nervous impulses that arose in the ectoderm were be- lieved to be transmitted from their region of origin through the nervous layer of the ectoderm including the oral disc to that of the oesophagus at whose inner edge they passed over into the nervous layer of the entoderm by which they were conducted to the appropriate ento- dermic muscles. Thus the inner edge of the resophagua was believed to be the region of nervous intercommunica- tion between the ectoderm and the entoderm. Histo- logical evidence that has led to the suspicion that this 89 90 THE ELEMENTARY NERVOUS SYSTEM view of the nervous connection of ectoderm with ento- derm is not wholly correct has already been given in the preceding chapter, but much more conclusive physiologi- cal evidence to this effect will be presented here. By appropriate lines of incision through the thin walls of sea-anemones it is possible to make preparations by which the courses taken by nervous impulses through the bodies of these animals can be determined with much accuracy (Parker, 1917 a). To test such preparations it is necessary to use a means of stimulation that is both accurately controllable and strictly local. Such a means is found in the mechanical stimulation produced by a delicate blunt glass rod. When the surface of a Metrid- ium is explored by such means the degree of sensitive- ness of the different regions is found to be as follows: Almost insensitive, the surfaces of the pedal disc, the lips, and the oesophagus ; slightly sensitive, the surface of the column between the sphincter and the oral disc, the oral disc between the tentacles and the lips, and the siphono- glyphs; slightly more sensitive, the tentacles and the equatorial portion of the column; fairly sensitive, the surface of the column near the sphincter ; and most sensi- tive, the surface of the column near its pedal margin. Stimulation of any of these regions was followed by a retraction of the oral disc due to a contraction of the longitudinal muscles of the mesenteries chiefly. The re- gions of stimulation, as the description implies, were al- ways in the ectoderm; the response was made by ento- dermic muscles. Hence this particular set of reactions was very appropriate as a means of testing the course of nervous transmission from ectoderm to entoderm. If the column wall of a sea-anemone is cut through in a complete ring equatorially, that is, if the column is gir- NERVOUS TRANSMISSION 91 died (Fig. 21), a stimulus applied to the pedal edge of the column will call forth a contraction of the oral disc even more readily than when it is applied to the portion of the column that is on, the oral side of the circular in- cision. Hence it must be admitted that there are trans- mission tracts that lead from the ectoderm of the pedal edge of the column directly through to the longitudinal muscles of the mes- enteries irrespective of such con- nections as may exist in the oesophagus. The same conclusion can be drawn from what is seen in prep- arations from which the whole oral disc has been cut off. On stimulating the pedal edge of such a preparation the portions of the longitudinal muscles of the mesenteries still remaining in the animal contract vigor- ously, showing that there is not only a direct connection between the ectoderm of the pedal edge of the column and the longitudinal muscles of the mesenteries, but that the oral disc, believed by the Hertwigs to contain a central- ized portion of the nervous system of the actinian, is in no way essential to the reaction noted. The belief that the oral disc does not contain an essential nervous center has already been vigorously set forth by Jordan (1908). If the column of a large sea-anemone with a pedal disc 10 centimeters or more in diameter is cut through in an oblong outline, 4 centimeters by 2 centimeters, a super- ficial piece of the column results that is attached to the rest of the animal only through its mesenteries (Fig. 22). FIG. 21. — Metndium with its column wall completely girdled by a cut that passes at all points en- tirely through the wall. Stimulus applied at z. 92 THE ELEMENTARY NERVOUS SYSTEM Nevertheless, when the middle of this piece is stimulated mechanically or by discharging on it a small amount of hydrochloric acid in sea- water, a withdrawal of the whole oral disc follows. This response ceases when all the organic connections of the piece with the rest of the ani- mal are severed by cutting through the attached mesen- teries, thus allowing the piece Fio. 22.— Metridium in whose column simplv to He ill plaC6 OH the wall an oblong incision, 4 cm. by 2 cm., a\\±:dTo1ne9:n?maaitoniybUytih!reC8een! sea-anemone. The cessation teries. stimulus applied at x. Of reSp0nse under these cir- cumstances shows that the transmission from the surface to the muscle must have been nervous and that it was not due either to the mechanical effects of the contraction of the piece itself on the deeper tissue, or to an accidental overflow of acid. Thus this experiment, like the two preceding ones, demonstrates a direct nervous con- nection between the ectodermic sensory apparatus of the col- umn wall and the longitudinal muscles of the mesenteries. If a sea-anemone is cut in two vertically in such a way that the resulting halves remain attached only by the lips (Fig. 23), not even the oesophagus or the oral disc remaining intact, and Fio. 23. — Metridium cut vertically in two except in the region of the lips. Stimulus applied at z. NERVOUS TRANSMISSION 93 if a mechanical stimulus is then applied to the column of one half, the portion of the oral disc in that half will be regularly withdrawn while that in the other half will be at most only rarely moved. Hence it follows that the region of the mouth, particularly of the lips, must be regarded as one poor in nervous connections. Conse- quently the view originally advocated by the Hertwigs and later by Wolff (1904) and by Groselj (1909) to the effect that the region of the mouth is the chief region of nervous connection between the ectoderm and entoderm is not supported by direct observation, for, as has just been shown, this region is one poor in its capacity for nervous transmission and, as was shown earlier, there are abundant direct connections between the ectodermic receptors and the entodermic muscles without recourse to such areas as the region of the mouth and the oesophagus. The conducting paths in the nervous organization of a sea-anemone can be dem- onstrated in a variety of ways. If a tongue of tissue is cut equatorially from the column of a sea-anemone so as to girdle it for half its circumference (Fig. 24), and a stimulus is applied to the free end of the tongue, such stimulus is only occa- sionally followed by a retraction of the disc, showing that this least sensitive portion of the column is not in FIG. 24. — Metridium from which an equa- torial tongue has been cut from the column. Stimulus applied at x. Fio. 25. — Metridium from which a tongue running from the pedal disc to the equator of the column has been cut. Stimulus ap- plied at x. 94 THE ELEMENTARY NERVOUS SYSTEM what would be called free nervous communication hor- izontally with the rest of the column. If, however, a tongue of the column wall is cut from the pedal edge of the col- umn up to its equatorial region and there left in organic connection with the rest of the animal (Fig. 25) and a mechan- ical stimulus is applied to the free end of this tongue, a response of the whole system of longitu- dinal mesenterio muscles commonly follows. If a similar tongue is cut from the oral edge of the column down to its equator, a stimulus applied to the free end of this tongue will again call forth a retraction of the oral disc. If the pedal edge of the column is cut off by an in- cision parallel to this edge and about 3 millimeters inside it, thus producing a band of tissue 4 to 5 centi- meters long and attached by only one end to the ani- mal (Fig. 26), a mechani- cal stimulus applied to the free end of this band is quickly followed by the retraction of the oral disc. If this band is anywhere completely cut across, a stimulus distal FIG. 26. — AlrtriJivm, seen from the oral pole, from which a portion of the pedal e Ige of the column has been partly cut off. Stim- ulus applied at x. NERVOUS TRANSMISSION 95 to the cut is never followed by a retraction of the disc even though the two faces of the cut are in contact. If a sea-anemone is cut through vertically except for its oral disc (Fig. 27), the mechanical stimulation of the column of one part is followed by a contraction of the longitudinal mesenteric muscles of both parts, thus dem- Fio. 27. — Metridium cut through verti- cally except for its oral disc. Stimulus ap- plied at x. FIG. 28. — Metridium cut through verti- cally except for its aboral disc. Stimulus applied at x. onstrating transverse nervous connections in the oral disc. If the sea-anemone is cut through vertically as de- scribed in the preceding paragraph, except that the con- necting portion is the pedal disc instead of the oral disc (Fig. 28), the stimulation of one part is followed by the retraction of the halves of the oral disc in both parts. Thus the pedal disc as well as the oral disc can serve the animal for transverse nervous connections. Finally, if a sea-anemone is cut vertically in two, ex- cept for a small connecting bridge near the pedal edge of the column (Fig. 29), the mechanical stimulation of the column of one part is followed by a retraction of the halves of the oral disc in both parts. These various tests and experiments show that the 96 THE ELEMENTARY NERVOUS SYSTEM longitudinal mesenteric muscles, by whose contraction the oral disc is depressed, may be called into action from almost any part of the outer surface of the sea-anemone. Paths of conduction must, therefore, exist between prac- tically every point of the exterior of the animal and the longitudinal muscles of the mesenteries, and for these paths of conduction the oral disc and the oesophagus are in no way essential. In fact, definite and circumscribed paths appear not to exist, but, as the directions of the vari- ous incisions and cuts FIG. 29. — Metridium, seen from the oral shOW, it is nCCCSSary to aS~ pole, cut through vertically except for a email connecting bridge on the pedal edge SUniC, first, an immediate of the column. Stimulus applied at z. connection between the superficial ectoderm and the deep-lying musculature, and, next, a veritable nervous network by which all manner of irregular and complicated incisions can be circumvented so long as a bridge of the natural tissue remains as a means of connection between the part stimulated and the part responding. In other words, some of the most char- acteristic reactions of sea-anemones are to be ascribed to the fact that these animals possess a well-differentiated nerve-net (Parker, 1917 a}. The location of this nerve-net is by no means clearly determined. At first thought it might be expected to occur in the nervous sublayer described in the ectoderm and in the entoderm by the Hertwigs. But doubt has already been expressed as to the nervous character of this layer (Parker, 1912). In Metridium the so-called nervous sub- NERVOUS TRANSMISSION 97 layer of the ectoderm is best developed in the region of the lips, but, as experimental studies have shown, this is precisely the part of the animal that is poorest in nervous conduction. It, therefore, seems probable that, though the so-called nervous sublayer of the Hertwigs must be penetrated by a host of neurofibrils from the overlying sensory epithelium, this sublayer must have some other function than that of nervous conduction, which must be carried out by another part of the animal. The part most probably concerned with nervous conduction, judging from the discoveries of the last few years, is the support- ing lamella. This layer has long been known to contain a network of cells and from the work of Havet (1901) and of Parker and Titus (1916) the richness of its nervous con- nections can no longer be doubted. It would, therefore, not be surprising if future investigation should show that the chief mass of the actinian nerve-net should be found to be located in the supporting lamella and not in the so- called nervous sublayers of the ectoderm and the entoderm. The rate at which the nerve-net of actinians transmits impulses has been determined for Metridium (Parker, 1918 b). When large specimens of this sea-anemone are prepared as shown in Fig. 26, and the stimulus is applied at the distal end of the tongue of tissue cut from the pedal edge, the longitudinal mesenteric muscles contract vigorously. If now the stimulus is applied to the root of the tongue, the muscular contraction occurs in less time than in the former instance. The difference between these two periods of time thus determined is the interval neces- sary for transmission between the two points stimulated. As the distance between these two points can be measured with accuracy, the rate of transmission can be easily de- 7 98 THE ELEMENTARY NERVOUS SYSTEM termined. At 21 degrees centigrade this proved to be from 121 millimeters to 146 millimeters a second, a rela- tively slow rate. In the particular form of preparation on which this determination was made, the tissue of the tongue exhibited no contraction during transmission, hence it is to be concluded that the type of transmission here measured was strictly nervous, and not muscular or a combination of nervous and muscular transmission. The type of stimulation and response that has been described in the last few paragraphs for the sea-anemone is one of the most extensive and vigorous that this animal can carry out. If the stimulation is sharp and prolonged not only do the longitudinal mesenteric muscles continue to contract, but the tentacles shorten and contract, the circular muscle of the column aids in diminishing the volume of the animal, and, as the water from its interior is discharged, the sphincter, like a puckering string, draws the column over the retreating oral disc till the whole animal is changed from a beautifully expanded flower-like form to a shrivelled, contracted, and almost amorphous mass. In this extreme condition probably all the thirteen or more sets of muscles in the actinian's body have passed into a state of extreme contraction, which justifies more or less the older conception of the sea- anemone as a simple muscular sac. This state of con- traction, a condition of extreme muscle tonus, may en- dure for hours or even days and illustrates well the chief characteristic of the musculature of many of these sim- ple animals. In these forms the musculature is not adapted to the quick and varied movement that we asso- ciated with it in the higher animals, but it exhibits a con- dition of prolonged contraction, a state of extended tonus, such as is quite inconceivable in the skeletal muscles of a NERVOUS TRANSMISSION 99 vertebrate or an arthropod. In fact, the condition in the higher animals that seems most nearly to resemble this state of profound tonus in sea-anemones is that which fol- lows strychnine poisoning or the poisoning from the bacil- lus of tetanus, a muscular spasm which, however, often results in death. The recovery of a sea-anemone from this state of ex- cessive muscular tonus is slow and gradual. It depends upon a reduction of the tonus, which exhibits itself in the form of a general relaxation of the musculature followed by a slow filling of the digestive cavity of the animal with seawater whereby it eventually assumes its distended form. This is carried out by the action of the ciliated siphonoglyphs through which currents of water are led into the interior of the animal. The process of recovery is, therefore, a relatively slow one as compared with the operation of contraction by which the water under rather high pressure is literally squirted from the animal's body. Thus the character of this response illustrates well the nature of the nerve-net, for it shows that from any single spot on the surface of the actinian's body practically its whole musculature may be brought into excessive but normal contraction. Is it possible through this net to affect one set of muscles rather than another or is the net an open conducting system leading as freely in one direc- tion as another? Theoretically such a diffuse condition has for some time been assumed, but it is questionable whether in even the simplest nerve-nets such an undiffer- entiated state really exists. The method of stimulation used in the preceding experiments, that by a fine glass rod, though apparently delicate, is in reality most harsh and unnatural as compared with normal stimuli for such ani- 100 THE ELEMENTARY NERVOUS SYSTEM mals. On applying suck a rod to the pedal edge of the column of a sea-anemone the oral disc as a whole is de- pressed, an operation that involves the simultaneous ac- tion, as far as can be seen, of all the longitudinal muscles in the numerous complete and incomplete mesenteries of the animal. This means that scores and perhaps hun- dreds of these muscles act in unison. Is the nerve-net capable of only this gross form of activity or does it rep- resent a system of finer gradations by which under an appropriate stimulus a part of this musculature can be excited to specific action while the rest remains essen- tially quiescent? An answer to this question can be seen in the follow- ing experiment. If a Metridium is allowed to remain for some time in running seawater in a situation relatively dark, its muscular tonus will be reduced to a minimum and it will assume the condition of fullest normal expansion. If, under such circumstances, it is generally and briefly illuminated, it will quickly shorten its length quite notice- ably though it will by no means go into what would be described as a state of contraction. This shortening of the animal as a whole is due to the simultaneous moderate contraction of its longitudinal mesenteric muscles. The fact that the shortening is symmetrical and uniform shows that a complete ring of these muscles have contracted in unison. If, instead of subjecting the fully expanded sea- anemone to a general illumination, light is thrown on only one of its sides, it responds usually by turning its oral disc toward the light, precisely as some flowers come to face the light. In the sea-anemone this is due to the contraction of those longitudinal mesenteric muscles that lie on the side illuminated, in consequence of which that side shortens and the oral disc is tilted in that direction. Thus NERVOUS TRANSMISSION 101 light, unlike the glass rod, can so stimulate the surface of an actinian that the subjacent nerve-net will call into action only that group of the longitudinal mesenteric muscles that lies close at hand and hence the nerve-net exhibits under a more normal form of stimulation a type of response much more delicate in character than what is seen when a glass rod is used. Probably these two types of responses are merely extremes of the same thing and differ only quantitatively ; but, however this may be, they show quite clearly that if the mechanical stimulus from such a source as a glass rod could be sufficiently refined there is not the least reason to suppose that it could not be made to call forth as finely graded a response as light does. At all events the nerve-net under these circum- stances does not appear to possess that property which is coming to be so commonly recognized in the nerve and muscle of the higher animals and which is exemplified under the head of the "all or none ' ' principle. * CHAPTER VIII JELLYFISHES THE ccelenterates whose neuromuscular organization has been most considerably studied are not the sea-anem- ones, but the jellyfishes. These animals belong either to the class Eydrozoa or Scypliozoa. The hydrozoan or craspedote jellyfishes are relatively small, even micro- scopic animals, characterized by the presence of a thin membrane, the veil, which extends from the edge of the bell partly across its open mouth. Because of their small size they are not very favorable for experimental work, and the best known from this standpoint are the larger forms, such as Carmarina and Gonionemus. The scypho- zoan or acraspedote jellyfishes on the other hand are mostly large, sometimes very large, animals and conse- quently much more experimental work has been done on these than on the craspedote forms. They are charac- terized by a freely open bell without veil. They include such well-known representatives as Pelagia, Cyanea, Aurelia, Cassiopea, and Rhizostoma. The neuromuscular system of a jellyfish is limited al- most entirely to the edge and concave or subumbrellar surface of the bell. In Aurelia, for instance, the edge of the bell has upon it at regular intervals eight groups of sense organs, the marginal bodies. Each group includes an eye-spot, a static organ, and several other organs prob- ably of the chemical sense. These receptors represent specialized and thickened portions of the superficial epi- thelium from which bands of nervous tissue spread cen- 102 JELLYFISHES 103 trally over the subumbrellar surface to the region of the large circular sheet of muscle that forms a sphincter-like organ midway between the centrally located mouth and the edge of the bell. It is the contraction of this muscular band that reduces the cavity of the bell and, by thus driv- ing the water out of this cavity, forces the animal for- ward. On the relaxation of this sphincter, the bell re- sumes its expanded form in consequence of the elastic action of its gelatinous tissue. The muscle libers in the great majority of coelenterates are like the smooth or non-striped variety in the higher animals, but the sphinc- ters of jellyfishes are exceptional in this respect and are composed of cross-striped fibers. Not only are they thus structurally exceptional, but they are also unlike the or- dinary coelenterate muscle in that they are incapable of long-continued tonic contraction (Jordan, 1912) and in this respect they also resemble the cross-striped muscle of higher forms. As already stated, each of the eight marginal bodies of Aurelia consists of a group of sense organs whose cells are a modified part of the superficial epithelium of the region. From the deep ends of these cells nervous prolongations reach out from the given sense organ toward the sphincter. The stretches of nervous tissue that thus emerge from each marginal body are made up not only of the basal prolongations of sense cells, but of many interpolated nerve cells whose form is usually that of a bipolar cell. In Aurelia these stretches of nervous tissue spread out almost at once on the plate-like sphinc- ter, but in Rhisostoma, according to Hesse (1895), the nervous tissue from each of the eight marginal bodies forms a fairly well circumscribed band, which retains its radial course from near the periphery of the bell till it 104 THE ELEMENTARY NERVOUS SYSTEM reaches the much more centrally located sphincter. In the region of the sphincter of Aurelia, Rhizostoma, and other such jellyfishes, these nervous bands spread out diffusely over the whole muscle layer, producing thus a nervous layer more superficial in position than the muscle on which it rests. The earlier workers were not able to state with cer- tainty the mutual relations of the elements in this nervous layer. The prevailing opinion was that the layer was composed of bipolar nerve cells. Eimer (1878) described the processes from these cells as branched and stated that the branches from various cells united to form a net- work, a state of affairs that Schafer (1879) was unable to confirm. Nor does the account of the nervous system of Rhizostoma, as given by Hesse (1895), lead to the con- clusion that a true network is present. With improved histological technique Bethe (1903, 1909), however, showed that the nerve plexus spreading over the sphinc- ter of Rhizostoma is a true nerve-net in which the proc- esses from the various nerve cells are continuous. This nerve-net overspreads the surface of the sphincter as a network, in which the processes from the cell-bodies, though limited chiefly to a plane parallel with that of the sheet of muscle, are not all thus restricted, but in many cases may be traced, on the one hand, distally in among the surface epithelial cells and, on the other hand, prox- imally among the muscle cells. In respect to their direc- tions the nerve processes are very unlike the underlying muscle fibers for while the nerve processes extend in a great variety of directions, the muscle fibers follow a most rigid arrangement and take a course circular in outline and concentric with that of the jellyfish as a whole. Thus on the outer surface of a very regularly arranged system JELLYFISHES 105 of muscle fibers is a diffuse and branching network of nerve cells. A clear idea of the mutual relations of the epithelium, nerve-net, and muscle layer in such a jellyfish as Rhizo- stoma can be obtained from a section across its sphincter at right angles to the contained muscle fibers (Fig. 30). Such a section would of course have a radial position in the jellyfish as a whole. On its outer face is the layer of epi- thelial cells that covers the sub- umbrellar surface of the jelly- fish. Next within is the nerve- net with its contained nerve cells and branching fibers, some of which can be traced in among the epithelial cells and others among the musole fibers. Deepest of all is the band of muscle fibers of the sphincter muscle cut at right angles and penetrated more or less by nerve fibers from the nerve- net. This sequence of tissues from the outer surface inward is the same as that met with in sea-anemones, in which the epithelium is most superficial, the nerve-net next, and the muscles deepest. Although the conditions thus far described are chiefly those in the acraspedote medusae, they reproduce in essentials the structure of the craspedote forms except in two particulars. In the cras- pedote jellyfishes there are no well-defined marginal bodies, but the sense organs of the edge of the bell are more diffuse in their arrangement. Further, the cras- Fia. 30. — Section at right angles to the sphincter of the bell of Rhizostoma; the subumbrellar surface is uppermost; e, epithelium of the subumbrellar sur- face; n, nervous layer; m, muscle layer. (After Bethe, 1903.) 106 THE ELEMENTARY NERVOUS SYSTEM pedote medusae, unlike those without the veil, are com- monly characterized by a double band of nervous tissue encircling the margin of their bells. On the other hand, the subumbrellar nerve-nets and muscle bands are essen- tially alike in both the craspedote and the acraspedote forms. The physiology of the neuromuscular mechanism of jellyfishes was independently investigated by Eimer and by Romanes. In 1873 Eimer presented before the Physi- kalisch-medicmische Gesellschaft in Wiirzburg a prelim- inary communication on his researches in this field, a pub- lished statement of which appeared in 1874, the year in which Romanes published his first brief note on the sub- ject. Romanes' two extended papers appeared, one in 1877 and the other in 1878, the year in which Eimer 's monograph came from press. The results of these two investigators were in essential agreement. In the main they entirely justified the view subsequently worked out by histological means that jellyfishes possess on their subumbrellar surfaces an intricate nerve-net. It was found that if the eight marginal bodies of such a jellyfish as Aurelia were excised, the pulsations of the bell, for the time being at least, ceased. If all but one were excised, the pulsations continued and emanated from the one remaining marginal body. Such an animal formed a very convenient preparation on which to study the course taken by the impulse to contraction. Both Romanes and Eimer showed that the bell of a jellyfish could be cut into a most complex pattern without interfering with the passage of the wave of contraction throughout its whole extent. If the single marginal body on a prepared bell of Aurelia is stimulated, a contraction appears in its immediate vicinity and spreads as two JELLYFISHES 107 waves, one to the right and the other to the left, around the bell till they fuse and become obliterated in a region opposite that from which they started. If such a bell is cut into a long spiral strip with the one remaining mar- ginal body at one end and the center of the bell at the other, a wave of contraction can be started from the mar- ginal body and will progress ordinarily over the whole FIG. 31. — Diagram of the jellyfish Aurelia from which the central mass and seven of the eight marginal bodies have been removed. The outline of the bell has been further compli- cated by a system of interdigitated cuts. (After Romanes, 1893.) length of the spiral. A much more complex preparation may be made by cutting out the center of a jellyfish bell, and then complicating it by a series of incomplete radial -incisions in alternation, one set starting from the inner edge of the bell and the other from the outer edge (Fig. 31). A bell thus complicated by so intricate a series of interdigitations will nevertheless transmit a wave of con- traction completely around its tortuous length. In fact, secondary interdigitations may be cut on the lines of any 108 THE ELEMENTARY NERVOUS SYSTEM of these primary cuts without necessarily interrupting the passage of the wave. Another extreme 'tost mav he car- ried out in which the jellyfish is reduced to two concen- tric rings attached to each other by only a small bridge of connecting tissue (Fig. 32). This type of preparation can be made by excising the center of a jellyfish and then Fio. 32. — Diagram of the jellyfish Aurelia from which the central mass and seven of the eight marginal bodies have been removed. The outline of the bell has been further complicated by a circular incision that has nearly divided the bell into two rings. The arrows show the course of the contraction wave as it emerges from the one remaining marginal body and passes over the bell. (After Romanes, 1893.) reducing it further by a circular cut parallel to its outer edge and midway between this edge and the center. The circular cut is so extended as to form a complete circle except for a fraction of an inch over which the outer ring is in organic connection with the inner one. When a wave of contraction is started from the one remaining mar- ginal body on the outer ring, it will, on arriving at the bridge, pass across to the inner one irrespective of the region in the bell at which the bridge may be located, thus JELLYFISHES 109 demonstrating how small an amount of tissue is neces- sary for the transmission of a wave. In Cassiopea, Mayer (1906, 1908) has shown that the two waves that emerge from the single marginal body are often of very different sizes. When two such dispro- portionate disturbances meet on the opposite side of the bell, the stronger wave commonly overcomes the weaker one and instead of having both disappear the stronger one continues its course around the bell. Such a wave once started may course indefinitely round and round a bell. This is particularly true if the bell has been so cut as to produce a very long circular stretch so that after the passage of the wave over a given portion of the stretch that portion has ample time to recover before the wave again passes over it. Harvey -(1912) found that such a "trapped" wave may course over a circular path for as long as eleven days with no appreciable decline in rate. As this rate in Cassiopea averaged 46,472 millimeters per minute, the wave must have travelled in the eleven days somewhat over 457 miles. As Eimer, Bomanes, Mayer, and others have shown, jellyfish bells may be cut into patterns of the utmost di- versity without, however, preventing the passage of their waves of contraction. As the majority of these patterns interrupt completely the course of the muscle fibers, but would always be circumvented by the nerve-net, it is gen- erally assumed that transmission over the complicated outline to which the body of the jellyfish has been reduced must be dependent upon the nerve-net, not the muscles. More conclusive evidence of the significance of the nerve-net in this respect, as contrasted with the muscle layer, has been obtained from several lines of work. Bethe (1903, 1909) has pointed out that in Rhizostoma the 110 THE ELEMENTARY NERVOUS SYSTEM sphincter muscle and the nerve-net are not coextensive. In Aurelia the sphincter forms a continuous circular band. In RTiizostoma, on the other hand, it is composed of a suc- cession of sixteen muscle fields, a pair between each pair of the eight marginal bodies (Fig. 33). These muscle fields are partly separated from each other by areas of interven- FIG. 33. — Diagram of a portion of the bell of Rhieostoma seen from the oral side; m, muscle field; n, area covered by the nerve-net but devoid of muscle. (Modified from Bet he. 1909.) ing non-muscular tissue over which, however, the nerve- net extends. Consequently, it is possible to cut prepara- tions from Rliizostoma in which one muscle field with an attached marginal body is connected with its neighbor only by an area of nerve-net (Fig. 34). When such a prepara- tion is stimulated through its marginal body, not only does the immediately adjacent muscular field contract but also the more distant one. It is, therefore, clear that in this case transmission from one muscle field to the next is accomplished by the nerve-net and not by the muscle fibers. Another piece of evidence bearing on this question has been advanced by Mayer (1906, 1908). Cassiopea is a jellyfish that possesses considerable powers of regenera- JELLYFISHES 111 FIG. 34. — Preparation of two muscle fields from Rhizo- stoma to one of which a mar- ginal body is attached; the preparation has been cut eo that the second field is attach- ed to the first, so far as its neuromuscular relations are concerned, only through the nerve-net, n. tion. Wlien the superficial tissue on its subumbrellar surface, epithelium, nerve-net, and muscle, is scraped off over a given area, the regenerative recovery is of such a kind that the epithelium and nerve-net appear soon and the muscle layer somewhat later. It is, therefore, possible to obtain in Cassiopea regenerating individuals in which the epithelium and nerve- net have been re-formed over the wound, but the muscles have not yet appeared. Such individuals may be cut into preparations in which two masses of muscle are connected only by a regenerated epithelium and nerve-net. When one muscle field in such a preparation is made to contract the other field also quickly contracts without the appearance of any movement in the intervening region. Hence it must be admitted that this intervening region, composed merely of epi- thelium and nerve-net, has the capacity of transmitting impulses to motion from, one of the muscle fields to the other. This experiment, like that of Bethe, supports the belief that transmission over the subumbrellar sur- face of jellyfishes is accomplished by the nerve-net and not by the muscle. The removal of all the marginal bodies from an acras- pedote jellyfish is an effective way of rendering the bell at least for the time being motionless. The same kind of an operation can be carried out on a craspedote form by cutting off the edge of the bell. In such cases the edge usually carries with it some muscle tissue and will con- tinue to pulsate for a long time, whereas the central por- 112 THE ELEMENTARY NERVOUS SYSTEM tion of the bell is quiescent. Loeb (1900) has shown that if the central portion of the bell of the craspedote medusa Gonionemus is thus rendered quiescent, its activity may be revived by placing it in a solution of % molecular so- dium chloride, which is about isotonic with seawater. If to this solution a small amount of calcium chloride or potassium chloride is then added, the pulsations cease as in ordinary seawater, for the solution is thereby rendered a balanced one, as seawater naturally is, in that the stim- ulating effects of the sodium are counteracted by the cal- cium or the potassium. Such a quiescent edgeless bell has upon its subum- brellar surface a nerve-net and a muscle layer. Whether in Gonionemus the sodium solution stimulates one or both of these has not been definitely determined, but in Rhizo- stoma a partial answer to this question has been found by Bethe (1909). When the eight marginal bodies of this jellyfish are removed, its bell is quiescent in ordinary sea- water. Under such circumstances, however, it can be made to contract by stimulating it electrically or chem- ically. If now a preparation is cut from a Rhizostoma so as to include one marginal body and two muscle fields, the latter separated from each other by a partial cleft that allows the two lobes thus formed to be in physiological connection only through their nerve-net (Fig. 35), it is possible to determine the relative sensitiveness of the marginal body as compared with that of the rest of the neuromuscular mechanism. When in such a preparation the lobe without the marginal body is immersed in a 0.62 molecular solution of sodium chloride and the lobe with this body is in seawater, the rate of contraction of the whole preparation remains constant and normal. If, how- ever, the lobe with the marginal body is now immersed in JELLYFISHES 113 the solution of sodium chloride and the other lobe is in seawater, the rate of contraction quickly increases to about three times what it formerly was. This shows that the particular concentration of sodium chloride used, 0.62 molecular, was stimulating for the marginal body but not for the nerve-net or the muscle. If a preparation that FIQ. 35. — Diagram of a preparation of two muscle fields from Rhizostoma, one with an attached marginal body and the other without such a body and so arranged that the two fields may be partly immersed in vessels containing different solutions. Thus the sensitive- ness of the marginal body may be compared with that of the rest of the neuromuacular structure. (Based on the description by Bethe, 1909.) consists merely of nerve-net and muscle and that will not respond to a 0.62 to 0.66 molecular solution of sodium chloride, is now subjected to an 0.8 to full molecular so- lution of the same salt, contractions will reappear for a short time at least, showing that the combination of nerve- net and muscle, though not so sensitive as the marginal body, is nevertheless open to stimulation. When a preparation in which there is only nerve-net and muscle is immersed in a sodium chloride solution, it is claimed by Bethe that both the region in which there is only nerve-net and that in which there is both nerve-net and muscle become more open to stimulation and that of 8 114 THE ELEMENTARY NERVOUS SYSTEM the two the region of the nerve-net alone is rendered rela- tively more sensitive than that composed of nerve-net and muscle. Bethe, therefore, concluded that, though the muscle may be open to direct stimulation, it is not so sen- sitive as the nerve-net and that consequently, when a por- tion of a Rliizostoma without a marginal body is made to contract, it probably does so through the action of the stimulating agent on the nerve-net rather than on the muscle. From this standpoint the neuromuscular mechan- ism of a jellyfish may be regarded as composed of three parts of graded susceptibility to stimulation: the mar- ginal bodies, which are most sensitive and are stimulated by ordinary environmental changes ; the nerve-net, which is less sensitive than the marginal bodies and is only rarely open to stimulation from the environment; and, finally, the muscle, which is the least sensitive of the three and is consequently only most rarely stimulated directly by environmental changes. This contrast in the sensitivity of muscular and nerv- ous components in the neuromuscular organization of the jellyfishes is quite in accord with what should be expected in the evolution of a receptor-effector system. The muscles of sponges are, in comparison with those in most other animals, extremely inert to stimulation and very sluggish in response. In this respect they are like the muscle of the coelenterate divested of its nervous connec- tions. It is only after such effectors as these have added to them receptors, such as are seen in the sensory sur- faces of the sea-anemones or in the marginal bodies of the jellyfishes, that a quicker and more efficient system is established, whose improvement over the old one is de- pendent upon the increased sensitiveness of the new mem- ber, the receptor, rather than on any considerable change in the original member, the effector. CHAPTER IX \ THE NEEVE-NET THE nerve-net of the lower animals contains the germ out of which has grown the central nervous systems of the higher forms. As a definite type of structure the nerve-net has been recognized only for a few years. Its discovery was brought about by the repeated attempts made since the declaration of the cell theory to resolve nervous tissue into its component cells. Although the cell theory as applied to animal tissues was enunciated by Schwann as early as 1839, it was not till more than half a century later that a clear and consistent idea of the nerve cell was arrived at. Nerve fibers seem to have been first really seen and figured by the Florentine physician Felix Fontana in 1781, but it was not till 1833 that Ehrenberg in the pre- liminary announcement of a monumental work on the fibrous structure of the central nervous organs, described certain corpuscles that proved to be what later investi- gators called ganglion cells. The connection of these two elements, vaguely intimated in 1838 by Eemak and sur- mised in 1840 by Hannover, was first really demonstrated for invertebrates in 1842 by Helmholtz and for verte- brates in 1844 by Kolliker, who showed that fibers with a medullary sheath, and therefore unquestionably nerv- ous, were directly connected with ganglion cells. From the time of these discoveries, it became necessary to assume that in some way or other ganglion cells were integral parts of the nervous system. 115 116 THE ELEMENTARY NERVOUS SYSTEM In 1855 Leydig discovered in the ganglionic bodies of spiders what appeared to be a finely granular material which he called punctate substance. Similar material was also shown to be a considerable constituent of the gray matter of the vertebrate nervous system. Hence, in addition to nerve fibers and nerve cells, a third kind of material was shown to be present in many nervous or- gans. This material, as was subsequently demonstrated in 1871 by Gerlach, consisted in reality of very fine fibrils which when seen in section appeared as minute points and hence Leydig 's name for it of punctate substance. The uncertainty of the relation of this fibrillar ma- terial to nerve fibers and to nerve cells was not removed until the Golgi method of silver impregnation began to be generally applied to nervous tissues. This method yielded such important results that in May, 1891, K61- liker could substantiate the claim that every nerve fiber in the body was at some part of its course directly con- nected with a ganglion cell, and in June of the same year Waldeyer, on the basis of conclusions drawn largely from Golgi preparations, promulgated the theory of the neu- rone, the first consistent account of the nerve cell. Ac- cording to this well-known doctrine the ganglion cell of the older workers is really the nucleated body of the true nerve cell, or neurone, whose processes are commonly of two kinds : fine protoplasmic processes from the cell body and nerve fibers with their final branchings. Both these kinds of processes may contribute to the formation of the finely fibrillar material already noted. This material was believed to be the means of intercommunication between neurones. The embryological investigations of His (1886) and other workers showed that the embryonic nerve cells, or THE NERVE-NET 117 neuroblasts, from which adult neurones were differen- tiated were in the beginning far separated from each other and only secondarily came together. Of their ini- tial separation there could not be the least doubt; the question that arose concerned the extent of their final union. On both anatomical and physiological grounds there were good reasons advanced to show that, though the separate neurones came into such contact relations as was necessary for the transmission of nervous im- pulses, they never fused to such an extent as to lose their identities. The fibrillar material, in which these inter- relations were known to occur, was, therefore, regarded, not as a continuous net, as Gerlach believed it to be, but as broken up into discrete neuronic systems separated one from another by an infinitude of minute interruptions, which, however, were capable of physiological continuity through what is known as a synapse. Thus each neurone, or true nerve cell, was believed to possess a certain amount of independence from its neighbors, though physiologi- cally united to them at least by transmitting contact. As the idea of the synaptic nervous system gradually unfolded itself to the more orthodox neurologists, there arose from another school of workers a very different con- ception of the interrelation of nervous elements. The impetus to this new movement came chiefly from the work of Apathy, who in 1897 maintained on the basis of histo- logical preparations of almost incredible clearness that the nervous elements of many animals were bound to- gether by a network of neurofibrils in which there was not the least evidence of interruption such as is implied in the synapse. This view was in a way a revival of the idea of a continuous network as maintained in a previous generation by Gerlach. The careful reader of Apathy's 118 THE ELEMENTARY NERVOUS SYSTEM papers will find it by no means easy to separate in them fact from speculation and, consequently, it is difficult to state in exact terms Apathy's real contribution to this subject ; but, however this may be, it is certainly true that the appearance of his publications excited others to a further investigation of the subject with the result that nerve-nets were proved to exist in a number of animals. As already stated, they were definitely identified by Bethe (1903) in jellyfishes, by Wolff (1904) and by Hadzi (1909) in hydro- zoans, and by Groselj (1909) in sea-anemones. In fact, the coelenterate nerv- Fio. 36. — Nerve network from a small j i blood-vessel in the .palate of the frog. OUS System Seemed LO DC (After Prentiss,.1904.) ., . ' , nothing but a nerve-net. Evidence was soon brought together to show that nerve-nets were at least components of the nervous sys- tems of echinoderms, worms, arthropods, molluscs, and even vertebrates, where they were especially asso- ciated with the digestive tract and the circulatory sys- tem (Fig. 36), including the heart. Thus nerve-nets were identified from the ccelenterates to the vertebrates, and some of the more ardent advocates of this type of nervous organization went so far as to assume that it was the only type of nervous structure really extant and that the evidence of a synaptic system rested upon his- tological artefacts that obscured the real relations of cell to cell. But this extreme position has not been justified by further research. It is now generally admitted that the conceptions of a synaptic system and of a nerve-net are not opposing ideas, but represent two types of nerv- THE NEEVE-NET 119 ous organization, both of which may exist side by side in the same animal. Judging from the fact that the nerve-net is apparently the exclusive type of nervous organization in the ccelenterates and that it becomes pro- gressively less and less evident the higher the animal series is ascended, it seems fair to conclude that the nerve- net is the more primitive type and that in the course of evolution it has given way more and more to the synaptic system which has finally come to be the dominating plan of nervous organization in the more complex animals. From this standpoint one of these animals might per- fectly well possess both types of nervous structure; nerve-nets having been retained in its more conservative portions and synaptic structures having been developed in its more progressive parts. Thus the nerve-net may be regarded as phylogenetically older than the synaptic sys- tem. If the cell unit of the synaptic system is called the neurone, it would not be inappropriate to designate the more primitive units of the nerve-net as protoneurones (Parker, 1918 a). Since the nervous system of the coelenterates consists very largely of nerve-nets, the activities of these ani- mals must reflect in a general way the peculiarities of this type of nervous organization. As the examples that have been given show, the nerve-net is almost always inti- mately associated with the muscles which it excites to action. Commonly the nerve-net is directly superimposed on the muscle layer or the two tissues may be more or less commingled. In either case the activity of the muscle may have a considerable influence on the state of the nerve-net, stretching it or squeezing it or in one way or another changing its physical condition. This relation has been made the basis of some important speculations 120 THE ELEMENTARY NERVOUS SYSTEM by von Uexkiill (1909) as to the physiology of nerve- nets, namely that in transmission impulses flow more freely into regions where the net is stretched than into those where it is not. Attention will be called to examples of this kind in subsequent chapters. As a result of the intimate relation usually existing between the nerve-net and the muscles that it controls, most organs that are provided with this type of neuro- muscular organization exhibit an extreme degree of au- tonomy. This is perhaps one of the most striking features associated with the nerve-net. It is well illustrated by such an organ as the tentacle of the sea-anemone, the au- tonomy of which was long ago recognized by von Heider (1879). In most actinians the tentacles are hollow finger-like projections from the oral disc. Their cavities communi- cate freely with the common digestive cavity of the polyp and their walls consist of an outer layer of ectoderm sep- arated by a supporting lamella from an inner layer of entoderm. In experimenting with tentacles it is desir- able that they should be of large size, and this is the case in Condylactis, whose tentacles may be from 12 to 15 cen- timeters long and 1.5 centimeters in diameter. In this particular sea-anemone the tentacle terminates in a blunt end provided with a pore, and, as in the tentacle of Metridium, there are a longitudinal ectodermic muscle and a circular entodermic one. If an expanded quiescent tentacle of Condylactis is touched near its tip witli a silver sound or a glass rod, the tentacle contracts, usually bending toward the stim- ulated side. The tentacle often sticks to the object with which it is touched and it may in contracting thus exert a considerable pull, showing that its surface is remark- ably adhesive and that its musculature is vigorous. If a THE NERVE-NET 121 piece of crab meat or fish flesh is brought into contact with a tentacle, it adheres firmly to the tentacle which quickly shortens and is usually covered by several adjacent ten- tacles. The piece of flesh is thus held on the oral disc of the animal, which contracts in such a way as to move the mouth gradually toward the food till this aperture is near enough to swallow the- flesh. The animal then slowly returns to a state of quiescent expansion. When a tentacle of Condylactis is cut off and allowed to float in seawater, it contracts to about one-third its former length and remains almost indefinitely in this condition. Relaxation apparently never fully returns. If such a tentacle is caught on its cut edge by a delicate metal hook, it may be held with its open basal end at the surface of the seawater and inflated by gently running seawater into it. In this way a severed tentacle may be made to expand to about two-thirds its normal length and, under these circumstances, it will exhibit just about that degree of distension and mobility as is seen in the attached tentacles. If now, more seawater is discharged into it, it is likely to elongate a little and then contract considerably, discharging much of its contained water. Obviously for experimental purposes the fairly distended tentacle offers the most favorable condition. Tentacles suspended in seawater and expanded to about two-thirds their natural length remain quiescent for considerable periods. From time to time, however, they show spontaneous movements consisting of slight contractions and twistings by which more or less of their contained fluid is discharged. If this is replaced, they will reexpand and thus periods of quiescence are followed by periods of activity. In these respects the severed ten- 122 THE ELEMENTARY NERVOUS SYSTEM tacles reproduce very closely the behavior of the nor- mally attached tentacles. In response to slight or to vigorous mechanical stimuli the severed tentacles reproduce in a most striking way the movements of the attached ones under like stimula- tion. When an attached tentacle is gently touched on one side midway its length, the tentacle as a whole con- tracts but without much bending. If it is touched on the tip, the response is mostly a terminal waving back and forth. If it is stimulated on one side near the base, the contraction is chiefly basal and on the stimulated side. These responses are reproduced quite clearly by isolated tentacles. Thus the responses of the two classes of ten- tacles to localized stimuli are strikingly similar. If a small amount of a 1 per cent, solution of acetic acid is discharged on an expanded severed tentacle, the tentacle contracts quickly with a curious appearance as though it were withering. After it has been washed with seawater, it may be distended again in about three to four minutes. A second and a third response have been elicited from such tentacles and these responses repro- duce most strikingly the movements of attached tentacles. To a tenth per cent, solution of acetic acid both classes of tentacles show a slight local shortening. To a hun- dredth per cent, solution they respond by a slight curv- ing. To a thousandth per cent, solution neither kind of tentacle shows any response whatever, as is also the case when pure seawater is discharged on them from a pipette. To seawater discharged on the tentacles from a small pipette, no noticeable response is made by either class, but to seawater containing the juice from a crushed mus- sel, the attached tentacle exhibits active writhings often THE NERVE-NET 123 accompanied by elongation. It is remarkable how strik- ingly similar to these are the responses of the isolated tentacles to the same juice. To a 1 per cent, solution of quinine hydrochloride in fresh water both classes of ten- tacles respond by quick contractions and often local con- strictions. In all the tentacular reactions studied the responses of the isolated tentacles agree most strikingly with those of the normally attached tentacles. Of course, the reac- tions of the isolated tentacles are not exact duplicates of those of the attached ones. They are feebler and less precise, but aside from differences such as these, which are quite clearly of an operative origin, the severed ten- tacle reproduces in a most striking way the responses of the attached tentacle (Parker, 1917 fc). The tentacles of Metridium are very small in com- parison with those of Condylactis and in consequence de- tailed tests cannot be carried out on them with the ease and certainty that they can on larger forms. Neverthe- less their general reactions are indicative of great au- tonomy. If the juices from fish flesh and other like food are discharged on the tentacles of an expanded Metrid- ium, these organs will exhibit first a considerable amount of irregular movement and finally point toward the mouth, where in fact under ordinary conditions they would de- liver the food itself. If, now, a tentacle is carefully cut off from Metridium, its original orientation in reference to the animal as a whole being kept in mind, the appli- cation of the stimulating juice will cause it, first, to un- dergo irregular movements and then to point in that di- rection that was originally toward the mouth. It is, therefore, evident that the tentacle of Metridium, like that of Condylactis, has within its own structure the neuro- 124 THE ELEMENTARY NERVOUS SYSTEM muscular mechanism necessary for carrying out a food response such as has been described (Parker, 1896). This autonomy of organs is nowhere better illustrated than in the foot or pedal disc of the sea-anemones. Prob- ably all actinians in which there is a well-differentiated pedal disc have some powers of locomotion. These are relatively slight and not often exercised in such a sea- anemone as Metridium marginatum, which is almost ses- sile; they are somewhat more evident in Sagartia lucice; and they are decidedly characteristic of Condylactis. Pedal locomotion in all these forms is accomplished by a wave of contraction that arises on one side of the foot and sweeps slowly across it to the opposite side. Appar- ently this wave is not fixed in relation to the secondary axes of the animal. It may arise at any point on the periphery of the foot without reference to the internal organization of the animal, but, having once arisen it sweeps across the center of the foot so that a line drawn from the point of origin through the center of the pedal disc marks the direction of locomotion. In a specimen of Condylactis whose foot measured 13 centimeters by 8 centimeters, the passage of a locomotor wave required on the average 3 minutes and the animal progressed by means of each wave on the average a little over 1 centimeter (Parker, 1917 c). To what extent are the creeping movements of the pedal disc dependent upon the animal as a whole? To answer this question experiments were carried out on Sagartia lucice. Fully expanded, attached specimens of this actinian were suddenly cut transversely in two with a pair of sharp scissors. The oral pieces thus cut off and carrying with them in each case the whole of the oral disc, tentacles and so forth, were discarded. The attached pedal discs and remaining portions of the columns were THE NERVE-NET 125 held under careful observation. These contracted at the level at which they had been cut so as to look like an ac- tinian normally withdrawn!. They soon filled themselves with water and, in twelve to fifteen hours after the oper- ation, many of them were creeping about precisely as the whole animal did. A single record will illustrate this condition. One of the animals without its oral disc began creeping and was observed to carry out 4 movements in 16.5 minutes, travelling in that period a distance of 7 millimeters. An animal with its oral disc intact that had been kept under similar conditions as a control, carried out 4 locomotor movements in 22.5 minutes, travelling in that period 6 millimeters. As the differences between these two sets of records are no more than may be seen in any pair of normal individuals, the locomotion of the two animals may be regarded as essentially identical. Individ- uals without oral discs not only creep as whole individuals do, but they also attach themselves to a glass surface as firmly as do those with oral discs. Furthermore, those without oral discs creep away from the light as consist- ently as do normal individuals. In fact, so far as the creeping is concerned, it is impossible to distinguish one class from the other, except perhaps that the operated animals are somewhat less inclined to creep than the normal ones are. These results are in entire accord with Loeb 's investi- gations (1895, 1899) in which he has shown that an Ac- tinia equina from which the oral end has been cut off, will creep more or less continuously on glass, and will attach itself firmly to a mussel shell just as a normal ani- mal will. They are also in accord with Jordan's results (1908), in which it was shown that the reflex excitability and muscle tonus of actinians is not under the control of superior nervous centers lodged, for instance, in the 126 THE ELEMENTAEY NERVOUS SYSTEM oral disc. They lead to the conclusion that the pedal disc and its immediately adjacent parts contain all the neuro- muscular mechanism that is necessary to creeping; in other words, this function is in no sense dependent upon assumed nervous centers in other parts. Thus the movements of the tentacles and of the pedal discs of actinians, parts whose nervous organization is that of the nerve-net, exhibit in a striking way an unusual degree of autonomy, and in this respect these organs are in strong contrast with such parts as the appendages of arthropods or of vertebrates whose activities are de- pendent upon a centralized nervous system and not a nerve-net, and disappear almost completely when they are severed from that part of the body in which the central organ is located. Thus the autonomy conferred by a nerve-net upon a given organ is one of the striking fea- tures of this type of organization as compared with that seen in animals with centralized systems. Autonomy is characteristic not only of coelenterate organs in which there are nerve-nets, but also of those parts of the higher animals in which there are similar structural conditions. The pedicellariae, spines, and am- bulacral feet of the echinoderms very probably belong under this head as well as the remarkable cloacal organ of holothurians whose pulsations have recently been studied by Crozier (1916). Autonomous organs with their nerve-nets seem to be less in evidence in worms and in arthropods than in the lower animals. In mollusks the foot is apparently largely autonomous and provided with a nerve-net, and the labial palps of the bivalve Anodonta are in this respect truly remarkable (Cobb, 1918). In vertebrates nerve-nets are abundantly present in con- nection with the circulatory system. The adult verte- THE NERVE-NET 127 brate heart is a muscle freely permeated by a nerve-net. It has already been pointed out that there is a sharply marked difference of opinion as to the nature of the heart's contraction, some investigators maintaining that the beat is dependent upon a rhythmic activity of the nervous tissue that is transferred to the muscle ; others, and these include the majority of modern workers, that the rhythm originates in the muscle itself, the nerve-net and other such structures being merely means of modify- ing the beat. Strong evidence in favor of the latter or myogenic theory of the heart-beat is found in the fact that the heart of the vertebrate embryo beats long before the first appearance of nervous tissue and that in tissue cul- tures single rhythmically contractile muscles fibers be- yond the reach of anything possibly nervous can often be found. Yet, notwithstanding this evidence, it is entirely possible that the cardiac nerve-net may have come to be a more intimate part of the rhythmic mechanism of the adult heart than is supposed by many, in which case this organ in its autonomy would have a strong resemblance to the actinian tentacle, for, like the tentacle, it may be removed from the animal and by appropriate means made to act in a strikingly normal manner for a relatively long time. Another portion of the vertebrate body that exhibits much autonomy and at the same time possesses a nerve- net is the digestive tube; especially the small intestine. Under normal circumstances the food in the small intes- tine is carried along from the end next the stomach to the outlet into the large intestine by a process of peristalsis, the essential part of which consists of a wave of con- striction that starts in the higher part of the tube and passes downward to its lower end. This wave of con- striction is preceded by a wave of relaxation which nat- 128 THE ELEMENTARY NERVOUS SYSTEM urally facilitates the progress of the food. Besides peri- stalsis the small intestine exhibits what has been called rhythmic segmentation. This consists of a set of con- strictions whereby a continuous band of food is broken up into a series of masses like a string of beads. Each mass is then sub- divided into two by a new con- striction and the two halves from adjacent masses fuse to form a new mass. By a repetition of this process the food becomes com- pletely mixed with the digestive secretions and is brought abun- dantly in contact with the absorb- ing intestinal walls. The walls that carry out these movements (Fig. 37) consist of a mucous epithelium on the inside followed by a submucous layer containing the submucous or Meissner's nerve plexus. This is followed in order by the layer of circular muscle fibers, the myen- teric or Auerbach nerve plexus, and the layer of longitu- dinal muscle fibers, that abuts against the serous layer. In both the submucous and the myenteric plexus the cells present all the characteristics of protoneurones and form a continuous network; these two plexuses may therefore be regarded as true nerve-nets. In addition to this in- trinsic nervous mechanism the small intestine receives nerve fibers from the vagi and the sympathetic chain. If these extrinsic nerves for a given part of the small intestine are cut, that part will still exhibit rhythmic con- Fio. 37. — Diagram of the nerv- ous organization in the intestinal wall of a vertebrate (after Lewis, 1910); m, mucous layer; « p, sub- mucous plexus; c m, circular muscle; m p, myenteric plexus; I m, longitudinal muscle; «, serous layer. THE NERVE-NET 129 tractions. In fact, Mall (1896 a) has shown that a piece of small intestine may be removed from the body, kept on ice twenty-four hours, and on being perfused in a warm bath, will contract rhythmically, an observation coirimned in its essentials by Cannon and Burket (1913). The small intestine must, therefore, be admitted to exhibit a high degree of autonomy. Magnus (1904) has attempted to ob- tain evidence as to the location of the layer or layers con- cerned with this autonomy. If a piece of intestinal wall is split so as to remove the mucosa and submucosa with the submucous plexus, the remaining portion will exhibit all the reactions that the section, of the intestine before the removal showed. Hence it is believed that the sub- mucous plexus is not necessary for the essential move- ments of the intestines. If, now, the remaining portion is separated into two sheets, one including the circular muscles and the other the myenteric plexus and the longi- tudinal muscles, that containing the circular (muscles shows no response except to mechanical stretching while the portion containing the myenteric plexus and the longi- tudinal muscles will still exhibit spontaneous rhythmic movements. It, therefore, seems probable that the move- ments of the small intestine depend largely upon the myenteric plexus and that this plexus represents a nerve- net that acts on the adjacent musculature much as the nerve-nets in the lower animals do. In this way the small intestine, though under the influence of extrinsic nerves, also retains a relatively high degree of autonomy depen- dent upon its nerve-net. Its activities, particularly its rhythmic segmentation, affords an excellent example of von Uexkiill's principle, for the segment of the intestine that is momentarily distended, whereby its nerve-net is stretched, is the region into which the impulses for re- newed contraction flow most freely. 9 CHAPTER X TRANSMISSION IN THE NERVE-NET A SECOND peculiarity commonly attributed to the nerve-net as contrasted with the synaptic nervous sys- tem is diffuse transmission. In the synaptic nervous system of the higher animals nervous impulses travel from the receptors in one direction only, over well-cir- cumscribed paths to given effectors; in other words, the synaptic nervous system exhibits a high degree of polar- ity. In the nerve-net, on the other hand, an impulse started at any point is believed to spread freely in all possible directions throughout the structure. This aspect of the nerve-net is best seen perhaps in the subumbrellar system of the medusae. As Romanes (1877, 1878), Eimer (1878), and especially Mayer (1906) have shown, this net may be cut into the most intricate patterns and yet so long as the original organic continuity from, point to point exists a nervous impulse may be started anywhere and will spread throughout the full extent of the tissue. Much the same form of diffuse transmission is to be seen in the sea-anemones, from almost any point of whose surface the whole retractor musculature may be brought into action. In other examples of the nerve-net evidences of be- ginning polarization can be seen and these may be re- garded as suggestive of the kind of steps by which the nerve-net was converted into the synaptic system. This is perhaps nowhere better seen than in the tentacles of actinians. When the tip of a tentacle is vigorously stimu- 130 A A TRANSMISSION IN THE NERVE-NET 131 lated, the whole tentacle is likely to respond, but when a point lower down on the side of the tentacle is stimulated, the reaction is chiefly from this point proximally ; in other words, transmission is more readily accomplished from the tip toward the base of the ten- tacle than in the reverse direction. The same kind of evidence has been shown by Rand (1909) to come from the reparative steps in regeneration. If a tentacle is cut off, the stump contracts vigorously and on reexpanding forms a ter- minal nipple; the cut face of the distal segment usually contracts only a little and seldom, if ever, closes the open wound. Yet, if such a distal piece is again cut cross- wise, its proximal part contracts and eventually forms a nipple, while the distal piece remains al- most unaffected (Fig. 38). The polarity thus exhibited is in the same direction as that which was shown by the reactions of the at- tached tentacle to mechanical stimulation. If the tentacles of a sea-anemone that has been thor- oughly anesthetized with chloretone are touched, no re- sponse whatever follows. If they are cut, neither the proximal nor the distal parts contract but both remain flabbily open. In consequence of these conditions it is believed that the polarity of the tentacle in so far as it is exhibited by the reaction just noted is of a nervous na- A B FIG. 38.— Diagram of the reactions of a tentacle of a sea-anemone to transection. On transecting a tentacle A , the distal end of the proximal piece a contracts and forms a nipple and the proximal end of the distal piece 6 remains open. On transecting the distal piece B the process is repeated in that the distal end of the new proximal piece c forms a nip- ple and the proximal end of the new distal piece d remains open. 132 THE ELEMENTARY NERVOUS SYSTEM ture, for it disappears on applying a drug that eliminates nervous activity (Parker, 1917 b). When the nervous structure in the tentacle of an ac- tinian is examined with the view of seeking some condition upon which this form of polarity can be based, a most simple relation is discov- erable. As Groselj (1909) has pointed out, the nerve fibrils that arise from the sense cells in the tentacles of actinians extend as a rule in a direction corre- sponding with the length of the tentacle. In Bu- nodes many of these cells are bipolar and in that f,-, • -. , -, -.• CaS6 OU6 nDrii CXtendS CllS- _ , . tallV alonff 1116 tentacle J and the other proximally, but there are also in this actinian a goodly number of cells that are unipolar and in such instances the fibrils almost invariably extend toward the base of the tentacle. In Cerianthus (Fig. 39) almost all the sense cells in the tentacular ectoderm are unipolar and their fibrils run almost without exception toward the base of the tentacle. As these fibrils transmit impulses away from the receptive cell bodies with which they are asso- ciated, it follows that in both these actinians nervous transmission must be predominately toward the base of the tentacle and that consequently the region of response would be largely proximal to the region of stimulation. This is what is to be seen in the neuromuscular reactions of most actinian tentacles and it is, therefore, believed Fio. 39. — Sense cells with their attached fibers in a tentacle of the sea-anemone Ceri- antkus. The arrow points toward the base of the tentacle, a direction taken by most of the fibers. (After Groselj, 1909.) TRANSMISSION IN THE NERVE-NET 133 that the polarity of these organs, as evidenced in the processes just mentioned, is dependent upon the prox- imal direction taken by the sensory fibrils in these struc- tures whereby the nervous impulses are led to flow pre- dominantly toward the base of the tentacle. This ana- tomical interpretation of the polarity of the tentacle is supported by the observation made by Chester (1912) that after two tentacles of Metridium have been grafted together base to base there is no change in their neuro- muscular polarity. Signs of polarization occur not only in the nerve-nets of the lower animals, but they are also evident in those of the higher forms. The rhythmic segmentation of the small intestine, already mentioned, may at times be a purely local process unconcerned with the progress of the food. Under such conditions the intestinal nerve- net must be acting as a diffuse mechanism. Commonly, however, segmentation is associated with peristalsis and the food is not only churned but moved along through the intestine. The direction of this motion is regularly toward the anus and the polarity thus exhibited is not unlike that seen in the actinian tentacle. That intestinal polarity, like tentacular polarity, probably rests on a structural basis is seen from an experiment by Mall (1896 b) in which a section of small intestine was cut free, turned end for end and healed into its old position re- versed (Fig. 40). After recovery the animals began to show serious digestive complications, and on killing and examining them it was found that the food had accumu- lated in the intestine at the stomach end of the reversed piece, a state of affairs indicative of marked and persis- tent polarity in the intestinal nerve-net. The heart of the lower vertebrates also exhibits pro- 134 THE ELEMENTARY NERVOUS SYSTEM nounced polarity, for contraction normally begins in the region of the sinus and progresses thence in sequence over the auricle, the ventricle and the bulb. The fact that by appropriate stimulation this sequence may be re- versed gives the heart a more diffuse character than the small intestine shows. But the uncertainty as to the myogenic or neurogenic na- ture of its beat leaves the heart a less clear example of these peculiarities than some other instances. As diffuse transmission is characteristic of the nerve-net and polarized transmission is a feature of the synaptic system, it fol- lows that the first signs of polarization in the nerve- net may be regarded as the initial step in the process of converting this organ into one of a synaptic type. As this process is apparently accomplished by the elongation of the neurofibril constitu- ents of the net in a particular direction, this feature may be regarded as the anatomical indication of the coming change. With the growth of such a feature the nerve-net begins to lose its diffuse condition and its transmitting fibrils come to form bands or trunks with some resem- blance to nerves ; in other words, the nerve-net loses some of its net-like character and comes to develop a partial Fio. 40. — Diagram to illustrate the preparation of the intestine in Mall's ex- periment on intestinal polarization. A, piece of intestine suspended by its mesen- tery and freed by two transections, but •till in normal position. B, the same, but with the piece of intestine healed into place in reversed position, a, anterior; p, poiterior. TRANSMISSION IN THE NERVE-NET 135 cleavage or grain. Consequently receptive cells, instead of being closely associated with, the muscles that they control, may come to lie at some distance from them and thus arises the necessity for extended transmission. This is clearly seen in the contrast between the tentacles and the general retractor mechanism of actinians. In the ten- tacles a system of ectodermic sense cells immediately over- lies the ectodermic longitudinal muscle and hence conduc- tion from the sense cells to the muscle is simple and direct. In the retractor mechanism, on the other hand, an ecto- dermic system of sense cells calls into action a distant entodermic system of retractor muscles in the mesenter- ies, and this involves not only a nerve-net but a net of such dimensions as to accomplish the extended transmis- sion necessary for the performance of general retraction. Such modifications of the nerve-net lead to conditions in which are realized the beginnings of unquestionable re- flex activity. In this form of response a definite motor or other efferent activity appears after the application of a specific stimulation. In other words, on the applica- tion of a definite stimulus not a diffuse response but a highly particularized one appears. This is well seen in certain reactions that depend upon oral and tentacular stimulation in the sea-anemone Metridium. If a small amount of seawater is discharged into the mouth of an expanded, resting Metridium, no response is usually noticeable. If the seawater contains hydrochloric acid -j^- , the actinian immediately opens the oesophagus widely and exhibits on its column a few well-marked ver- tical grooves. These disappear gradually as the oesoph- agus closes. If the position of these grooves is carefully noted, it will be found that one is always present for each siphonoglyph and that the others are distributed in ac- 136 THE ELEMENTARY NERVOUS SYSTEM cordance with the arrangement of the other pairs of com- plete mesenteries (Fig. 41). The grooves thus mark the lines of attachment of these mesenteries and are the result of the contraction of their transverse muscles, which are those concerned with the opening of the oesophagus. If fragments of fish meat are put on the lips of a fully expanded Metridium, they are carried into the animal by A .. B Fio. 41. — A, outline of the sea-anemone Metridium showing on its column the vertical grooves that accompany the opening of the oesophagus. B, transverse section of the same sea-anemone to show the relation of the oesophagus and grooves to the pairs of complete mesenteries whose transverse muscles open the oesophagus and form the grooves. ciliary action through an oesophagus that opens widely to receive them and during this operation the column of the animal is marked by the same vertical grooves that were seen in the experiment with acidulated seawater. As the pieces of food pass into the digestive cavity the grooves fade out. It is clear, then, that the transverse muscles of the complete mesenteries are concerned with the expansion of the oesophagus for the reception of food. If a piece of fish meat is placed upon the tentacles of an expanded Metridium, these organs become character- istically stimulated and if the meat is removed before it is brought by the tentacles to the animal's lips, the cesoph- TEANSMISSION IN THE NEKVE-NET 137 agus will still open, accompanied by the formation of ver- tical grooves on the column. This response can not be elicited by the application of weak acid to the tentacles. Under such circumstances a withdrawal of the. oral disc takes place. Thus it appears that not every form of effective chemical stimulus that can be applied to the ten- tacles is followed by an opening of the oesophagus. To appropriate stimuli, however, the tentacles and lips may act as receptors for the opening of this tube. No other parts of the bodj^ of Metridium have been found from which this cesophageal response can be called forth. The response, therefore, partakes of the nature of a true re- flex in that it implies definite transmission tracts from the tentacles and from the lips to the transverse muscles of the mesenteries, and these tracts are called into activ- ity only by specific forms of stimulation. Thus while dilute acid applied to the tentacles will excite the general nerve-net of Metridium, fish meat applied to these organs calls forth the oesophageal response only. Jordan (1908) has claimed that animals possessing only nerve-nets in their nervous organization exhibit reflex deficiency and this in general is undoubtedly true, but, as the present example shows, some of these animals may have involved in their nerve-nets specialized tracts that enable them to carry out simple but obvious reflexes. Such conditions indicate the steps by which a nerve-net may be converted into that type of central nervous organ that is char- acteristic of the higher animals (Parker, 1916 a, 1917 d). CHAPTER XI APPROPEIATION OF FOOD AND THE NERVE-NET IF the ccelenterates, whose nervous system is largely a nerve-net, nevertheless exhibit locally signs of a higher differentiation, may it not be possible that in some of their more complex operations they may show signs of all the higher nervous functions of the most complex animals? In attempting to find an answer to this ques- tion, so far as actinians are concerned, such activities as their feeding habits, their retraction and expansion, and their locomotion have been investigated. The appropriation of food is an activity with which the oral disc of actinians is principally concerned. The movements of the tentacles, mouth, and other such parts by which food is ingested were ascribed by Nagel (1892, 1894) to muscular action alone, but Loeb (1895) pointed out that cilia also play an important role. The parts that are immediately concerned in the appropriation of food are the five following: the tentacular gland cells, whose secretions render the tentacles adhesive, whereby pieces of food become attached to them ; the musculature of the tentacles, by which these organs are pointed toward the mouth ; the tentacular cilia, which sweep toward the ends of the tentacles and thus deliver the food to the mouth when the tentacles are pointed in that direction; the transverse muscles of the complete mesenteries, by which the oesophagus is opened; and the cilia of the lips and oesophagus, which in the presence of food reverse their 138 FOOD AND THE NERVE-NET 139 usual outward stroke and thus transport such materials to the digestive cavity. Besides these five sets of parts some actinians include in the means by which they appropriate their food a sixth system, namely, the mus- culature of the oral disc. In Stoichactis, for instance, as described by Jennings (1905), and in Cribrina, as re- ported on by Gee (1913), the mouth during feeding is moved by the oral musculature toward the food-bearing tentacles, a shifting which has also been observed in cer- tain corals (Carpenter, 1910). This operation, though it can be seen to occur in Metridium, is relatively so insig- nificant in this form that it may be passed over without comment; the important elements in the feeding of this actinian are the five already mentioned. Much confusion and uncertainty exists in the various accounts of the methods by which actinians obtain their food and more or less of this is due to the failure on the part of writers to designate the particular form of ac- tivity that they are for the moment discussing. Thus both ciliary and muscular activity are involved in the appropriation of food and have often been indiscrim- inately dealt with in accounts of this operation. Their significance for the animal as a whole is, however, very different and it is, therefore, highly desirable that they should be; kept clearly in mind as separate processes in any discussion in which they are involved. Of the five principal events that go to make up the act of food appropriation, three exhibit so little variation that they may be regarded as essentially uniform. These are the secretion of mucus, the beat of the tentacular cilia, and the opening of the oesophagus. In none of these are there during feeding any important readjustments which 140 THE ELEMENTARY NERVOUS SYSTEM are essential to the acquisition of food ; the production of mucus is apparently a strictly local response to a local stimulus ; the beat of the tentacular cilia is constant and irreversible ; and the opening of the oesophagus is as sim- ple and mechanical a reflex as can well be imagined. The idea that the oesophagus, as often intimated, exhibits per- istalsis is probably incorrect. At least a careful inspec- tion of this organ in action in Metridium gives no support to this idea. The two remaining events in the appro- priation of food, the responses of the oral cilia and the movements of the tentacles, are both open to significant changes and are of the utmost importance in judging of the relation of this process to the actinian as a whole. Unlike the tentacular cilia, the oral cilia, those of the lips and of the 03sophagus, may reverse the direction of their stroke so that the usual outward current can be con- verted into an inward one. This reversal is, under ordi- nary circumstances, a local response on the part of the cilia to certain dissolved substances in the food. Its rela- tive independence of the other activities of Metridium can be shown in a number of ways. Thus, though it is a response to food, excessive feeding has no marked influ- ence on it. Allabach (1905) caused a Metridium to gorge itself with food, a process that can result finally in dis- gorgement, and yet immediately after the animal had emptied itself, its oral cilia were found to reverse to food, which was thus passed down its oesophagus. Further, if pieces of meat are fed to the lips of the oral half of a Metridium cut transversely in two, the cilia reverse and the masses of food thus carried through the oesophagus are discharged at its open pedal end. By this means in the course of an hour or so many times the amount of food FOOD AND THE NEEVE-NET 141 that the body of a Metridium could contain can be passed through its oesophagus, and yet the ciliary reversal is as effective after this period of continuous feeding as before. Other evidence of the relative independence of the oral cilia as compared with other effectors is well seen in speci- mens of Metridium that have been narcotized with chlore- tone, by which all nervous activity is abolished. A piece of food placed upon the tentacles of such an animal calls forth no special response and either remains where it was placed or moves sluggishly off to the periphery of the disc under the action 'of the tentacular cilia. When, however, such a piece is put on the lips, the cilia reverse and the morsel is gradually carried down the oesophagus and discharged into the digestive cavity. The swallowing is usually not so rapid as in the normal animal, for, under this form of narcotization, the transverse muscles of the mesenteries do not respond to the food by opening the oesophagus and consequently the cilia are obliged not only to transport the morsel but to force it down a partly closed tube. This, however, they are usually able to do, and thus, quite independent of neuromuscular help, they bring about the swallowing of food and the rejection of non-food, for under these circumstances inert materials were found not to reverse the ciliary stroke. Thus, as Allabach (1905) has pointed out, the reversal of the ef- fective stroke of the oral cilia is a process that is largely independent of the physiological state of Metridium. In one particular only does this process appear to be related to the general condition of the animal. Ordi- narily the reversal of the oral cilia is accomplished by dissolved substances from the food, and in the earlier studies on this subject in Metridium this was believed to be the only way by which such reversal could be induced 142 THE ELEMENTARY NERVOUS SYSTEM (Parker, 1896). Torrey (1904 a), however, showed that in Sagartia this reversal could be brought about by me- chanical stimuli as well as by chemical means and that it was favored by a starved condition of the animal. Alla- bach (1905) also found that in Metridium a ciliary re- versal could be induced by mechanical means, and Gee (1913) has recently shown that specimens of Cribrina which have been in the laboratory some time do not ex- hibit a reversal to mechanical stimuli, whereas those still in their native pools give evidence of it. Further investigations have shown the correctness of Allabach's contention (1905) that in Metridium margin- atum some individuals on mechanical stimulation reverse their ciliary stroke readily, others1 less readily, and still others not at all, variations largely dependent upon whether the animals have been starved or fed. Two un- derfed specimens of Metridium which on being tested were found to reverse their cilia to clean filter-paper were vigorously overfed and after three hours were tested again with bits of clean filter-paper. In both instances the paper failed to bring about a reversal of the cilia and consequently was rejected. In another test made eighteen hours after feeding, the paper was engulfed, showing that the cilia had returned to the state characteristic of ani- mals that had lacked food. It is, therefore, clear that an underfed Metridium will reverse the effective stroke of its oral cilia to mechanical stimulation, though a small supply of food will obliterate this peculiarity and leave these organs incapable of such reversal. The occasion of this loss of the power to reverse the stroke of the oral cilia on mechanical stimulation has been ascribed by Allabach (1905) to the difference in metabolism between a well-fed and an underfed indi- FOOD AND THE NERVE-NET 143 vidual. This has been tested by cutting out the cesoph- ageal tubes from several specimens of Metridium, laying them open and experimenting with them as ciliated mem- branes. If they are carefully prepared from animals that have not been recently fed, they will show a well-marked ciliary reversal to pieces of clean niter-paper. To frag- ments of mussel they reverse the ciliary stroke in the way characteristic for food, and after a dozen or more such trials they will no longer reverse to pieces of clean filter-paper. Thus the isolated membrane exhibits all the changes that it does as a part of the whole animal and under conditions where it is quite obvious that the one change that it has suffered is fatigue. It is, there- fore, believed that the general metabolism of Metridium is not so much concerned with the change in the charac- ter of the response of the cilia to filter-paper as the fa- tiguing of the receptive mechanism of the ciliated surface is. In the undisturbed state this mechanism is at its greatest sensitiveness, but, on feeding, its efficiency dimin- ishes and hence filter-paper no longer excites a reversal, a change which is now called forth only by the more vig- orous stimulation from the dissolved products of the food. Hence the activities of the oral cilia are probably even more independent of the rest of the actinian than Alla- bach (1905) was inclined to insist upon. The feeding movements of the tentacles in actinians are obvious neuromuscular reactions, as their disappear- ance on narcotization with chloretone amply shows. The independence of the individual tentacles in their feeding reactions has been demonstrated in a number of forms, in which these responses have been observed after the tentacles have been cut from the polyp. That one ten- tacle can influence another through connections in the 144 THE ELEMENTARY NERVOUS SYSTEM oral disc has been proved for Condylactis and is prob- ably true for Metridium. The muscular responses of the tentacle in feeding, therefore, give much more opportu- nity for unified action than do the ciliary responses just considered. That tentacular responses in actinians change with continued activity has long been recognized. Jennings (1905) found that the tentacles of Stoichactis after they had been vigorously plied for a while with meat ceased for a time to react to food. Allabach (1905) noted that in Metridium the tentacular reactions became gradually slower or even ceased as feeding progressed, and the same is recorded by Gee (1913) for Cribrina. Evidence of this in Metridium was long ago published (Parker, 1896) and recent work on this point has been entirely confirmatory. Jennings (1905) attempted to explain this change as due to loss of hunger,1 but Allabach (1905) showed that it also occurred when the tentacles were stimulated, though the animal was not allowed to swallow the food. Her conclusion is that it is simply the effect of fatigue. Gee (1913), however, declined to accept this explanation because if an actinian that will ordinarily show this ten- tacular change after having been fed eight or ten times, is experimented upon when in a fresh condition and is made to contract about the same number of times, its ten- tacles are found not to have lost their responsiveness. But both Allabach and Gee have failed to recognize that ill ore are several kinds of fatigue. It is perfectly clear, 1 It is perhaps unfortunate that the term hunger should have been used, for it is somewhat ambiguous. Usually it stands for a well known sensa- tion due to movements of the stomach (Cannon and Washburn, 1912) ; less commonly for insufficient bodily nutrition. Pathology has long since demon- strated that these two phenomena are not necessarily connected, but in which sense Jennings intended to u»o the term is not always wholly clear. FOOD AND THE NEEVE-NET 145 from Gee's experiment, that muscular fatigue is not ac- countable for the change in the responsiveness of the ten- tacles, but it is entirely possible that it may have been caused by sensory fatigue. It is a common observation that if a sensory surface is placed under active stimula- tion, it is often only a short time before it will fall off very considerably in its receptiveness, and it is this form of fatigue in all probability that is accountable for the change in the tentacular responses of Metridium on con- tinuous feeding. A repetition of Allabach's experiment of placing food on the tentacles of Metridium and, after they have responded, of removing it from the lips before it has been swallowed has in all instances confirmed her results; namely, the tentacles fall off in responsiveness. In view of what has already been stated it seems impos- sible to explain this phenomenon except as a result of sensory fatigue. But there are also changes in the tentacular responses of actinians that are by no means so easily explained as are those that have just been considered. Jennings (1905) states that when the tentacles on the left side of an Aiptasia were plied with crab meat, they transferred the food to the mouth quickly five times, after which they reacted slowly on the sixth trial and hardly at all on the seventh. On trying the meat on the tentacles of the right side, it was found that the transfer to the mouth was quickly accomplished. Returning now to the left side, four sluggish deliveries were effected, after which the right side would now take no meat at all. Allabach (1905) states that Metridium can be fed from one side of its disc till no more food will be accepted, whereupon food will likeT-ise not be accepted by the tentacles of the opposite side. Gee (1913) has also recorded essentially the same 10 146 THE ELEMENTARY NERVOUS SYSTEM condition in Cribrina. From these observations it seems clear that changes induced in the muscular responses of the tentacles of one side influence to no small degree the reactions of the tentacles on the other side. As Jennings (1905) has put it, the animal reacts as a unit, one side influencing the other. Experiments of this kind have been repeated on Metridium, and though the results were not so striking as those described by the authors already quoted, it was clear that when a Metridium was fed per- sistently by means of the tentacles of one side and so as to avoid touching with the food those of the other side, the opposite tentacles w^ere nevertheless eventually in- fluenced in their reactiveness and became less responsive as the feeding proceeded. Here would seem to be a good instance of some such general effect as that of changed metabolism or the general utilization even of nervous experience. To ascertain whether changes in the tentacular re- sponses of one side of the disc are transmitted nervously to the other side, small pieces of mussel were fed to the tentacles of one side of a Metridium but removed before they were swallowed and then, after the tentacles of that side began to lose in responsiveness, those of the other side were tested to see if they too had lost in their ca- pacity to respond. The time in seconds required for the swallowing of each piece of food is recorded in the fol- lowing table. The rejection of a piece of food is indicated by the sign of infinity. It must be evident from an inspection of Table 3 that the right side of the animal gave no evidence of having been influenced by the left and that therefore there is no ground for the assumption that the experience of one side is transmitted nervously to the other. In other experi- FOOD AND THE NERVE-NET 147 TABLE 3. TIME IN SECONDS FOR THE TRANSFER BY THE TENTACLES OF METRIDITTM OP SMALL PIECES OF MUSSEL TO THE MOUTH WHEREUPON THEY WERE RE- MOVED AS THEY WERE ABOUT TO BE SWALLOWED. SIXTEEN TRIALS WERE MADE ON THE LEFT SIDE AND THEN THE SAME NUMBER ON THE RIGHT. oo INDICATES A DISCHARGE OF THE PIECE OF MEAT AT THE PERIPHERY OF THE ORAL Disc. Number of the Trial 1 2 3 4 5 6 7 8 0 10 11 12 13 14 15 16 Left side of disc 284 121 107 86 103 92 00 71 58 oo 97 108 CD oo 72 62 oo 00 Right side of disc 72 86 306 63 112 83 132 74 96 103 97 86 78 109 ments, in which the fragments of mussel delivered to the tentacles of the first side were allowed to be swallowed instead of being removed, the tentacles of the opposite side very regularly exhibited a decline in responsiveness. It is, therefore, believed that this change is due to the food introduced into the digestive cavity, and, since the pieces of food were very small, not to the accidental transfer of food juices from the side of the disc stimu- lated to the other, as suggested by Gee (1913). To remove any doubt on this point a modification of an experiment tried by Gee (1913) was adopted and by means of a fine glass syringe a considerable amount of mussel juice was injected through the column wall of small specimens of Metridium into their digestive spaces. This operation was easily accomplished, especially if the region through which the puncture was made was pre- viously anesthetized with magnesium sulphate. The in- jected juice apparently did not escape from the mouths of the animals, which, however, took in a considerable amount of seawater and enlarged, much as well-fed ac- tinians do. After an hour or so the tentacles of the in- jected actinians were tested with fragments of mussel and 148 THE ELEMENTARY NERVOUS SYSTEM found to be very noticeably insensitive to food. It, there- fore, seemed clear that it was the food in the digestive cav- ity rather than any accidental overflow that had influ- enced the tentacles. Reasons have already been pointed out for believing that the change in the responses of the tentacles after continuous feeding is due to sensory fatigue and not to a general metabolic change; this seems also to be true in the particular instance under consideration. Though the meat juice injected into the digestive cavity unquestiona- bly serves as material for metabolism and eventually must have its influence on the animal's general state, its first condition is that of a component of the fluid mixture that bathes the inner surfaces of the actinian. These surfaces include the cavities of the tentacles. As shown elsewhere (Parker, 1917 b), substances in solution in the digestive space of such organs as the large tentacles of Condylac- tis penetrate in a very short time the thin walls of these parts and thus make their way to the exterior. In doing so they must come in contact with the sensory ectoderm. Since the changes in the reactions of the tentacles pro- duced by food juices injected into the digestive cavity are in the direction of diminished response, and since these changes come over the tentacles with consider- able rapidity and before a modified metabolism depen- dent upon new food could have got much headway, it is believed that the loss of responsiveness in this instance, like that in the former case, is due to sensory fatigue and not to changed metabolism. In the first instance the fatigue was produced by the direct application of stim- ulating substances to the exterior of the tentacles ; in the second to the transfusion of those substances from the cavities of the tentacles to their sensory mechanism. If FOOD AND THE NERVE-NET 149 this explanation is correct, as there is good reason to suppose it is, the responses of the tentacles are like those of the oral cilia in that they are not especially dependent upon the condition of the animal as a whole. As Gee (1913) states, "the view that the seat of the modified re- sponsiveness lies very largely in the individual tentacles is more clearly in accord with what is known of the struc- tural organization of the sea-anemone than that the ani- mal acts as a unit." The feeding habits of sea-anemones thus prove on ex- amination to consist of operations none of which neces- sitate the assumption of activities other than those con- sistent with the nature of the nerve-net. There is no rea- son whatever to resort to the hypothesis of a controlling nerve center. All the activities are strikingly local and the changes that they exhibit are apparently entirely due to fatigue. In these respects they are in strong contrast with the feeding habits in the higher animals, a process which has become so deeply wrought into the make-up of these forms that its relation to the animal as a whole is most profound. While almost every one of the ele- ments involved in the feeding of actinians may be ex- perimentally isolated and made to act for itself in a re- markably local way, scarcely any such independence is observable in the parts concerned in the similar opera- tions of higher animals ; the jaws and their muscles, buc- cal glands and so forth in these higher animals exhibit a highly unified action dependent chiefly upon central nerv- ous connections such as is scarcely suggested in actinians, but as isolated elements they have almost no reactive power at all as compared with what is possible in sea- anemones. The feeding habits in actinians then empha- size the relative independence of parts rather than the action of the organism as a whole (Parker, 1917 d). CHAPTER XII OTHER COMPLEX RESPONSES AND THE NERVE-NET As the locomotor activities of most actuiians are ex- tremely limited, the chief protective response of these animals is general retraction whereby they are reduced greatly in bulk, their more delicate parts are brought under cover, and they shrink close to the substratum to which they are attached. In many instances, in fact, re- traction brings about a withdrawal of the body of the ac- tinian into deep, rocky recesses and the like whereby very efficient protection is secured. The reverse process, ex- pansion, is one which involves an enlargement and pro- trusion of the body as a whole and the opening of its folded surfaces and apertures in such a way that the operations of feeding, respiration, and so forth, may be resumed. Retraction and expansion are relatively simple processes. Retraction in its initial phases is chiefly the result of the action of the mesenteric muscles, the longi- tudinal muscles of the non-directive mesenteries depress- ing the oral disc, those of the directives serving chiefly to fold the siphonoglyphs, and the parietal muscles acting on the column wall. After the depression of the oral disc has proceeded somewhat, the contraction of the sphinc- ter muscle completes the process by bringing the oral disc under cover through the puckering effect of this muscle on the column wall. Incidentally the process of general retraction involves the expulsion of almost all the water contained in the digestive cavity of the actinian. 150 COMPLEX BESPONSES 151 The reverse operation, expansion, is dependent first of all upon the relaxation of the sphincter and of the mesenteric muscles; then follows the slow filling of the digestive spaces with seawater through the ciliary currents in the siphonoglyphs ; and probably as a last step the circular muscles of the column contract on the fluid contents of the body whereby the oral disc is forced well up above the pedal attachment, The details involved in the processes of retraction and expansion allow retraction to be accom- plished much more quickly than expansion. This rela- tion has all the appearance of an adaptation, for the quickness of a withdrawal may often be the essential part of the protection given by retraction, whereas there is nothing about the economy of an actinian, such as feeding, respiration, and so forth, that makes it vitally important for the animal to expand quickly. The conditions under which a Metridium remains fully expanded are by no means simple, but include an aggre- gate of factors. In the laboratory the fullest expansion was obtained when the animals were in well-oxygenated, cool, running seawater in the dark. Under such circum- stances this sea-aneinone will extend itself to as much as six times the diameter of its column, and hold its oral disc fully opened. In nature a greater degree of expan- sion than that seen in the laboratory under the circum- stances just stated apparently does not occur. This max- imum degree of expansion under natural circumstances has often been observed in sea-anemones in pools during the night or even during the day in dark situations such as under bridges and so forth. The elements that contrib- ute to this extreme expansion are certainly diverse. Of these light, temperature, food, oxygen supply, and water currents are of significance. 152 THE ELEMENTARY NERVOUS SYSTEM The influence of light on actinians is by no means uni- form but differs with different species. According to Nagel (1894, 1896), Adamsia, Anemonia,1 and Actinia are not responsive to light. Fleure and Walton (1907) have noted this lack of response in Anthea as well as in Adamsia. Pieron (1906 c, 1908 stalk being much slower than the hydranth. It would be natural to expect that one or other of these parts might serve as a pacemaker for the whole system, but of this there is no evidence. In a similar way a stalk may be cut crosswise in halves and the two halves will continue to show rhythmic contractions. As in the former case, both halves have a slower rate than the whole stalk had. Possibly in both cases the reduced rates give evidence of a general con- trol which is somewhat disturbed by the cutting, though of this there is no conclusive evidence. Besides the type of feeding that has just been con- sidered and that is apparently characteristic of quiet water, a second type is also to be noted (Torrey, 1904 &). When detritus of one kind or another is carried by a gentle current on to the expanded proximal tentacles of an erect Corymorpha, these are very likely to wave in- ward, carrying everything with them toward the distal tentacles, which in turn move quickly outward to meet the incoming proximal members and eventually trans- port their booty to the mouth. In this way under favor- able circumstances much food is doubtless obtained, but the success of this operation is inn. \\ more dependent upon accident and tin1 whole procodnt •• seems to have less HYDEOIDS 197 organized effort about it than the plan of feeding de- scribed for quiet water. The food accumulated by the two methods mentioned in the preceding paragraphs doubtless undergoes diges- tion in the interior of the polyp and is here moved about by the peristalsis of the proboscis already referred to. When the responses and activities of Corymorpha are compared with those of an anthozoan polyp, their inef- ficiency is most striking. This is especially well seen in the tentacular responses to food. In an anthozoan the tentacles when touched by a piece of food turn in many directions till they have more or less entwined the food. They become covered with a sticky mucus and they dis- charge their nettling filaments with great freedom. Finally by the action of their cilia and muscles the food is delivered at the lips. In Corymorpha the proximal tentacles are not provided with mucus and their one muscular response is to wave toward the mouth, a re- sponse that occurs as well when the food touches their aboral faces, and is consequently left behind by their re- sponse, as when it is on their oral face. No cilia are pres- ent in Corymorpha to help transport the food to the mouth. In Corymorpha the whole process of food gath- ering has a strongly marked mechanical character that makes it much less successful as a means of getting all the food within reach than the operations carried out by the anthozoan tentacle (Torrey, 1904 b). This lack of close adjustment, which has been noticed in the tentacles of Tubularia (Pearse, 1906) as well as in those of Cory- morpha, runs through all the reactions of Corymorpha as compared with those of the anthozoan polyps. Notwithstanding the general inefficiency of the re- sponses of Corymorpha, this polyp contains among its 198 THE ELEMENTARY NERVOUS SYSTEM muscles about the same array of types as are found in the sea-anemones. Some muscles, like the circular muscle of the stalk, are apparently quite without nervous connec- tions and respond to direct stimulation; others, like the circular muscle of the proboscis, probably respond as a general rule to direct stimulation, though they may be influenced by nervous impulses ; and finally muscles, like the longitudinal muscle of the stalk, are completely under nervous domination. Some of these, such as the longi- tudinal muscle of the proboscis, exhibit responses that are called forth in such a way that they are indistinguish- able from a reflex in the higher animals. From this stand- point Corymorplia reproduces in miniature all the con- ditions found in the sea-anemones, and this is further em- phasized by the lack of any general nervous center and the consequent great independence of all organs from the side of their neuromuscular activity. Corymorplia, therefore, does not seem to fill the gap between the ex- tremely simple effector system of sponges and the re- ceptor-effector systems of the coelenterates, but rather presents a reduced though not simplified state of the type in the sea-anemone. If the muscles of Corymorplia were more commonly open to direct stimulation than they are and if its activities presented less than can be inter- preted in the nature of a reflex, it might supply more nearly the requisites of an intermediate type, but, as it is, it resembles rather a reduced actinian than a form in any sense intermediate between sponges and sea-anemones (Parker, 1917 e). SECTION III. CENTRAL NERVOUS I ORGANS CHAPTER XIV CONCLUSION IT is intended that the present volume shall include an extended discussion of only the simplest examples of the elementary nervous system. In this concluding chap- ter, however, a brief outline will be given of the relations of this system to the central nervous system of the more complex animals. An outline of this kind must perforce block out only main contours, for, even were they known, it would be impossible in so limited a space as a single chapter to follow the intricacies to be met with in the evolution of that most complicated system of organs, the central nervous system. It was pointed out in the earlier part of this volume that the neuromuscular system probably did not origi- nate primarily as a nervous structure. The first trace of this system is to be seen in independent effectors, the smooth muscle of the lowest multicellular animals (Fig. 45). This tissue, as seen in the oscular and pore sphinc- ters of sponges, represents muscle unassociated with nerve and acting under direct stimulation from the en- vironment. Such independent effectors are apparently open to only a limited range of stimuli, particularly to those of a physical type, and are relatively slow and slug- gish in response. They reappear in the higher animals, as in the acontial muscles of the sea-anemones and in the 199 200 THE ELEMENTARY NERVOUS SYSTEM heart muscle of the vertebrate embryo, but in the evolu- tion of the neuromuscular mechanism they represent an initial stage and as such they are characteristic of only the simplest forms. The second step in this series is seen in the receptor- effector system of such animals as the coelenterates. This step has been accomplished by adding to the inde- pendent effector of the lowest forms a receptive element in the nature of a sensor}1- surface. Such receptors arose by the modification of those epithelial cells that were in closest proximity to the already differentiated muscle. These cells, in the form of sensory patches, became espe- Fio. 45. — Diagram of an independent effector, a muscle cell such as occurs in the lowest multicellular animals. cially receptive to categories of external changes, such as pressure changes, chemical changes, and changes in heat and light. The disturbances that these changes initiate in the receptive protoplasm are the means by which the sub- jacent muscles are brought into action. Thus the novel element in this combination, the receptor, serves as a trigger, so to speak, to discharge the underlying muscle, and, judging from the quickness with which this discharge is accomplished as compared with that with which the in- dependent effectors discharge, the efficiency of the new system is beyond doubt. The receptor-effector system in its simplest conceiv- able form consists of a patch of sensory cells attached to a group of muscle cells, but practically every instance illustrative of this particular combination is complicated by the fact that the central branches of the receptive cells CONCLUSION 201 are not only applied to the muscle cells but form among themselves a network of communication whereby the im- pulses that arise from a few receptive cells may be trans- mitted to many muscle cells instead of being limited to a restricted group. Here is to be seen the first trace of the nerve-net, an organ that in present forms makes the re- ceptor-effector system immensely more responsive and in past forms harbored the germ of a central nervous or adjuster ap- paratus that in the end profoundly affected the whole scheme of neuro- muscular organization. The simplest type of receptor- effector system with a simple form of nerve-net is such as is seen in the tentacles of many actinians where the receptive cells with their central branches and the muscle cells con- stitute the whole complex (Fig. 46). A more complicated type of this system and one that is more usual with the coelenterates is that in which, in addition to the two kinds of cells men- tioned, a third appears and from its position in the midst of the nerve-net adds by its numerous branches im- mensely to the intricacies of this structure (Fig. 47). This third type of cell was long ago recognized by the Hertwigs (1879-1880), who designated it a ganglion cell and attributed to it central functions such as were for- merly supposed to be of necessity associated with such elements. Cells of this kind, according to Havet (1901), intervened between the sensory cells and the muscles, and he, therefore, denominated them motor cells. ' As they are not massed together into ganglia and as they may FIG. 46. — Diagram of a simple type of receptor-ef- fector system such as is seen in the tentacles of sea-anem- ones. It consists of recep- tors r or sense cells whose basal nerve-net connects them with the deep-seated muscle cells m. 202 THE ELEMENTAEY NERVOUS SYSTEM have other functions than the control of muscles, it ap- pears that neither of these names is especially appro- priate. They are the cells in the simpler types of nerv- ous system from which the neurones of the more complex types have been derived. Hence the name protoneurone, as already suggested, is not inappropriate (Parker, 1918 a). A receptor-effector system whose nerve-net in- cludes protoneurones is, therefore, a step nearer a cen- tralized nervous system of the higher animals than one in which the protoneurone is not present. Although a receptor-effector system may consist of highly differentiated sense organs connected by a protoneurone nerve-net with distantly located muscles, it nevertheless is very far from representing even a simple central nervous system. First of all, such a receptor- effector system exhibits in a marked degree a state of diffuse- ness. The animals possessing such systems have no single nervous organ to which their nervous experience may be said to be referred and from which their im- pulses to response emanate. Their nervous activities are uncentralized. Each important organ, like the tentacle or the foot of an actinian or the column of Corymorplia, has within itself its own neuromuscular organization, and the autonomy thus conferred on this part is one of the characteristics of the effector-receptor system. This system is further characterized by great diffuse- ness of transmission. Any nervous impulse started up m Fio. 47. — Diagram of a com- plex type of receptor-effector sys- tem such as is seen in many parts of sea-anemones. It consists not only of receptors r, with their nerve-nets, and of muscle cells m, but also of the so-called ganglion cells g in the nerve-net. CONCLUSION 203 in it at any point is very likely to spread wave-like throughout its whole extent. This form of diffuse trans- mission is in strong contrast with that kind of transmis- sion that is carried out in the nervous systems of the higher animals where impulses follow definite paths in- ward and outward, reaching finally only special groups of muscles or even single muscles. These two features, FIG. 48. — Transverse section of the ventral nerve-chain of the marine worm Sigalion showing this chain as a thickened portion of the superficial ectoderm in which the sequence of tissues from the exterior inward is superficial epithelium e, ganglion cells g, and nerve fibers/. (After Hatschek, 1888.) the absence of a central station and the diffuseness of transmission, both of which are aspects of the same gen- eral condition, are the most striking characteristics of the receptor-effector system and bring this system into strong contrast with that final type of neuromuscular organization that is characteristic of the highest animals and in which a central nervous organ or adjuster is well differentiated. This most highly differentiated type of neuromuscular system in which an adjuster or central organ has arisen 204 THE ELEMENTARY NERVOUS SYSTEM between receptors and effectors represents the final step in the growth of this group of organs. As just indicated, the novel" element in this particular combination is the central organ or adjuster, and this arises in the region between the receptor and the effector and out of that material which in the elementary system constitutes the nerve-net. Its growth is associated with an inward mi- gration whereby it retreats from the surface of the ani- mal to a deeper situation and comes thus to gain what is significant of its growing importance, a certain degree of protection. In the coelenterates the nervous elements are mostly contained in the epithelial layers of the body and especially in the external epithelium, the ectoderm. In the worms, where the body has gained greatly in thickness and solidity as compared with the coelenterates, this in- ward migration is clearly seen. These animals no longer possess the diffuse system of the lower forms, but have a definitely centralized band of tissue extending the length of the body. The relation of this band to the ectoderm is well seen in a series of annelids. In Sigalion, a marine worm, this band, which consists of a brain and a ventral nerve-chain, is still a part of the superficial ectoderm (Fig. 48). In Nereis, by a process of delamination, the ventral chain has freed itself from the skin, though the brain is still a part of that layer. In the earthworm this process has been carried to its completion and the whole nervous band has separated from the ectoderm and mi- grated into the deeper parts of the animal (Fig. 49). This condition of a deep-seated central nervous organ is char- acteristic of the higher invertebrates such as the arthro- pods, mollusks, and so forth, and, as their embryology shows, it was brought about by a process analogous to that seen in the worms. In the vertebrates this inward CONCLUSION 205 migration of the central organ also takes place, but in this group it is accomplished, not by a process of delam- ination, but by an involution of a portion of the skin itself. The infolded layer of ectoderm thus produced assumes the form of a tube and the walls of this tube by differentiation give rise to the spinal cord and brain, the two chief organs in the central nervous system of these animals. The concentration of the diffuse nervous activities of the lower animals from the skin in general to a par- FIG. 49. — Transverse section of the ventral nerve-chain of the earthworm Allolobophora showing this chain as separated from the superficial ectoderm of the worm but still retaining on its ventral or more superficial side the ganglion cells g, and on its deeper side the nerve fibers /. (After Hatschek, 1888.) ticular part of this layer and the separation of this part from the rest of the skin, together with its migration into a deeper position in the animal, are rearrangements in which are retained a good deal of the original distribu- tion of the elements as seen in the more primitive sys- tems. Thus in the ccelenterates the elements of the re- ceptor-effector system consist of cell-bodies in the super- ficial part of the epithelium, followed by transmitting fibers, in the midregion, and finally completed' by muscle fibers in the deepest part (Fig. 3). When this system concentrates and breaks away from the skin, as it does in the annelids and the higher invertebrates, it is not sur- 206 THE ELEMENTAEY NERVOUS SYSTEM prising to find that the muscle occupies much of the in- terior of the body and that in the nerve chains the cell bodies are on the side next the skin and the transmitting trunks on that away from this layer (Fig. 49). The same sequence occurs in the vertebrates except that in this in- stance the epithelial layer of the skin instead of being left on the exterior of the animal is infolded and lines the central canal of the nervous system whose surface is derived from the exterior of the animal. Proceeding from this surface (Fig. 50) with its epithelial covering toward the other surface of the spinal cord the grey matter or region of the ganglion cells and fibrillar material is first passed through and then the white matter, the region of nerve fibers where transmission is , , . ,, ,, , . ,-p-i ,1 tllC Ciller lUnCtlOn. 1JQUS 1116 ... , , SeqU6nC6 01 USSU6S eStaDilSneCl in the receptor-effector systems of ccelenterates impresses itself on the differentiated cen- tral organs of the higher invertebrates and even the vertebrates. When an attempt is made to follow out the differen- tiation of the nervous elements in the evolving central nervous organ, it is necessary to begin with some of the processes that show themselves first in the nerve-net. As already pointed out, the nerve-net in its most typical con- dition exhibits diffuse nervous transmission and in this respect is in strong contrast with the central nervous or- gans of the higher animals whose transmission capaci- ties are highly specialized and limited. A kind of polar- 8ecFt!o°n 0? •^?»c°ord%tfrrvveerrt8e! brate (salamander); c, central canal; e, epithelium; g, grey substance com- posed of ganglion cells and fibrillar material; u>, whits substance or nerve CONCLUSION 207 ity that is indicative of this specialized state is well known to occur, however, in certain nerve-nets. For instance, it has been pointed out that in the tentacles of sea-anem- ones transmission is much more freely accomplished in a proximal direction than in a distal one. This was shown to result from the direction taken by the nerve- fibrils in the nerve-net of the tentacles. Probably a sim- ilar aggregation and alignment of fibers of the net into definite tracts is responsible for the simple reflex by which a stimulation of the tentacles of a sea-anemone by food will cause its transverse mesenteric muscles to con- tract and thus open its oesophagus. In some such way as this a diffuse nerve-net may be pictured as converted into the beginnings of a central nervous organ in that its network of transmitting fibers can be supposed to be re- arranged into systems of tracts whose connections are such that transmission becomes at once limited and specialized. But such a polarized derivative of a nerve-net is prob- ably still far from even a simple central nervous organ. Such a polarized nerve-net is to be met with in the my en- teric plexus of the vertebrate intestine, but the remote- ness of this structure from a central organ is well under- stood when we compare its activities with those of the vertebrate brain. The polarized nerve-net transmits in either direction. Its chain of protoneurones carry im- pulses up or down the series with indifference. This is in strong contrast with the state of these cells when they have arrived at the condition of neurones. It has long been known that in the spinal cord of vertebrates im- pulses may be sent over dorsal neurones into the cord to reappear in ventral neurones and excite muscle contrac- tion, but that it is impossible by stimulating ventral neu- 208 THE ELEMENTARY NERVOUS SYSTEM rones to excite the corresponding dorsal elements. The course that is open to transmission in one direction is not open in the opposite direction. This state occurs ap- parently wherever one neurone joins another and, as it is known not to be characteristic of transmission within a single neurone, it is assumed to be peculiar to the region in which the impulse is transmitted from one neurone to another. The special mechanism resident in this situa- tion and valve-like in its action is known as the synapse. It limits polarization in that it allows nervous transmis- sion to occur only in one of the two possible directions. It is perhaps the most general and definite criterion of a central nervous organ. It would be an extremely im- portant step if an easy test for the presence of synapses could be established. Anatomically this is probably quite impossible because of the very minute size of the struc- tures concerned. Physiologically it might be attained, and already an interesting suggestion in this direction has been made by Moore (1917). It has long been known that strychnine greatly heightens the reflex excitability of many animals and it has been commonly assumed that this action is due to the reduction, under the influence of this drug, of the synaptic resistances. Hence strychnine and other drugs having related effects may be used as a test for the presence of synapses. From this standpoint Moore's results are of extreme interest, for he has found that the drug has no effect on the neuromuscular re- sponses of coalenterates, a slight one on echinoderms, a much greater one on crustaceans and mollusks, a series that leads up to the well-known condition in vertebrates and suggests in its continuity that the effects are depen- dent upon the appearance and degrees of differentiation of the synapse. CONCLUSION 209 The conversion, therefore, of a system of protoneu- rones without synaptic interruptions into one of neu- rones related through synapses is the essential step in passing from a nerve-net to its derivatives, a true central nervous organ. In animals in which nerve-nets prevail, the lower invertebrates, the embryonic cells that give rise to their protoneurones are in the course of their devel- opment near neighbors so that the intimate relations which they bear to one another in the final net may after all be an expression of that closeness of relation that has existed between them from their embryonic states onward. It is not impossible that protoneurones that are united with each other in the nerve-net retain their strands of connection from embryonic stages when in the course of cell division they were never really entirely separated. With the neurones of the differentiated central nerv- ous system the case is usually very different. In the em- bryo of one of the higher animals the neurones are repre- sented by cells without processes, the neuroblasts, which, as they differentiate into neurones, do so in part by the production of extensive processes. In this way two neu- roblasts that lie at almost the opposite ends of the devel- oping nervous system may by means of their processes connect directly with each other as neurones in the adult. Thus certain neuroblasts in the developing cerebral cor- tex of mammals differentiate into neurones by throwing out processes that eventually reach in a remote part of the spinal cord motor neurones whose neuroblasts could have had no possible connections with those in the cortex. Both sets of cells are in the beginning widely separated and their final union depends upon the cell processes that have grown from one set to the other. How intimate this 14 210 THE ELEMENTARY NERVOUS SYSTEM union is, is not known, but that it is not a union of the kind seen in the nerve-net is quite clear. As was indi- cated in the discussion of synaptic transmission, the union is not complete, for it transmits in only one of the two possible directions. In picturing the conversion of a nerve-net into a central nervous organ, the longer and longer stretches over which nervous connections must be established as the higher and larger animals have been evolved must be kept in mind as a process which, as met with in embryonic life, implies a growing together of dis- A FIG. 51. — Stages in the differentiation of sense cells; A, sensory protoneurone from a coelenterate; B, sensory neurone from a mollusk; C, primary sensory neurone from a verte- brate. In each instance the peripheral end of the cell is toward the left, the central toward the right. tantly located cells, a process that thus gives ample op- portunity for the establishment of such relations as are seen in the synapse and which, therefore, has been of fun- damental importance in the evolution of the highest form of nervous system. The types of cells that the synaptic nervous system has received from the nerve-net are easily designated. The most primitive nerve cell from the standpoint of ani- mal phylogeny is the sense-cell, or receptive cell, such as occurs in the sensory epithelium of the coelenterates (Fig. 51, A ). In this type of cell, the cell body with its nucleus lies in the epithelial portion of the skin of such animals. CONCLUSION 211 This sensory protoneurone, for such it may be called, has been inherited by the higher invertebrates in an unmodi- fied form except that in some, as the mollusks for in- stance, its cell body shifts from a position in the superfi- cial epithelium into a deeper or subdermal one (Fig. 51, B). Finally in the vertebrates the cell body comes to occupy a still deeper position as the whole element as- sumes the form of a primary sensory neurone in that group (Fig. 51, C). A B FIG. 52. — Stages in the differentiation of nerve cells; A, protoneurone from the nerve- net of a coelenterate; B, motor neurone of an earthworm; C, primary motor neurone of a ver- tebrate. In B and C the receptive end of the neurone is toward the left, the discharging end toward the right. The protoneurones of the nerve-net, the second type of nervous element to be differentiated, give rise to the remaining neurones of the synaptic system. The general change that they undergo is relatively simple. As pro- toneurones in the coelenterate nerve-net, their cell bodies usually possess a central position in relation to the radi- ating system of nervous processes to which they give rise (Fig. 52, A). With the introduction of polarity and the synaptic relations, the cell body migrates toward the receptive end of the cell (Fig. 52, B), at which it is com- monly located in the fully differentiated neurone (Fig. 52, C). Neurones of this type fall under one or other of 212 THE ELEMENTARY NERVOUS SYSTEM two classes, primary motor neurones or intermmcial neu- rones. The primary motor neurones, the more primitive of the two, extend from the central organ to a muscle some of whose fibers they control. The internuncial neurones connect one part of the central apparatus with another and their courses, therefore, are entirely within the lim- its of the central organ. The more primitive central nervous organs such as those in the worms consist of lit- tle more than primary sensory and motor neurones. These of themselves constitute a sufficient basis for the simple types of reflex. But besides these there are a few internuncial neurones for internal connections. Above the worms in such forms as the arthropods and mollusks, the internuncial neurones become proportionally more abundant than the other types of neurones and in the ver- tebrates the internuncial type composes the chief mass of the central organs, a feature that reflects the enormous development of associative operations carried out by the vertebrate as compared with the invertebrate. But the nervous system of the higher animals is not only a system of tissues derived from a small group of especially sensitive cells associated with a still more primitive element, the musculature. It is a system that in its more differentiated examples has appropriated to itself certain other elements of the body than those that can be looked upon as direct descendants from an original source. In the vertebrates, at least, the primary sensory neurones have in some instances, as in the ear and in the organs of taste, appropriated ectodermic cells from the ordinary integument, and these cells, as taste buds and auditory hair cells, have come to form a constituent part of the sensory mechanism. Without doubt these appro- priated cells add to the effectiveness of the apparatus CONCLUSION 213 concerned. Thus the free-nerve terminations concerned with touch in the vertebrate skin require a much more vig- orous mechanical stimulus than do the auditory hair cells in the ear, cells that have been appropriated by just such a system as that seen in the tactile terminations. The same is true of the chemical senses. The free termina- tions of the primary sensory neurones of the human mouth (Fig. 53, A) are sensitive to solutions of ethyl al- cohol not less in concentration than 5 molecular, whereas those connected with taste buds (Fig. 53, B) are stimu- L iji l> & FIG. 53. — The appropriation of secondary sensory cells by primary sensory neurones in the vertebrates; A, primary sensory neurone associated with the mouth surface of a ver- tebrate and concerned with the common chemical sense of that surface; B, a primary sen- sory neurone that lias appropriated secondary sensory cells as represented by those in a taste bud in the vertebrate mouth. In each example the peripheral end of the neurone is toward the left, the central toward the right. lated by a solution of concentration 3 molecular (Parker and Stabler, 1913). Hence wherever a primary sensory neurone has appropriated secondary sensory cells, there is apparently an increase in the sensitiveness of the re- sulting1 combination. Not only have the sensory or afferent neurones of the differentiated nervous system appropriated cells not originally a part of the system, but the efferent neurones have also associated themselves with effectors not in the beginning under nervous control. Just as muscle arose as an independent effector, so other like elements have made their appearance in the animal series. To these categories belong cilia, nettle cells, chromatophores, luminous organs, glands, and so forth. Electric organs 214 THE ELEMENTARY NERVOUS SYSTEM are probably not of this nature, for in all cases they ap- pear to be modified muscles rather than independently produced structures. The others, however, give every evidence of independent origin and still exist in many animals as structures unassociated with nerves. Cilia proper and nettle cells seem never to have come under nervous domination, but chromatophores, luminous or- gans, and glands have all in one instance or another been appropriated by the nervous system. Through this agency they have been more or less completely removed from the class of organs whose action is called forth by direct stimulation and have been made subservient to the nervous system. Thus the system that arose secondarily around an independent effector muscle, has in the end gained such supremacy as to take to itself a number of independent effectors, any one of which might in the be- ginning have served as the nucleus around which the first nervous tissue could have taken origin. LITERATURE 215 LITERATURE ALEXANDROWICZ, J. S. 1913. Zur Kenntnis des sympathischen Nerven- systems eininger Wirbellosen. Zeitschr. allg. Physiol., Bd. 14, pp. 358-376, Taf. 12-13. ALLABACH, L. F. 1905. Some Points Regarding the Behavior of Metridimn. Biol. Bull., vol. 10, pp. 35-43. ALLMAN, G. J. 1863. Notes on the Hydroida. Ann, Mag. Nat. Hist., ser. 3, vol. 11, pp. 1-12. ALLMAN, G. J. 1871-1872. A Monograph of the Gymnoblastic or TubuJarian Hydroids. London, 4to, xxiv 4- 450 pp. APATHY, S. 1897. Das leitende Element des Nervensystems und seine topographischen Beziehungen zu den Zellen. Mitth. zool. Sta. Neapel, Bd. 12, pp. 495-748, Taf. 23-32. ASHWORTH, J. H. 1899. The Structure of Xenia hicksoni, nov. sp., with some Observations on Heteroxenia elizabethce, Kolliker. Quart. Jour. Micr. Sci., ser. 2, vol. 42, pp. 245-304. AUER, J., and MELTZER, S. J. 1909. The Influence of Calcium upon the Pupil and the Pupillomotor Fibers of the Sympathetic Nerve. Amer. Journ. Physiol., vol. 25, pp. 43-65. BAGLIONI, S. 1913. Physiologie des Nervensystems. Winterstein: Handbuch der vergleichenden Physiologie, Bd. 4, pp. 23^450. BANCROFT, F. W., and ESTERLY, C., 0. 1903. A Case of Physiological Polarization in the Ascidian Heart. Univ. California Pub., Zool., vol. 1. pp. 105-112. BAUER, V. 1910. Ueber die anscheinend nervose Regulierung der Flimmerbewegung bei den Rippenquallen. Zeit. allg. Phyisol., Bd. 10, pp. 231-248. BETHE, A. 1903. Allgemeine Anatomie und Physiologie des Nerven- systems. Leipzig, 8vo, viii + 488 pp., 1 Taf. BETHE, A. 1908. Die Bedeutung der Elektrolyten fiir die rhythmisehen Bewegungen der Medusen (I. Theil). Arch. ges. Physiol., Bd. 124, pp. 541-577. BETHE, A. 1909. Die Bedeutung der Elektrolyten fiir die rhythmisehen Bewegung der Medusen (II. Theil). Arch. ges. Physiol., Bd. 127, pp. 219-273. BIDDER, G. 1896. The Collar-cells of Heterowsla. Quart. Journ. Micr. Sci., new ser., vol. 38, pp. 9-43, pi. 2. BOHN, G. 1906 a. Sur les courbures dues a la lumiere. Compt. rend. Soc. Biol, Paris, ann. 1906, tome 2, pp. 420-421. BOHN, G. 1906 b. La persistance du rythine des marees chez 1' Actinia equina. Compt. rend. Soc. Biol., Paris, ann. 1906, tome 2, pp. 661-663. 216 THE ELEMENTARY NERVOUS SYSTEM BOHN, G. 1907 a. L'influence de 1'agitation de 1'eau sur les Actinies. Compt. rend. Soc. Biol., Paris, ann. 1907, toine 1, pp. 395-:!:^. BOHN, G. 1907 b. Le rythme nycthemeral chez les Actinies. Compt. rend. Soc. Biol., Paris, ann. 1907, tome 1, pp. 473— 176. BOHN, G. 1907 c. Introduction a la Psychologic dcs animaux a sy- metrie rayonee. I Memoire. Les etats physiologiques des Actinies. Bull. Inst. Gen. Psych., Paris, tome 7, pi». 81-129, 135-182. BOHN, G. 1908 a. De ['influence de roxyi^ene dissous sur les reactions des Actinies. Compt. rend. Soc. Biol., Paris, ann. 1903, tome 1, pp. 1087-1089. BOHN, G. 1908 b. Les facteurs de la retraction et de 1'epanouissement des Actinies. Compt. roul. Soc. Biol., Paris, anu. 1908, tome 1, pp. 1163-1166. BOHN, G. 1908 c. L'epanouissement des Actinies dans les milieux aspbyxiques. Compt. rend. Soc. Biol., Paris, ann. 1908, tome 2, pp. 317-320. BOHN, G. 1909. Les rythmes vitaux chez les Actinies. Compt. rend. Assoc. Frang. Av. Sci., 37 Sess., pp. 613-619. BOHN, G. 1910 a. Comparison entre les reactions des Actinies de la Mediterranee et celles de la Manche. Compt. rend. Soc. Biol., Paris. ann. 1910, tome 1, pp. 253-2.V>. BOHN, G. 1910 b. Les reactions des Actinies aux basses temperatures. Compt. rend. Soc. Biol., Paris, ann. 1910, tome 1, pp. 964-960. BOHN, G., et PIERON, H. 1906. Le rytbrne des marees et la phenomene de 1'anticipation rellexe. Compt. rend. Soc. Biol., Paris, ann. 1906, tome 2, pp. 660-661. DU BOIS-REYMOND, R. 1914. Physiologic der Bewegung. Winterstein : Handbuch der vergleichenden Physiologic, Bd. 3, Hiilfte 1, Teil 1, pp. 1-248. BRONN, H. G. 1860. Die Klassen und Ordnungen der Strablentbiere (Actinozoa). Leipzig, Svo, 434 pp. VON BRUCKE, E. T. 1916. Versuche an ausgeschnittenen und reiin- plantierten Flimmerscbleimbaut-Stiieken. Arch. gcs. Physiol., Bd. 166, pp. 45-54. BURROWS, M. T. 1912. Rhythmical Activity of Isolated Heart Muscle Cells in Vitro. Science, new ser., vol. .'Hi. pp. 5)0-92. CANNON, W. B., and BURKET, I. K. 1913. The Endurance of Anemia by Nerve Cells in the Myciiteric Plexus. Amer. Jour. Physiol., vol. 32, pp. 347-357. CANNON, W. P>., and \\'. \sinu i;\. A. I,. l!M2. An lv\j>laiiatioii of Hun- ger. Amer. Jour. Phi/siol., vol. 29, pp. 1 11-454. CARLSON, A. J. .11)01. The NVmms Origin of the Heart-beat in Lim- ulus, and the Nervous Nature of Coordination or Conduction in the Heart. Amer. Jour. Physiol., vol. 12, pp. 67-71. LITERATURE 217 CARPENTER, F. W. 1910. Feeding Reactions of the Rose Coral (Iso- phyllis). Proceed. Amer. Acad. Arts, Sci., vol. 46, pp. 149-162. CHESTER, W. M. 1912. Wound Closure and Polarity in the Tentacles of Metridium marginatum. Jour. Exp. Zool., vol. 13, pp. 451^170. CHUN, C. 1880. Die Ctenophoren des Golfes von Neapel. Fauna Flora Neapel, Monog. 1, xviii + 313 pp., 18 Taf. CLAUS, C. 1S78. Studien iiber Polypen und Quallen der Adria. Denkschr. Akad. Wiss., Wien, Bd. 38, pp. 1-64, Taf. 1-11. COBB, P. 1918. Autonomous Responses of the Labial Palps of Ano- donta. Proceed. Nat. Acad. Sci., vol. 4, pp. 234-235. CROZIER, W. J. 1916. The Rhythmic Pulsation of the Cloaca of Holo- thurians. Jour. Exp. Zool., vol. 20, pp. 297-356. EIMER, T. 1878. Die Medusen physiologisch und morphologisch auf ihr Nervensystems untersurcht. Tubingen, 4to, viii -}- 277 pp., 13 Taf. , ENGELMANN, T. W. 1904. Das Herz. Leipzig, 44 pp. FISCHER, P. 1888. Description d'une nouvelle espece du genre Ed- wardsia Quatrefages. Bull. Soc. Zool, France, ann. 1888, vol. 13, pp. 22-23. FLE.URE, H. J., and WALTON, C. L. 1907. Notes on the Habits of Some Sea-anemones. Zool. Am., Bd. 31, pp. 212-220. FOSTER, M. 1901. Lectures on the History of Physiology. Cambridge, 8vo, 310 pp. GEE, W. 1913. Modifiability in the Behavior of the California Shore- anemone Cribrina xanthogrammica Brandt. Jour. Anim. Beha- vior, vol. 3, pp. 305-328. GOSSE, P. H. 1860. A History of the British Sea-anemones and Corals. London, Svo, xl -)- 362 pp. GROSELJ, P. 1909. Untersuchunger iiber das Nervensystem der Aktin- ien. Arb. Zool. Inst., Wein, Tom. 17, pp. 269-308. GUTH, E. 1901. Untersuchungen iiber die directe rnotorische Wirkung des Lichtes auf den Sphincter pupilla? des Aal- und Froschauges. Arch. ges. Physiol, Bd. 85, pp. 119-142. HADZI, J. 1909. Ueber das Nervensystem von Hydra. Arb. Zool. Inst., Wien, Tom. 16, pp. 225-268. HAIME, J. 1854. Memoire sur le Cerianthe (Cerianthus membrana- ceus). Ann. Sci. Nat., Zool., ser. 4, tome 1, pp, 341-389, pi. 7-8. HARGITT, C. W. 1907. Notes on the Behavior of Sea-Anemones. Biol. Bull, vol. 12, pp. 274-284. HARVEY, E. N. 1912. The Question of Nerve Fatigue. Carnegie Inst., Washington, Tear Book, No. 10, pp. 130-131. HATSCHEK, B. 1888. Lehrbuch der Zoologie. Jena, Svo, iv -4- 432 pp. (incomplete). 218 THE ELEMENTAEY NERVOUS SYSTEM HAVET, J. 1901. Contribution a 1'etude du systeme nerveux des Ac- tinies. La Cellule, tome 18, pp. 385-419. HECHT, S. 1918. The Physiology of Ascidia atra Leuseur. Amer. Jour. Physiol., vol. 45, pp. 157-187. VON HEIDER, A. R. 1877. Sagartia troglodytes Gosse, em Beitrag zur Anatomic der Actinien. Sitzb. K. Akad. Wiss., Wien, math.-nat. Cl., Bd. 75, Abt. 1, pp. 367-418. VON HEIDER, A. R. 1879. Cerianthus membranaceus Haime. Sitzb. K. Akad. Wissen., Wien, math.-nat. Cl., Bd. 79, Abt. 1, pp. 204-254. VON HEIDER, A. R. 1895. Zoanthus chierchice n. sp. Zeit. wiss. Zool., Bd. 59, pp. 1-28. HEJJTEL, E. 1906. Experimenteller Beitrag zur Kenntnis der Pupillen- verengerung auf Lichtreize. Arch. Opth., Bd. 65, pp. 106-134. HERTWIG, 0., und HERTWIG, R. 1878. Das Nervensystem, und die Sin- nesorgans der Medusen. Leipzig, 4to, x 4- 186 pp., 10 Taf. HERTWIG, 0., und HERTWIG, R. 1879-1880. Die Actinien anatomisch und histologisch mit besondei'er Beriicksichtigiing des Nerven- muskelsy stems untersucht. Jena. Zeit., Bd. 13, pp. 457-460, Bd. 14, pp. 39-80. HESS, C. 1913. Neue Untersuchungen zur vergleichenden Physiologic des Gesichtsinnes. Zool. Jahrb., Abt. Allg. Zool. Physiol., Bd. 33, pp. 387-440. HESSE, R. 1895. Ueber des Nervensystem und die Sinnesorgane von Rhizostoma Cuvieri. Zeit. wiss. Zool., Bd. 60, pp. 41H57, Taf. 20-22. HICKSON, S. J. 1895. The Anatomy of Alcyonium digitatum. Quart. Jour. Micr. Sci., ser. 2, vol. 37, pp. 343-388. His, W. 1886. Zur Geschichte des menschlichen Riickenmarkes und der Nervenwurzeln. Abh. math.-phys. Cl. sachs. Gesell. Wissen., Bd. 13, pp. 479-513. His, W., JR. 1893. Die Entwickelung des Herznervensystems bei Wir- bilthieren, Abh. math.-phys. CL sachs. Gesell. Wissen., Bd. 18, pp. 1-65, Taf. 1-4. HOOKER, D. 1911. The Development and Function of Voluntary and Cardiac Muscle in Embryos without Nerves. Jour. Exp. Zool., vol. 11, pp. 159-186. HOWELL, W. H. 1906. The Cause of the Heart Beat. Jour. Amer. Med. Assoc., vol. 46, pp. 1665-1670, 1749-1753. HUNTER, G. W. 1902. The Structure of the Heart of Molgula man- hattensis (Verrill). Anat. An