THESE VOLUMES ARE DEDICATED TO THE MEN AND WOMEN OF OUR TIME AND COUNTRY WHO BY WISE AND GENEROUS GIVING'r HAVE ENCOURAGED THE SEARCH AFTER TRUTH IN ALL DEPARTMENTS OF KNOWLEDGE STUDIES IN GENERAL PHYSIOLOGY STUDIES IN GENERAL PHYSIOLOGY JACQUES LOEB FORMERLY OF THE DEPARTMENT OF PHYSIOLOGY NOW PROFESSOR OF PHYSIOLOGY IN THE UNIVERSITY OF CALIFORNIA THE DECENNIAL PUBLICATIONS SECOND SERIES VOLUME XV PART I CHICAGO THE UNIVERSITY OF CHICAGO PRESS 1905 Copyright 1905 BY THE UNIVERSITY OF CHICAGO PKEFACE I SHOULD not have had the courage to offer these volumes to the public, had not requests repeatedly come to me from physicians and biologists to render my publications, which are widely scattered, more easily accessible. There- fore, when the editor of the "Decennial Publications of the University of Chicago" invited me to make a contribution to the series, I mentioned to him, not without hesitation, the idea of collecting and republishing my papers on General Physiology. Through his initiative and kind as- sistance the idea has been carried out. No one will expect that a collection of papers on very diverse subjects can form attractive reading matter. Yet I may mention, by way of an apology, that, in spite of the diversity of topics, a single leading idea permeates all the papers of this collection, namely, that it is possible to get the life-phenomena under our control, and that such a control and nothing else is the aim of biology. Thus the reader will notice that in a series of these publications I have tried to find the agencies which determine unequivocally the direction of the motion of animals, and he will also notice that I consider a complete knowledge and control of these agencies the biological solution of the metaphysical problem of animal instinct and will. In taking up the problem of regeneration I started out with the idea of controlling these phenomena, and considered it my first aim to find means by which one organ could at desire be caused to grow in the place of another organ. Thus the experiments on heteromorphosis originated. , As far as the problem of fertilization is con- cerned, it seemed to me that the first step toward its solution should consist in the attempt to produce larvse artificially from unfertilized eggs in various classes of animals. ix 852 PREFACE It seemed desirable that the reader should be spared an undue amount of repetition, and for this reason a number of publications are omitted from this collection, and those printed are in many cases shortened. Among the papers which have been omitted are the preliminary notices and all those papers of which I am not the sole author. Occasion- ally I have made additions in the form of footnotes. Such footnotes have always been marked by the addition of [1903] at the end. Only a small number of these papers appeared originally in English, namely, VII, XXI, XXVI-XXXV, and XXXVII. The other papers were translated from the German by Pro- fessor Martin H. Fischer, to whom I wish to express my sincere thanks. The credit as well as the responsibility for the translation belongs entirely to him. In the reading of the proof I was assisted by Dr. Fischer, Dr. Rogers, Dr. Bullot, and Dr. Bancroft. Mr. Rogers made the index for the first volume. To all these gentlemen my thanks are due. JACQUES LOEB. BERKELEY, CALIFORNIA, October 14, 1904. TABLE OF CONTENTS PART I I. The Heliotropism of Animals and its Identity with the Heliotropism of Plants - 1 II. Further Investigations on the Heliotropism of Ani- mals and its Identity with the Heliotropism of Plants - 89 III. On Instinct and Will in Animals 107 IV. Heteromorphosis 115 V. Geotropism in Animals 176 VI. Organization and Growth - 191 VII. Experiments on Cleavage - 253 VIII. The Artificial Transformation of Positively Helio- tropic Animals into Negatively Heliotropic and vice versa 265 IX. On the Development of Fish Embryos with Sup- pressed Circulation 295 X. On a Simple Method of Producing from One Egg Two or More Embryos Which Are Grown Together 303 XL On the Relative Sensitiveness of Fish Embryos in Various Stages of Development to Lack of Oxygen and Loss of Water 309 XII. On the Limits of Divisibility of Living Matter 321 XIII. Remarks on Regeneration 338 XIV. Contributions to the Brain Physiology of Worms 345 XV. The Physiological Effects of Lack of Oxygen - 370 Xll TABLE OF CONTENTS PART II XVI. The Influence of Light on the Development of Organs in Animals 425 XVII. Has the Central Nervous System Any Influence upon the Metamorphosis of Larvae? 436 XVIII. On the Theory of Galvanotropism 440 XIX. The Physiological Effects of Ions. I 450 XX. On the Physiological Effects of Electrical Waves 482 XXI. The Physiological Problems of Today - 497 XXII. The Physiological Effects of Ions. II - 501 XXIII. Why Is Regeneration of Protoplasmic Fragments without a Nucleus Difficult or Impossible ? 505 XXIV. On the Similarity between the Absorption of Water by Muscles and by Soaps - 510 XXV. On Ions Which Are Capable of Calling Forth Rhythmical Contractions in Skeletal Muscle 518 XXVI. On the Nature of the Process of Fertilization and the Artificial Production of Normal Larvae (Plutei) from the Unfertilized Eggs of the Sea- Urchin 539 XXVII. On lon-Proteid Compounds and Their Role in the Mechanics of Life-Phenomena. — The Poison- ous Character of a Pure NaCl Solution 544 XXVIII. On the Different Effects of Ions upon Myogenic and Neurogenic Rhythmical Contractions and upon Embryonic and Muscular Tissue 559 XXIX. On the Artificial Production of Normal Larvae from the Unfertilized Eggs of the Sea -Urchin (Arbacia) - 576 XXX. On Artificial Parthenogenesis in Sea-Urchins 624 XXXI. On the Transformation and Regeneration of Organs 627 XXXII. Further Experiments on Artificial Parthenogenesis and the Nature of the Process of Fertilization - 638 TABLE OF CONTENTS Xlll XXXIII. Experiments on Artificial Parthenogenesis in Annelids (Chsetopterus) and the Nature of the Process of Fertilization XXXIV. On an Apparently New Form of Muscular Irrita- bility (Contact-Irritability?) Produced by Solu- tions of Salts (Preferably Sodium Salts) Whose Anions Are Liable to Form Insoluble Calcium Compounds XXXV. The Toxic and the Antitoxic Effects of Ions as a Function of Their Valency and Possibly Their Electrical Charge XXXVI. XXXVIII. On the Methods and Sources of Error in the Experiments on Artificial Parthenogenesis 646 692 708 Maturation, Natural Death, and the Prolongation of the Life of Unfertilized Starfish Eggs (Asterias Forbesii) and Their Significance for the Theory of Fertilization 728 XXXVII. On the Production and Suppression of Muscular Twitchings and Hypersensitiveness of the Skin by Electrolytes 748 766 INDEX 773 PART I THE HELIOTROPISM OF ANIMALS AND ITS IDENTITY WITH THE HELIOTROPISM OF PLANTS1 I. INTRODUCTION I INTEND to show in the following pages that animal movements depend upon light in the same way as the move- ments of plants. It is a well-known fact that animals, when light falls on them, move toward the source of light, like the moth, or move away from it, like the earthworm. It is also well known that certain plant organs have a tendency to turn toward or from the source of light when illuminated from one side only. While the conditions which govern the behavior of plants toward light have been well analyzed, especially by Sachs, little has been done to investigate the conditions upon which depend the movements of animals toward a source of light. It is the purpose of this paper to fill this gap, and to enumerate the facts which show that in reality the animal motions called forth by light depend upon the same circumstances as the motions which light produces in plants. The effects of light which we intend to study are purely mechanical, inasmuch as they consist in changes in position, as well as in the direction and the sense of the progressive movements of living animals. Consequently we shall regard as essential such circumstances as can help to explain the mechanical effects of the light. These circumstances, as in the case of all stimulations, are of a double origin: first, those belonging to the stimulus — in this case the light; and, i Pamphlet, Wurzburg, 1889. 1 2 STUDIES IN GENERAL PHYSIOLOGY second, those belonging to the structure of the organism. So far as the light is concerned, the circumstance which controls the orientation of the animal and the direction of its movements is the direction of the rays falling upon the animal.1 The condition which is of importance on the part of the animal is the symmetrical shape of the body. Sachs discovered that all plant organs which have a radial structure are orthotropic (this means that they bend, when light strikes them on one side, until their longitudinal axes lie in the direction of the rays of light), but that all dorsiventral structures are plagiotropic, i. e., they place their surfaces perpendicular to the rays of light. Symmetrically situated points at the surface possess a quantitatively and qualitatively equal irritability. In this way the organ of a plant is mechanically forced to orient itself in such a way that the rays of light strike symmetrical points at equal angles to the surface. If the plant, as for example the swarm spore of algae, is capable of a progressive motion, it must of course, in order to maintain this position, move in the direction of the rays of light. This is, indeed, found to be the case. I shall now show that quite generally in animals the direction of the rays of light controls also the direction of those movements which are caused by light; that, in addi- tion., quite generally in animals their orientation depends i In these experiments it is presumed that the animals move under the influence of only one source of light. It is explicitly stated in this and the following papers that if there are several sources of light of unequal intensity, the light with the strongest intensity determines the orientation and direction of motion of the animal. Other possible complications are covered by the unequivocal statement, made and emphasized in this and the following papers on the same subject, that the main feature in all phenomena of heliotropism is the fact that symmetrical points of the photosensitive surface of the animal must be struck by the rays of light at the same angle. It is in full harmony with this fact that if two sources of light of equal intensity and distance act simultaneously upon a heliotropic animal, the animal puts its median plane at right angles to the line connecting the two sources of light. This fact was not only known to me, but had been demonstrated by me on the larvae of flies as early as 1887, in Wiirzburg, and often enough since. These facts seem to have escaped several of my critics. [1903] HELIOTEOPISM OF ANIMALS on the form of the body in so far as dorsiventral animals move with their median planes in the direction of the rays of light, in which position the rays fall upon symmetrically situated points of the surface of their bodies at nearly equal angles. In this way the fact that a moth flies into a flame turns out to be the same mechanical process as that by which the axis of the stem of a plant puts itself in the direction of the rays of light. In both cases, however — in the fatal flight of the moth as well as in the orientation of plants— one point remains unexplained, namely: how can the light so change the state of the protoplasm as to bring about the mechanical effects just mentioned? At present we are not able to form a clear idea of this. A second condition which has a determining influence upon the mechanical effects of light on plants is the refran- gibility of the rays. Sachs has shown that it is chiefly the more refrangible rays which are able to bring about move- ments in plant organisms. We shall see that quite gen- erally the more refrangible rays are also more effective mechanically in the animal kingdom. Thirdly, we shall prove that the orientation of animals as well as of plants takes place when the intensity of the light remains constant. Very often we observe, for example in our eyes, that a change in the intensity of the light acts as a stimulus. In addition to these essential considerations of the effects of light in the animal kingdom, the following factors play a r6le, namely: Fourthly, light causes the orientation of animals (as well as of plants) only within certain limits of intensity. Fifthly, temperature influences the movements of orientation in animals and plants toward light — which is true for all phenomena of stimulation. To sum up: The conditions which control the movements of animals toward light are identical, point for point, 4 STUDIES IN GENERAL PHYSIOLOGY with those which have been shown to be of paramount influence in plants. Aside from the problem of proving by suitable experi- ments the stated propositions, it is also necessary for us to show what role the orientation toward the light plays in the economy of life of an animal. I shall therefore first describe the experimental proofs of the identity of animal heliot- ropism with plant heliotropism, and then show by individual examples what role heliotropism plays in the economy of life of animals. To discuss the latter point it will be necessary also to describe briefly the other forms of irritability pos- sessed by an animal. In a short article which appeared in January, 1888, I described the principal laws upon which depends the orien- tation of animals to light, and the identity of these laws with those governing plant heliotropism.1 II. THE ESSENTIAL PHENOMENA AND LAWS OF HELIOTROPISM IN PLANTS Assuming that the reader is acquainted with the orienta- tion of plants toward a source of light, it will suffice at this place to call attention briefly to the essential facts which bear upon our subject. In so doing I shall follow the presenta- tion given by J. von Sachs in his lectures on plant physi- ology.2 Straight stems or roots of growing plants bend when light falls on them on one side only, or with greater intensity on one side than on the other, until their tips lie in the direc- tion of the rays of light. Those organs which turn toward the source of light are called positively heliotropic; those which turn from the light, negatively heliotropic. i " Die Orientierung der Thiere gegen das Licht (thierischer Heliotropismus)," Sitzungsberichte der Wiirzburger physikalisch-medicinischen Gesellschaft, January, 1888. 2 Vorlesungen iiber Pflanzen-Physiologie, 2d ed. (Leipzig, 1887). HELIOTROPISM OF ANIMALS 5 It was formerly believed that the bending of the positively heliotropic parts of plants was due to the fact that the side which was turned away from the light grew more rapidly, because plants when brought into the dark at first grow more rapidly than they do in the light. But it was proved in Sachs's laboratory that negatively heliotropic organs also grow more rapidly in the dark. Because of the similarity of the geotropic and heliotropic movement in plants, Sachs came to the conclusion that the direction in which the rays of light penetrate the plant tissue determines the orientation of the plant toward light. He also proved that not all the rays of the visible sun spectrum bring about heliotropic movements, but only, or at least chiefly, the more refrangible rays. The less refrangible rays, which are of importance in assimilation, are ineffective heliotropically. If the light be previously passed through a dark-blue ammoniacal solution of copper, which absorbs all the red, yellow, and a part of the green rays, the heliotropic bending occurs in the same way as in completely white light. If, however, the light passes through a saturated solution of potassium bichromate, which lets through only red, yellow, and a part of the green rays, " the heliotropic shoots remain straight and vertical, no matter how intense the light is which passes through the solution." Finally, if the light "is passed through a solu- tion of quinine sulphate, the fluorescence of which completely absorbs the ultra-violet rays, the heliotropic curvatures nevertheless appear — a proof that they are caused princi- pally by the visible blue and violet rays." The best proof of the theory that the direction of the rays of light controls the orientation of plants was found by studying freely moving plant organs, the swarm-spores of algse. These swarm-spores make progressive movements like animals, and Strasburger1 proved that they move in the i STBASBUEGEB, Wirkung des Lichtes und der Wdrme auf Schwarmsporen (Jena, 1878). 6 STUDIES IN GENERAL PHYSIOLOGY direction of the rays, to or from the source of light. The more refrangible rays alone exercise this effect on the swarm- spores. They behave in the light which has passed through an ammoniacal solution of copper just as in diffuse daylight. On the other hand, they are not affected by light which has passed through a potassium bichromate solution, by light from a sodium flame, or by the light coming through ruby glass. The chlorophyll-bearing protoplasm of cells moves under the influence of light.1 The chloroplasts of a thread alga, Mesocarpus, turn "their broad surfaces toward the sky so that the rays fall upon them at right angles. If the direction of the rays is changed, the chloroplasts turn so that their broad surfaces are again at right angles to the rays. Direct sun- light, however, causes the chloroplasts to assume another position — they place their surfaces parallel to the rays which strike them." According to modern plant physiology, the whole proto- plasm of a multicellular plant is to be conceived of as a continuous mass, as a single protoplasmic body.2 More recent investigations have shown that when a plant organ is illuminated, that side of the organ which becomes concave from the effect of the light becomes rich in protoplasm, while the opposite convex side becomes poor.3 Multicellular organs behave in this regard like unicellular ones. Thus it appears that the light forces the protoplasmic mass to move in such a way that positively heliotropic protoplasm wanders to the side of the organ which is turned toward the light, while negatively heliotropic protoplasm wanders to the opposite side.4 Should it turn out that this phenomenon really occurs i STAHL,, Botanische Zeitung, 1880. 2 SACHS, loc. cit., p. 94. 3 Wortmann expressed his observations in this way. It is possible that in reality protoplasm on the concave side is only more opaque than on the opposite side. This difference in optical appearance may simply be the expression of a difference in the size of the colloidal particles. [1903] * See WORTMANN, Botanische Zeitung, 1887. HELIOTROPISM OF ANIMALS in all cases, it would prove that the protoplasm of a multi- cellular plant behaves just like the naked, creeping plas- modium, which is also heliotropically irritable. III. SUMMARY OF THE MECHANICAL EFFECTS OF LIGHT IN THE ANIMAL KINGDOM WHICH ARE THUS FAR KNOWN I shall in this chapter summarize briefly the facts and views in regard to the movements of animals under the in- fluence of light, so far as they are known up to the present time. These may be divided into three groups: 1. Casual observations of the older authors (Reaumur, Trembley). These are unprejudiced records of simple obser- vations. 2. Modern investigations on the effects of light from an anthropomorphic standpoint. The movements of animals are not attributed to mechanical causes, but to supposed human sensations of the animals. 3. Investigations according to the method of Sachs, which, however, have been applied only to Protozoa. The last- named observations are the most important in these three groups. The earliest account of the effects of light on animals which I have found in the literature is by Reaumur.1 He found that moths which are attracted by the candle flame "do not fly from flower to flower during the day." Since he saw chiefly the males fly into the flame, he raised the question as to whether or not the female moths emit light like glow- worms. "Do not the females of the nocturnal Lepidoptera emit a light too feeble to make an impression on our eyes, but sufficiently strong to act on those of their males?" He had observed, evidently, that the males of the glow-worm which are attracted by the light to the aboral end of the females likewise fly into the light. Reaumur was, moreover, i REAUMUH, :?' moires pour servir d, Vhistoire des insectes, Vol. I, 1, p. 330 (Amster- d.-. -.1748). 8 STUDIES IN GENERAL PHYSIOLOGY convinced that the glow-worm living in the woods could see. He made a glass window in a tree in which such worms lived and noticed that the animals gave a start upon the approach of a burning candle. Trembley made far better experiments.1 He found that "water fleas" can be driven around in a circle by a moving candle : By the light of a wax taper I observed polyps to which during the day I had given many water fleas; in the evening there were left in the glass some which the polyps had not consumed. I noticed that most of them had collected on the side toward the candle. I changed the position of the taper, and they followed it. As I had moved its position repeatedly, and each time had seen that the water fleas followed it, I moved the taper slowly around the glass without stopping. They followed, and thus made several trips around it. I have had the opportunity of repeating this ex- periment several times. Trembley's observations on the effect of light on Hydra were made with great care. After he had repeatedly observed that the polyps moved to the "brightest" side of the glass, he placed "a glass containing many green polyps in a case which had an opening on one side about opposite the middle of the glass." He reports as follows concerning their behavior: When I placed the glass so that the opening in the case was turned to the light, the polyps always migrated toward that side of the glass which was opposite this opening, in such a way that together they made the figure of a gable. I often turned the glass around, and after several days I observed the polyps again at the opening arranged as before (in the form of a gable). To vary the experiment still further, I fixed the dark case so that the open- ing was at times straight, at other times inverted, and again the polyps arranged themselves together. After he had discovered that polyps which had been cut in two could "move, eat, and multiply," he tried to see "whether 1 TREMBLEY, Abhandlungen zur Geschichte einer Polypenart, transl. by GOTZE (Quedlinburg, 1791). HELIOTKOPISM OF ANIMALS 9 these pieces would turn toward the light in the same way as the undivided polyps." He cut a number of polyps in two: the anterior halves he placed in one glass, the posterior halves in another. He £ound "in oft-repeated experiments that the animals in both glasses collected in the brightest regions in the glass." These are, as far as I know, the only extended observa- tions to be found in the old physiological literature of the effects of light upon animals. For a long time no further study of the effects of light upon animals was made. Johannes Muller mentions, in the preface to his Physiologic des Gresichtssinnes, that he made "investigations on the in- fluence of colored light on the vital phenomena of plants and animals," but, as far as I know, the results of his investiga- tions were never published. The modern anthropomorphic observations were intro- duced by Paul Bert. Bert raised the question: Do all animals see the same rays that we see?1 He meant to ask whether all rays of the visible sun spectrum are able to bring about animal movements. An experiment with Daph- nia pulex was sufficient for Bert to settle this question. He projected a spectrum and found that the animals became restless in all positions of the visible spectrum: Mes daphnies erraient disperses d'une maniere & peu pres 6gale dans toute F^tendue du vase obscur, lorsque soudain je fis tomber sur la fente un rayon colore", un rayon vert. Aussitot elles s'agiterent, se groupment toutes dans la direction de la trainee lumineuse et un tres-grand nombre s'en vint se heurter, montant et descendant sans relache centre la paroi qui recevait la lumiere. Or, un semblable re"sultat fut obtenu pour toutes les regions du spectre visible. Le rouge, le jaune, le bleu, le violet meme atti- raient les daphnies. Seulement il fut facile de remarquer, qu'elles accouraient beaucoup plus rapidement au jaune ou au vert qu'a toute autre couleur. !BERT, Archives de physiologic, 1869. 10 STUDIES IN GENERAL PHYSIOLOGY On either side of the spectrum the animals remained at rest. In addition to this, Bert made another experiment. He had a spectrum projected on a trough, and observed how the animals distributed themselves over the different parts of the spectrum. L'immense majority se placa dans le jaune, le vert, Forange; une assez grande quantity se voyaient encore dans le rouge, un certain nombre dans le bleu, quelques-unes de plus en plus rares a mesure qu'on s'eloignait dans les regions plus re"frangibles du violet, au del& du rouge, au del& de 1'ultra-violet ; dans les regions invisibles, en un mot, on n'en trouvait que d'isole"es en promenade accidentelle. From these facts Bert concluded that Daphnia behaves in the spectrum much as a man would, who, when reading a book, would move into the brightest part of the spectrum, into the yellow light. Lubbock repeated Bert's experiment on Daphnia.1 One- half of a dish was covered by a yellow screen; the other half was left uncovered. In the uncovered half 1,904 animals collected, while 3,096 gathered under the yellow screen. From this Lubbock concludes that Daphnia has a "preference" for "yellow." But one would suppose that in the uncovered part of the dish there was at least as much yellow light as under the yellow screen; or did the majority "hate" the blue light? When Lubbock covered one-half of the trough with blue glass and left the other uncovered, he found 2,0-16 animals under the blue glass, and 2,954 in the uncovered part of the trough. Whether one is to conclude from this that blue light is in the sense of Lubbock "disagreeable" to Daphnia is not stated. When half of the trough was covered with red glass, there collected 1,928 animals under the red glass, while 3,072 collected in the uncovered por- i LDBBOCK, " Die Sinne und das geistige Leben der Thiere," Internationale wissenschaftliche Bibliothek, Vol. LXVII (1889). HELIOTKOPISM OF ANIMALS 11 tions of the dish. When half of the vessel was covered with an opaque porcelain screen, Lubbock found 2,048 animals collected under it, and 2,932 animals in the un- covered half. From these and similar experiments Lubbock concludes that the animals have a decided preference for yellow light. I also have made some experiments on the effects of rays of different refrangibility on Daphnia, and found that when the more refrangible rays (blue and violet) fell upon the animals they hastened to the source of light and moved up and down on the light side of the vessel. When I made the same experiment with the less refrangible rays, the effect was weak or did not take place at all. The result conforms with other facts which are to be described later. I shall, therefore, not revert to the Daphnia and their alleged ''preference for yellow." Lubbock has employed a similar method in his experi- ments on wingless ants;1 these, however, led to much more fruitful results than his experiments on Daphnia. In an experiment in which a vessel was covered with strips of red, green, yellow, and violet glass he found that 890 animals collected under the red glass, 544 under the green, 495 under the yellow, and only 5 under the violet. There is no doubt in this case that the animals collected under those glasses where they were struck by the less refrangible rays. Other experiments showed that red glass acts like an opaque body. The observation of Lubbock that ants avoid the ultra- violet part of the spectrum is also worthy of note. For the sake of completeness the experiments of Lubbock on bees and wasps must be mentioned, in which it was found that under otherwise similar conditions blue objects smeared with honey were preferred to those of another color. 1 LUBBOCK, " Ameisen, Bienen und Wespen," ibid., 1883. 12 STUDIES IN GENERAL PHYSIOLOGY The most extended experiments on the influence of light on the orientation of animals were made by Graber.1 His "comparative studies on light-sensations" (Vergleichende Licht-Gefiihl-Studien), as he called his investigations, cover about fifty species. His method is that followed by Lub- bock. The faultiness of this method and the errors of interpre- tation of the results obtained stand out more clearly in Graber' s writings than in Lubbock's. Graber covers one- half of a vessel with a partially or completely opaque screen, and after a time notes how the animals are dis- tributed in the vessel. If most of the animals are under the opaque screen, Graber says that they are "fond of the dark" and "hate the light;" or in the reverse case, that they are "fond of the light" or of "the white" and "hate the dark." He therefore uses the conceptions of "white" or "bright" and "dark," which designate certain effects of liylii upon a human being for the conceptions of great or small intensity of the light; and in saying that animals which "prefer the light" also "hate the darkness" he makes a second mistake in that he maintains that strong and weak light have opposite effects. We shall see, however, that these effects are similar and differ only in degree. He makes the same mistake in experimenting on rays of different refrangibility. The most important among the facts ob- served by him in this connection is this, that animals which "prefer the light" with a few exceptions also "prefer" blue, while those which "hate the light" "prefer" red. His ideas are expressed in the following remarks, which, however, I do not fully understand: The question arises as to the cause of this truly striking rela- tion between the love for white light and for blue light, on the one hand, and between the dislike for white light and for blue light, i Grundlinien zur Erforschung des Helligkeits- und Farbensinnes der Thiere (Prag, 18&4). HELIOTROPISM OF ANIMALS 13 on the other hand. If the law had reference only to white light, and not also to colored light — red, blue, etc. — which is, however, by no means always the case, one might at first be inclined to be- lieve that the animals which prefer red avoid mixed light because it contains many of the hated short waves of the blue and violet light; for this very reason it would be more agreeable than dim light to the animals which prefer blue, for dim mixed light is poor in all rays, and therefore also in blue. Yet the objection might be raised against this explanation that mixed light contains as much red for those animals which prefer red as it contains blue for those animals which prefer blue. Yet this objection could again be weakened by the assumption that, since the animals which prefer red also prefer darkness, they prefer a minus of their chosen color to a plus of the color they dislike. Graber finally considers it best "to await further investi- gations in a field where great darkness still prevails." We see that Graber in regard to the effects of monochromatic light again establishes a contrast in effects where, as we shall see, a similarity exists. Graber was prevented from cor- rectly interpreting his results by attributing the movements of animals to sensations instead of to physical causes. If he had given up the anthropomorphic standpoint, he would soon have discovered that his experiments show that the more refrangible rays are more effective in causing the orientation of an animal than the less refrangible ones. In none of the investigations of Bert, Lubbock, or Graber has the influence of the direction of the rays on the orienta- tion been studied. Graber, for example, took it for granted that an animal moves to the light because, as he expressed it, "it is fond of the light" or "the white." If it moves in the opposite direction, it "is fond of the dark." Lubbock remarks incidentally that "ants do not like light in their nests, probably because they do not deem it safe." This sums up the opinions and results of the authors who sought to explain anthropomorphically the phenomena which interest us here. 14 STUDIES IN GENEKAL PHYSIOLOGY Finally, I have to mention the heliotropic investigations on Infusoria which were made along the lines mapped out by Sachs. To bring these investigations before the reader I shall describe the more important observations which have been made on Euglena. The influence of the direction of the rays of light on these Infusoria was first demonstrated by Stahl:1 Those individuals which did not swim about freely remained with their pointed posterior ends attached to the cover-glass or to other objects, while their free anterior ends were, according to con- ditions, either turned toward or away from the source of light. The longitudinal axes of both the motile and sessile Euglenge coincided as nearly as possible with the direction of the rays of light. The motionless ones behaved like the free-swimming ones whenever the direction or intensity of the light was suddenly changed, except that they reacted more slowly. If, for example, the glass slip was suddenly rotated through an angle of 180°, the position which the animals occupied originally with reference to the source of light was slowly reassumed, while the swimming individuals left their former path and moved in the original direction toward the light immediately after a change in its direction. Engelmann studied in Euglena the relation between the effect of the rays of light and their refrangibility.2 After he had established the fact that when a drop of Euglenae is only partially illuminated the animals gradually accumulate in the lighted area, he brought the animals into a micro- spectrum. Here they collected on the more refrangible side of the spectrum. The orientation of Euglena therefore depends on the direction of the rays, and especially on that of the more refrangible ones. It must finally be men- tioned that the anterior ends of the Infusorise are most sen- sitive to light; yet the pigment spot is not, as might be supposed, the most sensitive, but the colorless protoplasm in front of this. Besides these direct effects of light in phenomena of 1 Botanische Zeitung, 1880. 2 Pflugers Archiv, Vol. XXIX (1882). HELIOTROPISM OF ANIMALS 15 orientation, which alone interest us here, there are also cer- tain indirect effects on the orientation of low forms of life. These were also first observed by Engelmann. When the supply of oxygen is cut off from certain chlorophyll-bearing organisms, they remain in that part of the spectrum in which assimilation takes place. In water with its normal amount of oxygen, as Engelmann found, Stentor viridis, Bursaria, and the green slipper animalcule do not react to light.1 If, however, the supply of oxygen from without is interfered with, "the insufficient supply can be compensated for by a production of oxygen by the chlorophyll granules within the mesoplasm." Under these conditions the animals return to the light side of the drop when they accidentally get into the shady part. When the animals are brought into a micro- spectrum, they collect in those regions which promote assimi- lation. The opposite effect takes place, however, when the supply of oxygen from without exceeds the normal. When Engelmann passed a stream of pure oxygen through the water, the animals moved from the lighted into the shaded part of the drop. Such an indirect orientation toward light as is determined by assimilation is shown also in the behavior of the purple bacteria.2 These, as Engelmann found, collect in those regions of the spectrum which are most absorbed by the coloring matter of the bacteria. These are the most important facts which up to this time are known concerning the influence of light on the orienta- tion of animals. Thus far only the observations made on Infusoria are sufficient to warrant the conclusion that ani- mal movements depend on light in the same way as the movements of plants. In the rest of the animal kingdom either the facts necessary for this conclusion are lacking, or false statements and conceptions are prevalent. So far as ., p. 387. 2 ENGELMANN, Botanische Zeitung, 1888. 16 STUDIES IN GENERAL PHYSIOLOGY the latter are concerned, it is wrong, as we shall see, to say that certain animals "are fond of the light" and seek those regions in space where light is most intense, while others "are fond of the dark" and betake themselves to those regions which are darkest. In contradiction of this idea I shall prove that the direction of the progressive heliotropic move- ments of animals is determined solely by the direction of the rays, no matter whether the animals move from regions in which light is less intense to those in which it is more intense, or vice versa. Further than this, it is fundamentally wrong to say that an assumed "preference for color" determines the orientation of animals toward rays of different ref rangibilities ; that, as Graber says, the animals which "are fond of blue" "hate red," and that those which "are fond of red" "hate blue." In contradiction of this iclea I shall prove that there are no animals which "are fond of" red or "hate" blue, but only such as move toward a source of light or away from it ; and that these movements occur in the same way under the influence of the more refrangible rays as under that of the less refrangible rays, only with this purely quantitative difference, that the more refrangible rays, as in plants, are much more effective than the less refrangible ones, which usually have no effect. I consider it inadvisable to represent the movements ob- served in animals as the expression of a "color preference," or a "color sensation," of a "pleasurable" or "unpleasur- able sensation," as do most animal physiologists and zoolo- gists who have studied the effects of light in the animal kingdom. I do not propose to base an analysis of the movements of animals on such hypothetical, anthropomorphic sensations and feelings, but on such conditions as determine the course of phenomena in inanimate nature as well. Real natural science began when, instead of fabulizing over the HELIOTROPISM OF ANIMALS 17 nature of gravitation, men determined accurately the details of the movement of falling stones, of pendulums, etc., and described them in the most simple and definite terms. In biology, especially in regard to the mechanical effects of light which concern us here, the task of the investigator can only be to determine and describe the circumstances upon which depend the movements of animals under the influ- ence of light. IV. REMARKS ON THE METHOD OF EXPERIMENTATION. THE HELIOTROPISM OF AN ANIMAL USUALLY BECOMES EVIDENT ONLY AT A DEFINITE EPOCH IN ITS EXISTENCE. — THE HELIOTROPISM OF AN ANIMAL CAN EASILY BE OBSCURED BY A SPECIAL FORM OF CONTACT-IRRITABILITY The facts which I have to prove are so simple that almost all technical apparatus can be dispensed with. If one attempts to demonstrate that the orientation of the animals is controlled by the direction of the rays of light, care must be taken that light falls upon the animals from only one side. To accomplish this it is sufficient to carry on the experiments in a room which is lighted from one side only. Since the animals with which we are dealing in this discussion are dorsiventral and place their median planes in the direction of the rays of light, progressive movements are possible in only two directions — either toward the source of light (when they will be called positively heliotropic), or away from the source of light (in which case they will be called negatively heliotropic).1 Diffuse daylight was used as the source of light, and only where specially mentioned was sunlight employed. i Some botanists designate the movements of motile plant organisms toward a source of light as " phototactic," in contrast to the "heliotropic" movements of sessile plants. Since the observations of Sachs, Stahl, and Wortmann, however, leave no room for doubt that the processes are identical in both cases, it seems to me that this separation is not justified. Otherwise a " phototactic " animal ought to become "heliotropic" when its progressive movements are prevented. For this reason I use the same term for similar processes. (See WOKTMANN, Botanische Zei- 18 STUDIES IN GENERAL PHYSIOLOGY There are two methods by which the second fact, that only the more refrangible rays bring about orientation, can be proved, namely, by experimenting with prismatic spectra or with colored screens. All authors who have studied the behavior of plants behind colored screens have obtained the same result — that it is only, or more especially, the more refrangible rays which are heliotropically active. Studies on the behavior of plants in prismatic spectra have led to harmonious results, in so far as they confirm the gross results obtained by using colored screens; yet opinions differ as to the efficacy of the more limited portions of the spectrum. Since for the present I wish to show only that the laws governing the orientation of an animal toward light correspond to the laws governing the orientation of plants toward the same stimulus, it was necessary to use as a basis the really established data of plant physiology, and I therefore shall confine myself to the proof of the fact that the more refrangible rays of the spectrum are exclusively, or almost exclusively, effective. To do this I proceeded as is usual in plant physiology. In order to have only the less refrangible rays act on the animals, I passed the diffuse daylight 'through a solution of potassium bichromate or ruby glass ; to study the influence of the more refrangible rays, I chose cobalt glass or an ammoniacal solu- tion of copper. The screens were examined spectroscopically. The dark-red glass which I used completely absorbed the more refrangible rays, and let through only the red, yellow, and a part of the green rays. The dark-blue glass absorbed the less refrangible red and yellow and a part of the green rays, with the exception of a small region in the outer red. Since, however, the heliotropic phenomena appear only weakly or not at all behind dark-red glass, while they occur just as in diffuse daylight behind dark-blue glass, the few red rays which penetrate the dark- blue glass cannot be HELIOTROPISM OF ANIMALS 19 responsible for the heliotropic phenomena which take place so energetically behind this screen, but can be due only to the activity of the more refrangible rays. The other external conditions which must be considered in heliotropic investigations are so simple that they do not call for any special explanations. Where they are of impor- tance they will be self-evident. It is very essential, however, to realize that the helio- tropism of an animal often manifests itself clearly only dur- ing a definite, often decisive, period of its existence, only to diminish again or to disappear entirely later. It was only by observing for weeks and months the animals described in this treatise, which for the most part I raised myself, that I have been able to establish this fact. The caterpillars of Porthesia chrysorrhoea, for example, are energetically positively heliotropic only during a certain period of their existence, when they have just left the coc- coon in which they have wintered, and have not yet taken food. At this time the entire existence of these animals is a function of the light. Under natural conditions they hatch out on a warm spring day. The light compels them to creep ' to the tips of the branches, where they find their first nourishment in the young buds. When fed they are still positively heliotropic, but very much less so than before. If anyone should examine them in this condition, he would scarcely pronounce them heliotropic. It is not, however, a certain date of the year which gov- erns this heliotropism ; for whenever I forced the animals to leave their nest (by raising the temperature), whether at the beginning of summer or of winter, they were indefatigable in their attempts at creeping toward the source of light. Winged ants are pronouncedly dependent on light only at a definite period of their existence — at the time of their nuptial flight. The same animals which were actively helio- 20 STUDIES IN GENERAL PHYSIOLOGY tropic at the time of the nuptial flight were practically indifferent toward the light a few days previously. In the same way, later on their heliotropism was entirely pushed aside again by another form of irritability, frequently encountered in the animal kingdom, and to which I shall soon return. Fly larvae also possess very different forms of heliotropic irritability at different epochs in their existence. Negative heliotropism is not very distinct in the newly hatched larvae ; but the animals turn their ventral surfaces toward a suffi- ciently intensive source of light without otherwise being influenced by the direction of the rays of light. Full-grown larva?, however, place their median planes very sharply in the direction of the rays of light, provided the light is suffi- ciently intense. I believe that this periodic appearance of heliotropic irritability plays a great role in the ecology of animals. The periodic migrations of many animals, such as birds of passage, might be explained in this way. It is a well-known fact that the irritability of an animal in the larval stage may be entirely opposite in kind to that of the adult stage. This phenomenon is very common. The larva of the fly is negatively heliotropic, while the imago is positively heliotropic; this is also the case with June-bugs and many other animals. I encountered this inversion of the sense of heliotropism when the animal changed from the larval stage to the mature state so frequently that for a time I thought it a universal rule. Such, however, is not the case. Caterpillars, for example, behave toward light as does the imago, as I know from my own experience and from what I can find on the subject in the literature. The behavior of an animal is determined by the sum of all the forms of its irritability. The heliotropic irrita- bility, therefore, may be obscured by a more powerful irrita- bility of another sort. This is often due to a special kind HELIOTKOPISM OF ANIMALS 21 of contact-irritability, which, so far as I know, has not yet been recognized. Many insects are compelled to bring their bodies in contact with the surfaces of solid bodies in a very definite way. My attention was called to this phenomenon in my experiments on animal geotropism, in which I allowed the animals to move about on geometrically simple bodies bounded by plane surfaces. I noticed that the animals rarely remained on the plane surfaces, but collected about the edges, particularly the vertical ones. // is worthy of note that certain animals always seek the concavity of the angle between the sides of hollow cubes, while others just as con- stantly move on the convex side. The caterpillar of Por- thesia chrysorrhcea is an example of the latter type. The other form of this contact-irritability, which leads the ani- mals to the concavity of the angles, is very common. The following observations show how this form of irritability might easily be confused with the irritability toward light, and so lead to a misconception of the behavior of the animal toward light. I studied for several weeks a large number of moths of the species Amphipyra. The animals are remarkable in that they are more given to running than to flying. The rapidity of their running movements calls to mind the lively movements of cockroaches and ants. While formerly I had found that all butterflies are positively heliotropic, I observed that Amphipyrse when let loose, did not fly to the window, but to the nearest wall or to the floor, where they ran about nimbly and crept under the first suitable object, like cock- roaches. This looked as though the animals fled from the source of light. Yet it could be shown that the animals move toward a source of light, and that the inclination to creep into crevices depends upon the contact-irritability, which was mentioned before. The following experiments always succeeded : In the evening, when a lamp was brought into 22 STUDIES IN GENERAL PHYSIOLOGY the neighborhood of a box containing the animals, those which reacted at all always flew with great violence to the side of the vessel which was turned toward the light. In no case did they fly in the opposite direction. The experiment was unequivocal and could be interpreted in but one way. So far as the contact-irritability is concerned, the animals collected in the four concave vertical edges when kept in a cubical wooden box, which was covered on top with window glass. In this position they assumed an indifferent orienta- tion toward the source of light. To make perfectly sure of this fact, I employed the following method : I placed a plate of window glass so close to and parallel with the plane of the floor of the vessel containing the animals that they could just wedge themselves in between the floor and the window glass. The glass plate was entirely exposed to the light. Those animals which by chance came to the edge of the glass plate crept under it, and remained in this position exposed to the light, in contact, however, both above and below with solid bodies. On the next day all the animals were under the glass plate. The animals are therefore forced to bring their bodies in contact with other solid bodies, and it is this (and not the light) which causes them to creep under solid bodies. I placed a ball of paper in the vessel containing the animals ; a part of them crept under the paper and a part into its folds. In nature these butterflies remain in the clefts on the bark of trees or on the ground in mead- ows. Forficula auricularia are found in great numbers in verti- cal crevices (such, e. g., as the spaces between gate and gate- post, in the entrance to gardens). I obtained the animals for my experiments by hanging a cloth of cotton on the top of a small grape vine. The animals collected in the folds of the cloth. These animals in reality move away from the light ; that is to say, they are negatively heliotropic ; but it HELIOTROPISM OF ANIMALS 23 would be wrong to attribute their tendency to creep into the folds of the cloth to their negative heliotropism. When I experimented on these animals with the glass plate, I found that they wedged themselves under it, and remained there exposed to broad daylight, rather than creep away from it. Inside of a box the animals collected in the concave edges ; and it was very noticeable that the animals rarely ran over the free surfaces, but nearly always along the edges, as if it were ever necessary for them to have their sides in contact with solid bodies. I believe that this form of contact-irritability is identical with the important phenomenon, observed by J. Dewitz,1 that spermatozoa are compelled to turn a certain side of their bodies toward solid bodies. Because of this contact-irrita- bility a spermatozoon is never able to leave a cover-glass or a glass slide when once it comes in contact with it. I have observed the same phenomena in hypotrichal Infusoria. These always turn one side of their bodies, the ventral, toward solid bodies. They further resemble the spermatozoa observed by Dewitz in that they alter the direction of their movement always in the same sense, so that on the cover- glass of a microscopical preparation are found only Infu- soria which move in one direction, while on the glass slide they seem to move in the opposite direction. In order to distinguish this form of contact-irritability from other forms of contact-irritability (such as the rolling up or progressive or retrogressive movements when touched), I shall call the peculiarity, possessed by some animals, of orienting their bodies in a definite way toward the surface of other solid bodies, stereotropism. The co-operation of other forms of animal irritability with heliotropism is so simple as to be self-explanatory wherever we may encounter it in our experiments. 1 J. DEWITZ, Pfliigers Archiv, Vol. XXXVIII (1886). 24 STUDIES IN GENERAL PHYSIOLOGY V. THE POSITIVE HELIOTROPISM OF THE CATERPILLARS OF PORTHESIA CHRYSORRHGEA I will enumerate the observations which show the identity of animal and plant heliotropism in the caterpillars of Por- thesia chrysorrhoea. I shall mention only such experiments as in my experience were always successful under the given conditions, and which may be taken as the prototype of the experiments made upon all the animals treated of in this discussion. 1. The direction of the progressive movement in animals is determined by the direction of the rays of light. — I placed a large number — about a hundred specimens of the small gregarious caterpillars of Porthesia chrysorrhoea which had just crept out of the web in which they had passed the win- ter— into a test-tube. They had not fed as yet, and in this hungry condition they were exposed to the light. The tem- perature of the room was necessarily more than 12°-15° C., as otherwise they would have crowded together and fallen asleep again — a state in which they react neither to light nor to gravity. Experiment 1. — If the test-tube is laid on a dark table, so that the longitudinal axis of the tube is perpendicular to the plane of the window, the animals, which are at first scat- tered about irregularly, all assume the same orientation. They creep to the upper portion of the test-tube, turn their heads toward the window, and with their ventral surfaces and their heads turned toward the light creep in a straight line toward the window side of the test-tube. The process requires from one to five minutes, according to the tempera- ture and the condition of the hibernated animals. All with- out exception, provided they are not sickly, move in the direction of the rays of light to the window side of the test- tube. If the tube is turned about an angle of 180°, the HELIOTROPISM OF ANIMALS 25 process is repeated, the animals creeping to the window side of the glass just as before. If, however, the position of the glass remains unchanged, the animals remain permanently crowded together on the window side of the test-tube. Experiment 2. — If the test-tube is laid on the table with the longitudinal axis parallel to the plane of the window, the animals gradually scatter uniformly over the whole of the upper part of the tube. The lower portion of the vessel is in ] consequence again free from animals. If the longitudinal axis of the test- tube lies at even a slight angle with the plane of the window, the animals move to the end of the tube nearest the window, and remain there in their customary position. Experiment 3. — The test-tube is placed perpendicular to the plane F of the window, and at the beginning of the experiment the animals are collected at the window side B of the test-tube (Fig. 1). That half of the vessel which lies nearest the window is now covered with an opaque paste- board box, K. The following then occurs : The animals soon appear at A on the room side of the pasteboard box ; as soon, however, as they emerge from the box K into A, they turn about, direct their heads toward the window, move to the edge of the pasteboard, and remain at the boundary between the covered and the uncovered portions of the tube, at A and especially at the top of the test-tube. The remark- able thing is that they are not distributed evenly over the whole brightly illuminated part of the test-tube. The explanation is as follows : As soon as the animals near the window at B are covered by the pasteboard, the weak rays of light reflected from the walls of the room fall upon them. 26 STUDIES IN GENERAL PHYSIOLOGY The animals follow the path of these rays and arrive at the uncovered portion of the tube. As soon, however, as the strong rays of diffuse light fall upon them at A, they turn about and direct their heads toward the window, until they come again under the pasteboard which shuts out the diffuse light. They are then again attracted by the light of the room, and so on, until they come to rest at the boundary between the two regions at A. At the beginning of the experiment, before the animals stop moving it can really be seen that they are driven around in a narrow circle. If at the beginning of the experiment the animals are collected, not on the window side, but on the room side of the test-tube at C, they move toward the window until they reach the pasteboard at A. If the tube is pulled away from the window for some distance, while the pasteboard remains stationary, the animals begin to move, until they reach the edge of the pasteboard. If the tube is placed horizontally with the longitudinal axis parallel to the window, the animals distribute themselves over the whole length of that portion of the tube which is not covered by the pasteboard, collecting, however, always on the window side of the tube. According to the prevailing views of zoologists and ani- mal physiologists, the movement of caterpillars toward the light is determined by the animals' "fondness for light." They, therefore, move from a region of less intense light to one of greater intensity. That the essential feature, how- ever, is the direction of the rays, and not a difference in their intensity,1 is evident from the following experiments. Experiment 4. — The animals are in a glass cylinder a, some 3cm. in diameter. Light can enter it from all sides (Fig. 2). The inside of a second test-tube 6, which has the 1 In different parts of the tube. [1903] HELIOTKOPISM OF ANIMALS 27 same diameter, is covered with dull black paper, except for a strip about 2mm. wide. The two test-tubes are placed together on a table so that their longitudinal axes lie per- pendicular to the plane F of the window, and the transparent side cd of the glass 6 is turned up; the animals move along the illuminated side cd from a to 6, with- out stopping at the boundary between them, until they reach the window side c of the cylinder. The total amount of light which strikes a caterpillar in the glass b, however, is less than in the glass a, since all lateral rays are cut off in the former and the animal is struck by rays of light only on its ventral side; in test- tube a light falls upon the animals from all sides, though the rays from above and in front are of course the most intense. The animals there- fore move toward the source of light in the direction of the rays of light, even if by so doing — to judge from human sensations — they are led from a "bright" to a "dark" place. In such an experiment no animals are found, as a rule, scattered over the rest of the surface of the glass b. If both glasses are turned around so that a is nearest the window side, the animals of course again move from b to a. The experiments described here were carried on in diffuse daylight. In sunlight, however, the results are the same as in diffuse daylight. When the glass is placed with the longitudinal axis in the direction of the rays, the animals move in the direction of the rays toward the sun and collect at the end of the glass which is turned toward the sun, even though in their hungry state they cannot bear the high tem- perature. When the test-tube is placed with the longitudinal axis perpendicular to the rays, the animals scatter over the STUDIES IN GENERAL PHYSIOLOGY whole length of the tube, remaining, however, upon its sunny side. Orientation takes place more quickly in direct sun- light than in diffuse daylight, I-:..'l>erinient />. — A small pencil SS of direct sunlight is allowed t«> fall on a table obliquely to the plane of the win- dow through the window F (Fig. 3). Rays of diffuse daylight fall upon the remaining portions of the table. If at the beginning of this experiment all the animals are at the end a of the test-tube —which is so placed on the table that a is in direct sunlight, while the other half ft is in diffuse day- light, and is nearer to the plane of the window than a — the fol- lowing occurs: The animals move from a through the pencil of direct sunlight into ft, which lies in the diffuse daylight, where they remain at the cup of the test-tube. They pass from the direct sunlight into dif- fuse daylight without even attempting to return into the sunlight. This experiment can be explained only by the assumption that the orumtalion of the animals is determined by the direction of the n///x. The animal can and must folloir the rays of diffuse light which have the direction ft-* a. If, as is customary with zoologists, we believed that these animals love the light — or, more correctly, that they prefer the more intense light — it would be impossible to see why they do not remain in the direct sunlight, or at least why they do not hesitate to go into the diffuse light. From iclxtt /ms heen sa/W, no one, I believe, will donht tJiat Ihe direction of the progressive movements of the cater- HELIOTEOPISM OF ANIMALS 29 pillars of Porthesia chrysorrhcea is determined by the direction of the rays of light, and not by differences in the intensity of the light in different parts of space. Positively heliotropic animals are compelled to turn their oral pole toward the source of light and to move in the direction of the rays toward this source. 2. The dependence of orientation on the refrangibility of the rays. — I shall now show that it is the more refrangible rays of the visible spectrum which are chiefly concerned in bringing about the orientation of the caterpillars of Por- thesia chrysorrhcea. Experiment 1. — If we place the test-tube on a table and cover it with a box of dark-blue glass, the animals behave as if the vessel were uncovered. Without exception, they move in a straight line to the window side of the vessel and remain there. If instead of blue glass we use red, which to our eyes seems much brighter than blue glass, no change occurs in the orientation of the animals at first; after a long time, however, the animals collect under the red glass on the win- dow side of the vessel. In direct sunlight, however, orienta- tion takes place more quickly. Exactly the same phenomena are observed if an ammoniacal solution of copper is sub- stituted for the blue glass, or a solution of potassium bichromate for the ruby glass. This is also true in the following experiments, where I may not always call special attention to it. This experiment shows (1) that the more refrangible rays have the same effect as mixed rays, and (2) that the less refrangible rays bring about movements in the same way as the more refrangible ones, only their effect is less intense. The experiment also proves that it is wrong to say, as do the anthropomorphists, that the animals "are fond of" blue and "hate" red; for, were this true, the animals should have been forced to move to the room side of the test-tube when under the red glass, yet they moved 30 STUDIES IN GENERAL PHYSIOLOGY toward the window. The animals neither "are fond of" blue nor "hate" red, but they are like plants, simply positively heliotropic, and the blue rays are more effective heliotropically than the red. There is, as I shall state here once for all, no difference in direction between the movements called forth by blue light and red light ; there is only a difference in the velocity and precision with which these heliotropic movements take place. Experiment 2. — The longitudinal axis of the test-tube is again perpendicular to the plane of the window. The small caterpillars are at the beginning of the experiment on the room side of the tube. The window half of the test-tube is covered with dark-blue glass. The experiment goes on as if the tube were uncovered; the animals move to the window side of the test-tube, where they remain under the blue cover. If the same experiment is repeated, only so that the blue cover is placed over the room side of the test-tube, the animals again move to the winjdow, where they remain. The experiment proves that the more refrangible rays alone have the same effect as mixed light ; and the fact that the animals leave the uncovered portions of the test-tube to creep under the dark-blue cover corroborates what has already been said, that positively heliotropic animals move in the direction of the rays of light even when in so doing they pass from a place of greater intensity of light to one of less intensity. Experiment 3. — The test-tube again lies horizontally, with its longitudinal axis perpendicular to the window. At the beginning of the experiment the animals are on the window side of the test-tube. If the window half of the tube is covered with red glass (which may seem much brighter to us than the blue glass of the previous experi- ment), immediately after the red glass has been placed over the animals they appear on the room side of it, and collect at the boundary between the covered and uncovered parts of HELIOTROPISM OF ANIMALS 31 the tube. If at the beginning of the experiment the animals were on the room side of the test-tube, they move until they reach this boundary. We therefore get the same results by using red glass that we got by using opaque pasteboard in a previous experiment. Taken together with the preceding ones, this experiment proves that pre-eminently the more refrangible rays of mixed daylight are heliotropically effective. Although , as we have just seen, the rays passing through red glass or a red solution are not absolutely ineffective, yet the weak light which is reflected from the walls of the room, and which contains some blue rays, is more effective than the diffused light reflected from the sky after it is filtered through red glass. It is for this reason that the animals on the window side under the red cover migrate to the boundary of the red screen where they are held by the rays of diffuse daylight. Experiment 4. — If, as before, we place the test-tube with the longitudinal axis perpendicular to the window, and cover it with red glass on the window side and with blue glass on the room side, the animals collect under the blue glass at its boundary with the red glass. Experiment 5. — If we place the test-tube with its longi- tudinal axis parallel to the window, the animals scatter over the whole length of that part of the tube which is covered by blue glass. From all these experiments it follows that it is chiefly the more refrangible rays which determine the orientation of the caterpillars of Porthesia chrysorrhoea toward light. The only difference between the heliotropism of these animals and the heliotropism of plants is this, that the less refrangible rays are not so completely ineffective in the case of the caterpillars of Porthesia chrysorrhcea as apparently are in many plants. This point must, howevei be studied more accurately with the aid of a spectrum. 32 STUDIES IN GENERAL PHYSIOLOGY 3. The dependence of the orientation on the intensity of the rays of light. — It is a peculiarity of all animal as well as plant structures that only external stimuli of a certain inten- sity can call forth reactions. It can easily be shown that at the approach of twilight there comes a time when the rays of diffuse daylight coming through a window no longer attract caterpillars of Porthesia chrysorrhoea. If the animals are between two sources of light of differ- ent intensities, that having the greater intensity is the more effective. This can easily be shown by bringing the animals into a room into which light enters from opposite directions. Other conditions being the same, the animals move to the window nearest them. A maximum limit for the intensity of the light cannot be established, as direct sunlight is in itself effective. Artificial sources of light above a certain intensity and containing the more refrangible rays affect the animals in the same manner as the natural sources of light. In a dark room caterpillars are attracted by a kerosene flame as markedly as moths; the caterpillars, however, are not burned, because they move so slowly that they have time to turn back before the zone of fatal temperature is reached. Such animals as are attracted by direct sunlight may also be attracted by the candle flame, exactly as is the case in posi- tively heliotropic plants. 4. At a constant intensity light acts as a continuous source of stimulation. — If the test-tube which is placed with its longitudinal axis perpendicular to the window is left undisturbed, the animals remain permanently on the side nearest the window. Under these conditions we can also safely open the room side of the vessel without a single animal changing its position or escaping from its cage. It is remarkable, however, that when the test-tube has been left undisturbed all day, the animals keep their position during the night. In this way I have kept animals for several days HELIOTROPISM OF ANIMALS 33 in a test-tube open on the room side ; but when I turned the vessel through an angle of 180° in the daytime, hardly two minutes elapsed before all the animals had moved to the open end of the vessel which was now turned toward the window. Under these conditions they of course escaped from the test-tube. A position which the animals have assumed under the influence of light is usually not changed when the light is removed, unless some other stimulus comes into play. 5. On negative geotropism and contact-irritability in the caterpillars of Porthesia chrysorrhcea. — The reader may perhaps have noticed that in all of these experiments on caterpillars the test-tubes were always placed with their longitudinal axes horizontal. This was due to the fact that the animals behave like plant structures, not only in regard to their heliotropic, but also in regard to their geotropic, irritability. Just as is frequently the case in positively heliotropic plants, we find that the caterpillars are also nega- tively geotropic ; that is, they are compelled by gravity to creep vertically upward until they come to rest in the highest part of the test-tube. These experiments were made in a dark room, with the long axis of the test-tube in a verti- cal direction. If the test-tube is inverted, the animals again creep to the top ; if left undisturbed, the animals remain in the uppermost regions of the test-tube. It is necessary in these experiments, as in those on heliotropism, to have the temperature of the room at least 15°, preferably as high as 20-22 °. It is simplest to put the test-tube in one's pocket with its longitudinal axis vertical. In a few minutes the animals are found at the highest point in the tube. An increase in temperature increases the geotropic irritability of the animals. « It must now seem questionable whether in our former discussion of the heliotropism of these animals we were 34 STUDIES IN GENERAL PHYSIOLOGY justified in taking as the effect of light the movements of the animals to the top of the test-tube ; it might, indeed, be a geotropic phenomenon. To decide this the animals were placed in a test-tube which was lined with thick black paper except for a strip 2 mm. wide. The uncovered strip was turned downward, so that light could enter the vessel only from below. Diffuse daylight was reflected through the slit from below by means of a mirror. The animals collected in the lower, lighted portion of the glass vessel. Their helio- tropism is therefore more powerful than their geotropism, even when only weak diffuse daylight is used. The geotropic experiments succeed only when the animals have been in the light for some time and have not yet come to rest. When the animals are kept in the dark for a long time and the test- tube is not disturbed, they do not creep upward. The orienting effect of the light always exceeds that of gravity. The effects of gravity, like the effects of light, usually appear only during certain periods in the life of the animals; at any rate, they cannot always be demon- strated with certainty. The contact-irritability of the caterpillars of Porthesia chrysorrhcea shows itself by the way in which the animals remain in the corners and convex sides of solid bodies. I covered the boxes in which I cultivated my caterpillars with large, square glass plates. These did not close the box tightly, so that the animals could creep out and creep upon the glass plates. Only rarely, however, were they found on the free surface of the plates. The animals moved along the rough edges of the plate until they reached the window side of the dish. I confirmed this observation almost daily for months. When I placed the animals upon the outside of a •cubical block, they collected by hundreds in one of the upper corners. Of course, only a few have room in the corner itself, but, as is generally the case with these ani- HELIOTROPISM OF ANIMALS 35 mals, when a few have collected in a spot the others on arriving hold fast to the sides of those already there. An animal at rest acts upon a creeping one as a convex edge. On the other hand, I have never observed that the animals within the cubical box collect on concave edges. From this it follows that the friction of gliding over the convex corners is the source of the stimulation which compels the animal to come to rest there ; in moving over the concave corners this friction, of course, does not take place. These three forms of irritability control mainly the daily life of the animals. We find them in great numbers in fruit trees and bushes, where they pass the winter in their nests; as soon as the warm weather comes, they leave their nests. Positive heliotropism and negative geotropism com- pel them to creep upward to the tips of branches, and contact- irritability holds them fast on the small buds. We can easily show that neither smell nor a special mystical "instinct" leads the animals to the buds, as we are able to compel them by the aid of light to starve in close proximity to food. The animals move to the window side or to the top of a test-tube in which they are kept. If then a branch covered with buds is pushed into the test-tube on the room side, the animals nevertheless remain where light and gravi- tation have compelled them to go and are holding them. If, however, they once are on the buds, the latter act as a stimulus which may be even stronger than the light. It is in such a case impossible to draw the animals away from the food by means of light. All these forms of irritability can best be demonstrated on animals which have just left the nest in which they have spent the winter, and which have not yet eaten anything. As soon as they have eaten and are about to moult, their irritability decreases, and at the time of moulting it is almost impossible to show any effect of light or gravity upon them. 36 STUDIES IN GENERAL PHYSIOLOGY 6. The effect of temperature on the caterpillars of Por- thesia chrysorrhoca. — The caterpillars of Porthesia chry- sorrhoea behave toward a source of heat in a manner opposite to that in which they behave toward light ; they move away from the source of heat. If the animals contained in an opaque vessel are brought in the neighborhood of a hot stove, they leave the side of the vessel which is nearest the stove. Yet the heat does not compel the animals to move in a straight line, as they do when struck by the more refrangible rays of light. This directing effect of the more refrangible rays of the visible spectrum is greater than that of the dark heat rays. In this way it is possible for the same animal which flees from the source of the dark rays of heat nevertheless to move in the direction of the sun's rays to the sunny side of a vessel. It is a well-known fact that irritability in a tissue is a function of the temperature. I have already mentioned that at a temperature of less than 13° C. the animals are no longer affected by light. It can be shown that heliotropic irrita- bility increases with an increase in temperature. If the animals are kept during the day in a room having a tem- perature of about 18°, it is found that they no longer respond to light when beyond a certain distance from the window. If, however, the temperature of the test-tube is increased a few degrees, the animals move the more quickly to the win- dow side of the tube the higher the temperature. It can easily be demonstrated that the orientation takes place more rapidly, and that the direction of the progressive movements coincides more nearly with the direction of the rays of light, whenever the temperature is raised. If, however, the tem- perature is increased to 30° or over, the animals become very restless; they raise the anterior ends of their bodies higher than is usual in their movement, and so decrease the velocity of their progressive movements. The most suitable tern- HELIOTROPISM OF ANIMALS 37 perature for demonstrating their heliotropic activity lies between 20° and 30°. The experiments on the caterpillars of Porthesia chry- sorrhcea are typical. I have repeated them on some hundred species of insects, but I have never found a positively heliotropic insect whose dependence upon light was of a different kind from that found in Chrysorrhoea. This fact has given me the impression that all animal proto- plasm, as perhaps all plant protoplasm, is heliotropically irritable, and that where this is apparently not the case the heliotropic reaction is inhibited, either temporarily or permanently, by other causes. For this reason it would be useless to publish here every single experiment I have made. This would result in repeating each time the same phenomena, only under the name of a different insect. Since there are only negatively and positively heliotropic animals, it would be of secondary interest to know to which of the two classes the individual animals belong. But I believe it necessary to show by concrete examples what part heliotropism plays in the habits and ecology of animals. VI. THE POSITIVE HELIOTROPISM AND THE SLEEP OF BUTTERFLIES Our knowledge of the behavior of butterflies toward light has, on the whole, remained at that point which is marked by the statement of Keaumur that "it- is a singular fact that those butterflies which shun the daylight are pre- cisely those which fly into lighted chambers." The paradox has not yet been explained why those butterflies which are not to be seen by day fly into the flame at night, while the day butterflies apparently do not possess the tragic "instinct" of the night Lepidoptera. There is no lack of conjecture on this point. Eomanes believes that the lamp is a "strange object" to the moths, and that "the desire to examine this 38 STUDIES IN GENERAL PHYSIOLOGY strange object" drives the moths into the flame. We find, however, that the caterpillars of Porthesia chrysorrhoea creep as well toward the sun as toward a lamp. Yet, according to Romanes, the sun ought to be a familiar ob- ject to these animals. Such anthropomorphic opinions as those of Romanes are evidently as useless in the analysis of life-phenomena as the speculations of metaphysicians— e. #., Hegel's — on physical phenomena. A scientific analysis of the behavior of moths toward light leads to a very simple explanation of the paradox. Experiment 1. — Specimens of Sphinx euphorbiae, Bom- byx lanestris, and other moths are kept in a large glass box. The box is placed in a room into which only daylight and no artificial light enters. As soon as the animals begin to fly, at the approach of twilight or later, they collect at the window side of their boxes. Whenever the box is reversed the animals fly back to the window side. This experiment is rendered more complete by the following observations: I kept the pupae of moths in an open box. Most of the moths hatched at night. On the following morning I always found them collected at the closed window of the room. Here they remained all day exposed to the light. Finally, when I caused the moths to fly by day, I noticed that they flew to the window as do all other positively heliotropic insects. These experiments show that the animals are attracted, not only by a lamp, but also by diffuse daylight. They also show that Reaumur's idea that moths shun daylight is wrong. The experiments indicate that the animals are positively heliotropic toward diffuse daylight, although, as we shall soon see, this positive helio- tropism may during the daytime be obscured by another form of irritability. Experiment 2. — I brought some specimens of Sphinx euphorbias into a room which had a window only on one HELIOTEOPISM or ANIMALS 39 side. On the wall of the room opposite the window I placed a kerosene lamp. At the approach of twilight, when the animals began to fly about, I brought them into the middle of the room, so that they were equidistant from the lamp and the window, and left them alone. They flew to the window. Yet, when I brought them into the immediate neighborhood (within about a meter) of the lamp, they flew into the flame. I repeated this experiment and convinced myself that they always flew to one of the two sources of light, either the window or the lamp ; to the latter, however, only when they were in its immediate neighborhood. This experiment shows that the animals do not even pre- fer artificial to the natural light, but that the artificial light attracts them only when its intensity is greater than that of the diffuse daylight, which is the case at night when the animals are within a certain distance of the lamp, varying with the intensity of the flame. The heliotropic sphere of attraction of an electric arc light is therefore larger than that of a candle flame, and the number of moths attracted by it correspondingly greater. Experiment 3. — It must yet be proved that it is chiefly only the more refrangible rays of light which determine the movements of the moths. I studied the behavior of Sphinx euphorbia, which began to fly at about 9 o'clock in the evening. The animals were contained in a large box, 40 cm. long, the upper wall of which was of glass. Whenever I turned the box the animals at once flew to the window side and crowded against the upper glass wall through which the light came. When I placed a red glass over the window side of the box, the animals at once flew to the room side. They collected at the edge of the red glass, but on the room side of it, where they were not covered by it. Here they attempted to fly upward. When I used blue glass instead of 40 STUDIES IN GENERAL PHYSIOLOGY red, they flew under it to the window side of the box. At fifteen minutes past 9 o'clock they came to rest and no longer reacted to light. When exposed to daylight on the follow- ing day, they did not stir, and made no attempt to creep away from the light, although sufficient opportunity was offered. I repeatedly established the fact that the movements of night butterflies are determined by the more refrangible rays of the spectrum on other specimens of Sphinx euphorbise. It was therefore not to be expected that in lamplight any other than the more refrangible rays would bring about movements. I have convinced myself that the moths of Geometra piniaria are readily attracted by the light of a lamp when behind blue glass, but not when behind red glass. The night butterflies, therefore, shun neither diffuse nor intense light, nor do they prefer artificial light to diffuse daylight ; the correct expression of the facts is rather this, that most species react to light only at night, when they are positively heliotropic like the day Lepidoptera. We find in butterflies periodic variation in irritability (as in many plants), and these variations correspond to the changes of day and night. As certain flowers open their calices only by night, while others open theirs by day, so certain butterflies fly only by day, while others fly only by night. Both classes of butterflies, however, are positively heliotropic; and it seems as if the irritability of the night butterflies toward light is not less, but even greater, than that of the day but- terflies ; for the intensity of the light which causes heliotropic phenomena in moths is apparently much less than the mini- mal intensity which stimulates day butterflies to heliotropic movements. The phenomena of sleep in butterflies are perhaps more complex than the corresponding phenomena in plants. One thing is, however, certain— that the periodicity of the noc- turnal movements of butterflies does not change during the HELIOTROPISM OF ANIMALS 41 first two or three days if the animals are kept in the dark. Under these circumstances the moths become restless at the usual time. Reaumur showed that moths begin to fly in the evening when kept in a box. I must leave it undecided for the present whether this periodicity finally disappears if the animals are kept still longer in the dark. I have tried repeatedly to cause Sphinx euphorbia to fly in the daytime by a sudden diminution in the intensity of the light. When I protected the animals from all jarring I never succeeded between 6 and 12 o 'clock in the morning. Yet I was easily successful in the afternoon, long before the beginning of twilight. I will cite here several of my experiments. One morning I placed a Sphinx euphorbias, which had begun to fly at 9 o'clock on the previous evening, on the window cur- tain, where it remained quietly. At 2:45 I returned it to its glass box, which stood in a dark corner and into which light fell only through a narrow slit. An hour went by, but the animal did not leave its place. It then moved to the light side of the box, without flying. I carried the animal back to the window, where it remained quietly. After twenty min- utes I returned it again to the dark box. Half an hour later, at half -past 4, it finally began to fly. The next day I allowed it to remain at rest near the win- dow, and it did not begin to fly until 9 P. M. at well-advanced twilight. On the following day I kept it in the dark box, and at half -past 3 in the afternoon it had already begun to fly. At noon on the succeeding day a heavy storm came up and it grew quite dark. The moth, which until then had remained quietly at the window, began to fly. I have had the same experience with other examples of this species. These facts seem to indicate that it is possible to influence the time of waking of Sphinx euphorbise by diminishing the intensity of the light, but only when they would soon wake up without artificial interference. 42 STUDIES IN GENERAL PHYSIOLOGY The day butterflies are positively heliotropic like the night butterflies. The only striking feature is that in certain day butterflies the intensity of the light must be very great to bring about heliotropic movements. Specimens of Papilio machaon (which I had raised) remained at rest during the day at a window where they were exposed to the diffuse day- light and could be carried around on the finger; as soon, however, as they were brought into direct sunlight, they flew toward the window in the direction of the rays of light, and this with such force that they dropped down as if stunned. In direct sunlight they pressed themselves closely against the window pane. In diffuse daylight the animals, if they moved at all, crept toward the source of light ; but in direct sunlight they flew toward it. My attempts to attract Papilio machaon by the weak light of a kerosene lamp were unsuc- cessful. I will add at this point my general observations on the caterpillars of butterflies. I have not found these periodic variations in heliotropic irritability in most caterpillars, not even those of Sphinx euphorbiae. The caterpillars which I studied reacted to light at all times of the day and night. The caterpillars agree, however, with the day and night butterflies in so far as they are all, without exception, positively heliotropic. This positive heliotropism is most marked in the cater- pillar of the willow-borer, which lives in the stems of the willow where it is not at all exposed to light. Such cases are also known in plants. Koots, for instance, are helio- tropically irritable, and yet, as Sachs points out, under nor- mal conditions their heliotropism is of no use to them. They can certainly not have acquired it through natural selection. According to the Darwinian theory, we would expect that the caterpillars of willow-borers should be nega- tively heliotropic, or at least indifferent to light. But the HELIOTROPISM OF ANIMALS 43 behavior of an animal is merely the resultant of all its forms of irritability, and so it may happen that an animal is positively heliotropic even when it has no opportunity to make use of it. The larvse of many saw-flies behave just as the caterpillars of Lepidoptera. I have made observations on the larvse of Nematus ventricosus, which are exactly like those on Porthesia chrysorrhoea, which have been described. I have not yet succeeded in demonstrating a heliotropic reaction to diffuse light in the indigenous pupae. Wilhelm Mtiller, however, has observed effects of light in South American species.1 The pupa3 can move at three joints. Only a lateral movement to the right and left is possible in some of the species; in other species only a dorsal move- ment of the body is possible; in a third species of pupae a combination of both kinds of movements is possible. Miiller observed that all three classes of movements can be brought about under the influence of light. He found that some pupse turned not only away from the light, but also toward it. He also found that when the animals had been exposed to the dark for some time, they "needed some time to become susceptible again to the influence of light." In interpreting the phenomena Mtiller follows the Darwinian idea, so that the thought never occurs to him that he might be dealing with phenomena similar to the heliotropic phenomena of plants. The negative geotropism of the Lepidoptera. — The movements of very young or recently hatched animals have for the most part been misunderstood, because they have always been considered a function of mysterious "instincts" of the animals, while the direction of their motions is in reality determined by definite external forces. The same cause which prescribes the course of a falling stone or deter- mines the orbits of planets, namely gravitation, determines i MILLER, Zoologische Jahrbiicher, Vol. I (1886), pp. 568 ff. 44 STUDIES IN GENERAL PHYSIOLOGY also the path which a butterfly follows that has just emerged from the pupa case. The geotropic irritability is at that time especially strong ; the newly hatched animals remain restless, and are compelled to run about until they come to a vertical wall, on which they can put the longitudinal axes of their bodies vertically, with their heads upward. Here they remain quietly until their wings are unfolded. The powerful mani- festation of negative geotropism at the time of hatching is no isolated phenomenon in insects. In summer, we find great numbers of the ecdyses of the Iarva3 of Epheuieridae on the banks of streams. They are found on blades of grass or steep banks, with their longitudinal axes usually vertical and the head upward. That gravity, and not light alone, plays the chief role here is shown by the fact that I have found the ecdyses in the same position under bridges where no light could strike them from above. This observation on the larv« of Ephemeridse makes it impossible for us to accept the idea that the "purpose" of the orientation of the freshly hatched imago of a butterfly is that the wings may unfold; for negative geotropism appears in the larvae of EphemeridaB at a time when no wings are present. The caterpillars of butterflies are also negatively geotropic like the freshly hatched moths, even though not so markedly. Immediately after hatching geo- tropism is much stronger in the imago of the butterfly than heliotropism — a phenomenon rarely observed in the animal kingdom. If a freshly hatched imago is on a vertical wall, it does not change its orientation toward the center of gravity even when the direction, ref rangibility, or intensity of the light is changed. What is true of the heliotropism of Lepidoptera, that it is most marked during certain periods of their existence, holds good also for their geotropism. Amphipyra is ener- getically negatively geotropic immediately after moulting. HELIOTROPISM OF ANIMALS 45 Several days later the animals assume every possible position with reference to the vertical. They prefer to remain on vertical walls, yet they will creep just as readily into hori- zontal folds and crevices. VII. THE POSITIVE HELIOTROPISM OF PLANT LICE Anyone closely studying a rose covered with wingless plant lice will notice that they are arranged in a definite way on the plant. On a vertical stem they rest with the head downward ; on the leaves they are usually found on the underside, mostly on the principal veins. Here one also notices a certain regularity in their orientation, in so far as the animals on the principal vein turn their oral poles toward the stem, and their aboral poles toward the point of the leaf. The orientation of the animals seems therefore to be controlled by the structure of the plant, and not directly by external forces. But the plant lice do not behave on all plants as on the rose. On a palm, for example, I found no such definite orientation of the animals toward the plant, even though in this case also they show a preference for the lower surfaces of the leaves. Yet it might seem reasonable to suppose that light or gravity compels the plant lice to seek the lower surfaces of the leaves. I twisted several leaves of Cineraria, the dorsal sides of which were covered with plant lice, so that the dorsal sides were directed upward and toward the window, and fixed the leaves in this position. I watched the animals for two days and found by actual count that the animals remained at rest. I repeated the same experiment on the plant lice of palm leaves, but also with negative results. My experiments on the orientation of new-born wingless plant lice were practically negative when I removed them from the plant and placed them in a glass vessel. Yet in 46 STUDIES IN GENERAL PHYSIOLOGY the older wingless animals I could notice an inclination to move toward the source of light. When their wings had sprouted, however, the orientation of the plant lice was extraordinarily definite. In this state they are perhaps the most suitable animals we have for demonstrating the phenomena of heliotropism. Not all species are equally irritable ; Cineraria afforded me the best specimens. I have never found a species of plant louse which was not definitely positively heliotropic. I kept the plants near a closed window. The animals were attracted by the sun to the window, where they crept upward. When the animals are lightly touched with the point of a pen, they fall down a second or two later. If a glass vessel is held under them, a large number of these animals can be collected in an unin- jured condition in a short time. I found it much better to work with such animals as have already flown from the plant, than to collect the winged animals from the plant itself. To obtain the winged plant lice in great numbers it is necessary only to allow a plant which is covered with them to dry out gradually. Under such conditions the wings grow out very rapidly. All the experiments which were made with Porth<-xiv Planaria at the beginning of the exper- \ ) iment are brought to the window side a of the vessel, but so that they are not FIG. 66 struck by the light, all, or almost all, the animals are found in a few hours, or on the following day, under the opaque paper be where the intensity of the illumi- nation is least. If the same experiment is made with nega- tively heliotropic Limulus larvse, the larvae move to the room side a of the vessel, and remain there permanently. It is clear, under these circumstances, that when these Planarise are left quietly for some days in a cylindrical vessel abed (Fig. 67), all the animals finally collect at the two sides c and d, as was observed by Dr. Wheeler. Heliotropic animals in the same vessel either go immedi- ately to the window side a or the room * + side b of the vessel, and remain there. This mode of reaction to changes in the intensity of light occurs probably also in angleworms; perhaps, too, in other animals. It is, moreover, pos- sible that heliotropism and photokinesis are associated in certain animals1 — a subject which I still wish to investigate.2 IE, gr., Spirographis spallanzanii. 2 1 had previously noticed that in some animals, which I at that time considered negatively heliotropic, the typical heliotropic experiment did not succeed very well. 1 attributed this to secondary circumstances. I now consider it possible, however, that the experiments which I described, for example those on the larvse of beetles, indicate as much the existence of photokinesis as negative heliotropism. I shall make further experiments in this direction. TRANSFORMATION OF HELIOTROPIO ANIMALS 289 2. There are photokinetic animals which react more rapidly to changes in the intensity of the light than do Planarians. I noticed this form of reaction at Naples in certain Annelids living in tubes; for example, Serpula uncinata. The gills of the animals are often exposed to the light. When the hand is moved between the animals and the source of light, they quickly draw back into their tubes as soon as they are struck by the shadow. In order to see whether positive and negative changes in the intensity of the light had the same effect, I made the following experi- ment: A glass aquarium which was closed by a glass cover was set upon an isolated table about 2 m. distant from the window. When I closed the shutters rapidly, the worms quickly withdrew into their tubes, much as does a snail when touched suddenly. The shutters did not close absolutely, and it was always light enough in the room to observe the animals. After some time the animals would again stretch out their gills. When I now suddenly opened the shutters quickly, the animals did not react. Even when the animals had withdrawn into their tubes, an increase in the intensity of the light did not again bring them out. It is therefore only the decrease in the intensity of the light which acts as a stimulus upon the animals. One notices, however, that these reactions cannot always be relied upon. Andrews has noticed such reactions also in Annelids whose gills are free from eyes or eye-like organs.1 VI. ON SOME PHYSIOLOGICAL CONDITIONS WHICH DETER- MINE THE DEPTH-DISTRIBUTION AND DEPTH-MIGRATION OF MARINE ANIMALS 1. Investigations concerning the depth-distribution of marine animals seem to show that we meet with a consider- able amount of animal life only in two regions of the sea — IE. A. ANDREWS, Journal of Morphology, Vol. V (1891). 290 STUDIES IN GENERAL PHYSIOLOGY at the surface down to a depth of perhaps 400 m. and at the bottom of the sea. Some of the surface, or pelagic, animals show a periodical depth-migration. They come to the sur- face at night and move down during the daytime. In the Mediterranean, Chun found another migration of a greater period. Animals which always come to the surface in win- ter, or at least during certain hours of the day, live at greater depths in summer. The first experimental investigations on the cause of these depth- migrations were made by Groom and myself, and led to the conclusion that in the Nauplii of Balanus per- foratus heliotropism alone suffices to account for the fact that they rise to the surface at night and move down during the day.1 These animals are positively heliotropic to weak light, but strong light soon makes them negatively helio- tropic. They are, in consequence, driven into the depths during the day from the surface of the water. They can, however, not go very deep, as the intensity of the light decreases with an increase in the depth of the illuminated layer of water, and becomes so weak at a certain depth that the Nauplii again become positively heliotropic. They must, in consequence, again move toward the surface. As soon, however, as they again come into more intense light, they become negative again. It can therefore be easily seen why these animals do not go to the bottom of the ocean during the day, but are forced to remain in a layer of water which is not too far below the surface. When the light becomes weaker toward evening and in the night, the Nauplii are again forced to move to the surface of the water in conse- quence of their positive heliotropism. 2. The question may now be asked whether all animals which are found at the surface of the sea are constantly, or at least under certain conditions, positively heliotropic. I 1GKOOM UND LOEB, Biologisches Centralblatt, Vol. X (1890) TKANSFOBMATION or HELIOTROPIC ANIMALS 291 have made experiments on the small animals obtained in the surface dredgings at Woods Hole, on Copepods, and on the larvaa of crustaceans, worms, and molluscs, and have thus far been able to find no pelagic animal of these classes which is not either permanently, or at least at times, positively heliotropic.1 3. It would be incorrect, however, to assume that heli- otropism is the only condition for the depth-distribution of animals. Just as in the vegetable kingdom positive heli- otropism and negative geotropism often act together toward the same end, we must expect similar conditions in the ani- mal kingdom. In a paper on geotropism I have already shown that certain starfish and Actinians, which always live near the surface of the water, are compelled to creep con- stantly upward, owing to a peculiar form of irritability, and I have made it probable that this irritability is negative geotropism.2 I have since been able to convince myself that in certain animals which would be forced by their positive heliotropism alone to go to the surface, other conditions are at work which co-operate with heliotropism. This is the case, for example, in the freshly hatched larvae of Loligo. These animals are constantly positively heliotropic, and, besides, live at the surface of the sea. When these animals are introduced into a long eudiometer tube filled with sea- water which is set up vertically, they move to the surface of the sea-water. As long as the effective rays of light fall into the tube from above, the positive heliotropism of the larvae would compel them to move upward. I found, how- ever, that the animals come to the surface of the water also in the dark room. Moreover, when the eudiometer tube is exposed to the light, with the upper part of the tube covered 1 1 have even found a young pelagic fish at Woods Hole which is as pronouncedly positively heliotropic as the insect larvse described in a preceding paper. I was not able to ascertain the species to which it belonged. [1903] 2 Part I, p. 176. 292 STUDIES IN GENERAL PHYSIOLOGY with a cap impermeable to light, the animals which are very energetically positively heliotropic should not go under the dark cap. The latter does, however, actually occur. Hence another and more powerful circumstance is at work besides the positive heliotropism, and this might be negative geot- . ropism. The temperature in the tube was everywhere the same in these experiments. The need for oxygen does not compel the animals to come to the top in these experi- ments, for when the eudiometer tube is filled entirely with water, and the FIG. 68 open end is turned downward into a larger vessel filled with water, the animals nevertheless move from this vessel to the cap of the eudiometer, and remain here even though fresh oxygen can reach the animals at this point only through diffusion from below. The significance of this negative geotropism (which I believe it to be) for the upward movement of the Loligo larvae is shown still more beautifully in the following experiment : AB (Fig. 68) represents a vertical section through the plane of the window; CD is a eudiometer tube filled with sea-water and closed with a cork. If the larvse at the beginning of the experiment are collected at the room side C of the tube, they all move to the window side D in consequence of their positive heliot- ropism, and remain there. If, on the other hand, the tube is so placed that it is at an angle with the horizontal — for example, in the position Cf> — the Iarva3 gradually but steadily rise by active swimming movements away from the window to the elevated end Cl and remain there. They therefore leave the window side and go to the room side, in spite of their positive heliotropism. To accomplish this the angle Cf>C must in this case be not smaller than about 20 °.1 1 The papers of Wolfgang Ostwald seem, however, to indicate the possibility of a purely physical explanation of this experiment. [1903] TRANSFORMATION OF HELIOTROPIO ANIMALS 293 Of course, negative geotropism is not so closely linked with positive heliotropism in all pelagic animals. If. for example, the Copepods mentioned in this paper are intro- duced into the eudiometer tube, and the upper end is covered with the opaque capsule, the animals also rise in the tube when they are positively heliotropic. They do not, however, go under the opaque cap, as do the Loligo larvae, but remain at the highest uncovered portion of the tube. They therefore move upward entirely, or at least mainly, through their positive heliotropism. 4. The effect of temperature upon the depth-migration and depth -distribution of marine animals must yet be studied experimentally. We saw that the larvae of Poly- gordius are negatively heliotropic and positively geotropic at high temperatures, but become positively heliotropic and negatively geotropic at lower temperatures. The same trans- formation of heliotropism occurs in Copepods. It can scarcely be assumed that other animals will not show the same phenomenon. In water whose surface reaches a very high temperature in summer such animals must disappear during that period from the surface, as the high tempera- ture makes them negatively heliotropic, and perhaps also positively geotropic. The negative heliotropism and posi- tive geotropism drive these animals into the depths. As soon, however, as they reach cooler layers of water below the surface, they again become positively heliotropic and negatively geotropic; they must then rise again until they have reached the warmer layers of water. Here they soon again become negatively heliotropic, and perhaps positively geotropic, when they must again sink ; and so on. In this way such animals are kept floating at a certain distance under the surface of the water during the summer. When, however, the temperature becomes sufficiently low in the winter, the animals may rise to the surface without becoming negatively heliotropic or positively geotropic. 294 STUDIES IN GENEEAL PHYSIOLOGY Another condition must, however, be taken into con- sideration, which may have the effect that animals which are constantly positively heliotropic must leave the surface of the water at higher temperatures. As is well known, the processes of oxidation, and consequently the demand for oxygen, rise considerably with an increase in temperature. It is natural that when the demand for oxygen exceeds the supply, the animal can execute no, or only weak, swimming motions, and in consequence falls to the bottom. At any rate, the latter can, indeed, be observed at high tempera- tures. Loligo larvae which hold themselves at the surface by swimming motions sink passively as soon as the tempera- ture exceeds 30° C. In conclusion I wish to add that I made experiments on most of the animals mentioned in this paper with colored light, and found a universal confirmation of the fact, which I discovered before, that the more strongly refrangible rays of the visible spectrum are the most active heliotropically, as in the case of plants, IX ON THE DEVELOPMENT OF FISH EMBRYOS WITH SUPPRESSED CIRCULATION1 1. ONE of the methods which may lead us deeper into the physiology of development consists in removing one link in the chain of such processes in order to see how the further development is influenced by such a step. I recently, came upon such an experiment, which I wish to detail in the fol- lowing pages. The experiment consisted in preventing, by a specific cardiac poison, the beat of the heart and the circu- lation of the blood in an embryo. I had, indeed, expected that under such circumstances the embryo would not die immediately, but I did expect that its further development would certainly be impossible. In this, however, I was mis- taken. Development went on in spite of the elimination of the activity of the heart ; in some cases as long as four days —which was nearly half, or one-third, of the duration of the embryonic stage. The consequences of the elimination of the activity of the heart are also in some ways different from what one might expect. The observations have not been completed on all points because of the lateness of the season and lack of material, but I intend to fill in these gaps next season. The experiments were made upon a marine fish, Fundu- lus, which is very common at Woods Hole. The eggs were fertilized artificially in normal sea-water. In order to prevent the action of the heart and the circulation in the developing embryo, the eggs were put, half an hour after fertilization, into sea-water to which a sufficient amount of potassium chloride had been added. Potassium chloride can 1 Pflugers Archiv, Vol. LIV (1893), p. 525. 295 296 STUDIES IN GENERAL PHYSIOLOGY be called a specific cardiac poison only in so far as the heart can be brought to a standstill with a much smaller dose than is required to poison the remaining organs of the embryo. 2. At a temperature of 20° C., and with a plentiful sup- ply of oxygen, the embryos of Fundulus develop in about twelve to fourteen days in normal sea-water. Under these conditions the heart begins to beat about sixty to seventy hours after fertilization. If Fundulus embryos four to six days old are placed in sea-water to which 1.5 g. of KC1 have been added to each 100 c.c. of sea-water, the heart ceases to beat, and death ensues in about one hour, at the outside, in all the embryos. That the poisonous effect of the potas- sium is not limited to the heart is shown by the fact that the embryo becomes exceedingly restless before the heart ceases to beat. An entirely different series of phenomena ensues when the Fundulus eggs are, introduced into the same salt solution about one-half hour after fertilization. The eggs develop in an entirely normal way, the embryos live until the fifth or sixth day, and the cramp-like movements do not set in. Furthermore, I noticed in a few of these embryos very weak and very slow pulsations of the sinus venosus on the third or fourth day. This activity of the heart did not, however, appear in all the embryos, and when it did appear it did not last long. In no case, however, was the beat of the heart sufficient to cause a circulation of the blood. The circulation in the yolk-sac of the Fundulus embryo can be demonstrated more clearly and easily than the circulation in any of the ordinary preparations used for this purpose. In the normal embryo the circulation is very marked, even some seventy -five hours after fertilization; but it does not matter how long one waits in the case of the embryos poisoned by adding 1.5 g. of KC1 to each 100 c.c. of sea-water — never did I succeed in discovering even the slightest indication of DEVELOPMENT OF FISH EMBRYOS 297 a circulation of the blood in the vessels. In spite of this fact, a complete circulatory system was developed which did not differ markedly, in regard to the direction and the branching of the vessels in the embryo and in the yolk-sac, from that of a Fundulus embryo developed in normal sea- water. Heaps of red-blood corpuscles -were found in the large blood-vessels, such as the arteries of the yolk at the point where they leave the embryo. The important result of these observations is therefore the fact that a complete vascular system, which is probably identical in its main distribution and in its two vascular divisions with that found in the normal embryo, can be formed without a circulation, and therefore without blood- pressure. A difference between the two, which also deter- mines the limit of the possible identity of the vascular sys- tems in the normal and abnormal embryos, is found in this, that the lumina of the vessels in the poisoned embryo are exceedingly irregular. The lumen of a vessel is in extreme cases rosary-like, narrow and wide spots alternating with each other. This is due to the absence of a sufficient intra- vascular pressure. 3. It might have been possible that a circulation lasting only a short time had been present which I did not dis- cover. I therefore made experiments with much stronger KC1 solutions — such in which not even a trace of cardiac activity ever appeared. In extreme cases I addded 5 g. of KC1 to 100 c.c. of sea-water. I had previously ascertained that a four-day-old embryo which had developed in normal sea-water dies in two minutes in a 3 per cent. KC1 solution. Nevertheless, the fertilized eggs developed normally in a 5 per cent. KCl solution. They developed for three to six days, and formed — what is of interest to us here — a heart and a typical vascular system in the embryo and yolk-sac. On the other hand, the development of the embryo as a 298 STUDIES IN GENERAL PHYSIOLOGY whole was markedly retarded in this concentrated salt solu- tion, so that a definite judgment of the vascular system could be made only so far as the main stems and their branches were concerned. These main stems corresponded with the main stems of the normal embryo. Yet I was never able to discover even the slightest evidence of a heart-beat, much less a circulation. The lack of hydrostatic pressure within the vessels was particularly evident here from the irregularity in the diameter of the blood-vessels; nevertheless, a large number of branches, which gradually decreased in caliber, sprang from the main vessels. In this case, therefore, it is unquestionably true that the process of branching and the growth of the blood-vessels are independent of blood-pres- sure. 4. The experiments with weak KC1 solutions also deserve mention. In a series of experiments I added 0.25 to 0.5 g. to 100 c.c. of sea-water; normal development occurred in these solutions. The heart-beat and the circulation developed apparently normally. The control eggs, which had been taken from the same culture, but raised in normal sea- water, completed their development in twelve to sixteen days, when the embryos hatched. They lived some four to six weeks after escaping from the egg. In the two KC1 solutions, however, but one embryo, which lived for a day, hatched on the twelfth day in the 0.5 per cent. KC1 solution. All the remaining embryos died between the twelfth and sixteenth day. Death unquestionably resulted from a poisoning of the heart, and not from a general intoxication. 5. The experiments cited above show thai a KC1 solution of a definite constitution is the more poisonous the older the embryo. . One might think that the chemical constitution of the individual elements of the heart changes with develop- ment ; but how can we harmonize with this the fact, which has been mentioned above, that the heart of a four-to-five- DEVELOPMENT OF FISH EMBRYOS 299 days-old embryo comes to a standstill, and the embryo dies immediately (that is, in less than an hour), in a 1.5 per cent. KC1 solution, while an embryo of the same age which has been kept in this solution from the beginning continues to live in it, and may even show slight evidences of a heart- beat? To say that the embryo adapts itself or becomes accustomed to the poison gives us no new view of the ques- tion. Might it not be possible that the KC1 is the more poisonous the greater the work done by the heart in the unit of time, and in consequence the greater the chemical changes going on in it? According to this, it would be intelligible why a normal embryo, when put into a 1.5 per cent. KC1 solution, dies within a short time, while an equally old embryo which has grown up in the poisonous solution is alive. at the same time, and can even show evidences of a heart-beat. The heart of the former beats strongly, while that of the latter works only faintly — so faintly, indeed, that the blood does not even circulate. The embryo can live in a 0.5 per cent. KCi solu- tion as long as no great demand is made upon the activity of the heart. As soon as the heart begins to beat more strongly at the time of maturity, the embryo dies. This relation of the toxicity of the potassium to the development of energy in the protoplasm, or rather to the chemical changes deter- mining this development of energy, would hold not only for the heart, but also for all the other tissues. The entire ques- tion could be decided experimentally, if this has not already been done. 6. All the remaining organs, especially the brain, eyes, ears, and mesoblastic somites develop in the Fundulus embryo without a circulation, without apparent anomalies. Only in one place, where no one has thus far suspected it, did a depend- ence on the circulation show itself in an unexpected way- in the marking of the yolk-sac, and possibly (but I wish to 300 STUDIES IN GENERAL PHYSIOLOGY make further experiments in this direction) also of the embryo. The yolk-sac of the Fundulus embryo has a very characteristic tiger-like marking in the second week. Numer- ous chromatophores, which contain in part black, in part reddish -brown, pigment, develop on the surface of the yolk- sac of Fundulus. In the early stages of development, on the third day, no definite relation can be discovered between the circulatory system and the chromatophores. The chromatophores are scattered about irregularly upon the blood-vessels, and in the spaces between them. As soon as the circulation is established, however, the chromatophores begin to creep upon the vessels, and in the latter periods of the development, from the tenth day on, the chromatophores are no longer found in the spaces between the vessels, but have all crept upon them. But that is not all. The chro- matophores of the yolk-sac of Fundulus have the character- istic amoeboid appearance as long as they lie in the spaces between the vessels. Their diameter in any direction is greater than the diameter of an average-sized blood-vessel, and much larger than that of the capillaries. As soon as a chromatophore has reached a blood-vessel, however, it accommodates its entire mass to the surface of the blood- vessel, so that it finally loses its amoeboid appearance and apparently forms only a layer about the blood-vessel. The chromatophore cannot leave the surface of the blood-vessel after it has once reached it. This relation is most apparent where a blood-vessel branches. The chromatophore then branches in the same way as the blood-vessel. If the circu- lation of the blood is prevented by the addition of KC1 to the sea-water, the chromatophores and the blood-vessels both develop, but the chromatophores do not creep upon the blood-vessels. The tiger-like marking of the embryo- sac of Fundulus is apparently, therefore, a function of the circulation, in so far as the chromatophores are compelled to DEVELOPMENT or FISH EMBRYOS 301 spread over the surface of the vessels, in consequence per- haps of a chemical stimulus. Whatever may be the cause which compels the chromatophores to creep upon the blood- vessels, my observations certainly show that the distribution of the chromatophores, and therefore the marking of the yolk-sac, is dependent upon the arrangement of the blood- vessels. I will not enter upon this point in greater detail here, as a separate paper on this subject will appear in the Journal of Morphology. I wish only to point out that this is the first case, to my knowledge, in which the physiological explanation of the marking of an animal organ has been found. 7. Nowhere will the mystic find a richer field of unex- plainable purposefulness than in the developmental history of the higher animals. In these everything apparently comes into being at the right time and at the right spot, as though each element knew what role it had to play in the whole. The heart also begins to beat, apparently, just at the right moment; and I always had the idea — and others will per- haps have shared it — that if the activity of the heart were interfered with, development would soon cease. Our experi- ments, however, show that, if we do not consider the extreme cases, the development of an embryo in a KC1 solution can keep on normally for three days after the formation of the heart, even though no circulation is established. This lati- tude for the time of the beginning of the heart-beat is, when compared with the total time of development, very far from the precision expected of a clock-work. 8. In conclusion I wish to emphasize what seems to have been definitely established by these experiments, and what is yet to be determined by further experiments. I consider it certain that the origin, the pathway, and the branching of at least the larger blood-vessels are independent of the blood-pressure. For this reason it is possible for a vascular 302 STUDIES IN GENERAL PHYSIOLOGY system, which is identical with the vascular system of the normal circulation in the points mentioned, to develop even in the absence of a circulation. The mechanical causes for the growth of the vessel-walls are, therefore, not to be sought inside the vascular lumen, but in all or in single cells of the vessel-wall. The giving off of branches is determined by internal causes acting within the cells of the vessel-wall, or through stimuli arising in their neighborhood which affect these walls, as external stimuli affect the formation of stolons in Hydroids. It is possible, however, that the angles at which the branches arise from the main vessels do not corre- spond absolutely with the angles found in normal embryos. This point still remains to be investigated. Another ques- tion which I leave open is whether the circulatory system which is formed in the absence of a circulation is closed or not ; that is, whether the capillary branchings of the arteries of the yolk pass over into the capillary branches of the veins. Mere morphological study speaks in favor of this idea, but to settle this point definitely further experiments must be made. I consider it as certain that the tiger-like marking of the yolk-sac of Fundulus is dependent upon the vascular system. ON A SIMPLE METHOD OF PRODUCING FROM ONE EGG TWO OR MORE EMBRYOS WHICH ARE GROWN TOGETHER l 1. IN the effort to extend my work on heteromorphosis to the embryo, I have discovered a simple method of pro- ducing at will from a single egg two or more embryos which are grown together. My experiments were made on sea- urchins, but it is possible that they can be made with just as great certainty on every other holoblastic egg. Ten minutes after having been artificially fertilized in normal sea-water, eggs of Arbacia were introduced into sea-water to which 100 per cent, of its volume of distilled water had been added. The eggs absorbed so much water in the diluted sea- water that their membranes burst and part of their proto- plasm flowed out. The eggs then consisted of two connected spheres of protoplasm (P and Pl , Fig. 69), as the extruded drop of protoplasm in consequence of its surface tension assumes a spherical form, as does the protoplasm remaining behind inside the membrane. As segmentation has not yet begun at this time, only one of the two droplets contained a nucleus (Fig. 69). When after some time I returned these eggs into normal sea-water, each of the two spheres of protoplasm developed into an entirely normal and complete embryo. In many cases the two embryos remained connected. More often, however, one of the embryos went to pieces in the course of its early development (in about the morula or blastula stage); and finally many double embryos were gradually separated from each other, in consequence of their l Pfliigers Archiv, Vol. LV (1894), p. 525. 303 804 STUDIES IN GENERAL PHYSIOLOGY active movements in the blastula and gastrula stage. These latter isolated embryos continued their development nor- mally. In this way either separate or "Siamese" twins were formed from a single egg. It often happened that a repeated outflow of the protoplasm occurred, and then three or even a larger number of joined proto- plasmic drops were formed from one egg. In a number of cases, which was by no means small, I obtained, in consequence, joined triplets or quadruplets.1 I suspect, however, that in the gastrula stage many of these multiple embryos are separated from each other through the active movements of the egg, as triplets are relatively rare in the pluteus stage. On the other hand, it was a simple matter to obtain double plutei in large num- bers. All these double and triple plutei lived as long (about two weeks), and were as well and as complete in their form, as the plutei produced from a normal egg. 2. As has already been mentioned, the eggs were intro- duced into the diluted sea-water before the beginning of cleavage, and as only one nucleus was present at this time, only one of the two drops of protoplasm contained a nucleus. Nevertheless, both drops developed into a complete embryo. How did the drop of protoplasm which was at first without a nucleus obtain a nucleus ? This happened in a very simple way in the course of cleavage. Cleavage did not take place in the sea-water which had been diluted 100 per cent., but as soon as the eggs were returned to the normal sea-water, cleavage began. The first line of cleavage was perpendicu- lar to the common diameter of both spheres (Fig. 70). The 1 These facts have been questioned by one author on the basis of inadequate and imperfect experiments made by him. During my experiments on artificial parthogenesis I havo had a chance to verify amply the statements made in this paper. PRODUCING ACCRETE EMBRYOS FROM ONE EGG 305 cleavage sphere then developed in the normal way (Fig. 71), and finally a cleavage plane appeared in the extraovate (Fig. 72). In this way the nucleus becomes distributed through the egg. The further development is simple. The external form of the double sphere is maintained, while both parts divide into smaller cells; each of the two spheres forms a separate blastula cavity, which may communicate in certain cases, although they do not do this as a rule. The spheres FIG. 71 then continue to develop into gastrulse and plutei. I wish to emphasize especially that both embryos develop from the beginning as entire morulse and blastulse, and that no half- formation of any sort appears. In other words, the develop- ment goes on as if two independent eggs had been laid side by side, or had been glued together, and each had cleaved and developed entirely independently of the other. The protoplasmic connection of the two double embryos acts, however, differently from the way the gluing together of two eggs would do, as is evidenced by the deformities in the skeletal parts of the two plutei. 3. I have repeated these experiments with eggs in various stages of cleavage. Under these circumstances, also, the protoplasm always flows out in such a way that the cells remain joined together and double spheres are formed. I always obtained the same results, namely, double or multiple embryos. Only in the eggs which had developed very far — 306 STUDIES IN GENEKAL PHYSIOLOGY for example, such as were ruptured in the sixty-four-cell stage — did I usually obtain something different, namely, abnormally formed skeletons. Nevertheless at times I ob- tained under these conditions also real "Siamese" twins. 4. So far as the theoretical importance of these experi- ments is concerned, the following seems certain : As the burst- ing of the membranes and the flowing out of the protoplasm is a purely mechanical process determined by osmotic forces, no reason is at hand for assuming that qualitatively or quan- titatively the same constituents leave the egg in each case ; according to direct observation, the opposite seems to occur with much more certainty. Nevertheless, the extraovate develops into a complete embryo. It follows from this that every part of the protoplasm can form an embryo. So far as the nucleus is concerned, the two spheres obtain entirely different constituents of the nuclear substance. Nevertheless, they develop in exactly the same way to similar embryos, as Driesch has already shown by other methods. Thirdly, my experiments show that the number of embryos which can arise from an egg is determined by the geometrical form which we give the protoplasm, in so far as each completely or almost isolated sphere (or ellipsoid) of protoplasm deter- mines the formation of a separate blastula, and as the num- ber of blastulse determines the number of embryos. In this way I obtained last year double embryos from a normal egg when I introduced it, in the two-cell stage, into somewhat concentrated sea-water for some time, in which it was unable to segment further. When I brought such eggs back into normal sea-water, each of the two hemispheres broke up into several cells at once. The cells of each of the hemispheres probably adhered to each other, but not to the cells of the other hemisphere, so that an isolation of the two hemispheres was obtained in this way. In one egg I obtained double blastula3. The method deserves to be worked out more care- PRODUCING ACCRETE EMBRYOS FROM ONE EGG 307 fully. Herbst has apparently observed the same thing in eggs which he raised in salt solutions which had a qualita- tively abnormal constitution. The experiments of Driesch, who found that when the separate cleavage cells are isolated in the four-cell stage, each of the cleavage cells is still able to develop into a completely normal embryo, can also be explained in this way. When the mass of the protoplasmic sphere becomes too small, the formation of a blastula, of course, becomes impossible on geometrical grounds alone, for the size of the cells probably does not fall below a certain minimal volume during cleavage. Even though these experiments can leave no room for doubt that an embryo may arise equally well from every part of the protoplasm and from every part of the nucleus, as soon as these parts are isolated to a certain degree, and can assume a spherical or ellipsoid form, the fact nevertheless remains that in many eggs conditions are apparently present which determine the position of the median plane of the embryo and the further orientation of the various parts of the egg. These circumstances may, however, be purely sec- ondary in character, and may depend upon the mass and distribution of the nutritive yolk, the position of the micro- pyle, and similar secondary conditions. On the other hand, my experiments do not seem to agree with the assumption that each part of an egg can give rise only to a certain part of the embryo. 5. We must now raise the question whether the forma- tion of twins or double embryos in mammals can come about in a way similar to that described in these experiments. Driesch isolated individual cleavage spheres by rupturing the membrane through shaking. It is probably impossible that an egg can be ruptured in this way in the Fallopian tubes of a mammal. On the other hand, it may be possible that the scheme of my experiments corresponds to natural 308 STUDIES IN GENEKAL PHYSIOLOGY processes in the formation of twins. For I have found that from the moment of the entrance of the spermatozoon into the egg the osmotic pressure1 of the egg increases greatly. If unfertilized eggs are introduced into dilute sea-water, their volume increases relatively little. As soon as the sper- matozoon enters the egg, however, or when an egg which has just been fertilized is brought into the same salt solution, its volume increases very markedly, as I have determined by actual measurements. This fact shows that the spermatozoon brings about chemical changes in the egg which cause an increase in its osmotic pressure.1 I will return to the dis- cussion of this point in my more complete description of these experiments. So far as osmotic pressure is concerned, great differences exist between the eggs from one and the same individual. Even when the sea-water was only slightly diluted, a small percentage of the sea-urchin eggs burst ; and I do not doubt that this may occasionally happen in normal searwater. Whatever may be the actual process in the formation of twins from one egg in mammals, it seems probable that all multiple formations from one egg are caused primarily through complete or partial mechanical, or at least physical, division and isolation of the substances of the egg. 1 Or rather, its power of absorbing liquid. [1903] XI ON THE RELATIVE SENSITIVENESS OF FISH EMBRYOS IN VARIOUS STAGES OF DEVELOPMENT TO LACK OF OXYGEN AND LOSS OF WATER1 IT is probable that the series of successive changes in form which we call the development of an animal embryo is accompanied by a corresponding series of physiological changes. While we are well acquainted with the changes in form, so far as their mere morphology is concerned, we know but little concerning the changes in the physiological reac- tions of the embryo in its various stages of development. It is a well-known fact that the embryo has a greater vital- ity than the completely developed animals.2 Systematic investigations, however, are lacking as to whether this vital- ity decreases steadily with the progress of the development of the embryo, and as to whether this decrease is the same toward different variables. In order to obtain an answer to these questions, I studied the relative sensitiveness of the fish embryo (Fundulus) to lack of oxygen and loss of water in different stages of its development. I found in general that the embryo is the more sensitive to lack of oxygen, the older it is. Yet the sensitiveness increases more rapidly at first than later. On the other hand, the experiments on the effect of withdrawal of water gave a totally different result. The germ of the embryo is much more sensitive to loss of water in the first stages of its development (during cleavage and before the beginning of the formation of the embryo proper) than after the formation of the blastoderm, and its sensitiveness decreases with the increase in the devel- 1 Pflilgers Archiv, Vol. LV (1894), p. 530. 2 ZUNTZ, Pfliigers Archiv, Vol. XIV ; and PFLttGEK, ibid. 309 310 STUDIES IN GENERAL PHYSIOLOGY opment of the embryo. The details are given in the fol- lowing pages. I. THE RELATIVE SENSITIVENESS OF THE EMBRYO TO LACK OF OXYGEN 1. The fact that embryonal development soon stops, and that the embryo dies without oxygen, has been shown to be true so often by experiment that it is not necessary to discuss it here.1 I tried to determine, first of all, how long and how far development could go in various stages of development at the same degree of lack of oxygen; and, secondly, how long the embryo could remain exposed to the same degree of lack of oxygen in different stages of development without losing its power of development. The stages of development which were studied were the following : (1) the freshly fer- tilized egg; (2) the egg after the formation of the blasto- derm, but before the formation of the embryo, about twenty- four hours after fertilization ; (3) the egg after the beginning of the formation of the embryo, about forty-eight hours after fertilization (the embryo usually had at this time optic vesicles in which the lens was just being formed) ; (4) the embryo just after the circulation was established, seventy- two hours after fertilization ; and finally various later stages. The egg of Fundulus is especially well adapted to these ex- periments, because it is very tough, develops fully in the aquarium, and the fish hatches in the aquarium. The entire period of development takes in summer, at a tempera- ture of about 24° C., about twelve to fourteen days. The method used in these experiments was similar to that of Bunge in his well-known experiments on the need of oxygen in lower animals. a About 6 c.c. of potassium hy- droxide and pyrogallol are put into a test-tube (according to Hempel's directions). Into this test-tube is introduced a 1 For the literature on this subject see DttsiNG, PflUgers Archiv, Vol. XXXIII. 2 BUNGE, Zeitschrift fiir physiologische Chemie, Vol. XIV, p. 322. RELATIVE SENSITIVENESS OF FISH EMBRYOS 311 second smaller one containing the eggs and a little sea-water (about 2 to 3 c.c.). The out6r test-tube is then sealed. By putting splinters of glass into the bottom of the outer test- tube the smaller test-tube is kept above the surface of the pyrogallol. This arrangement allows one to observe the eggs, and at the same time to shake the apparatus in order to accelerate the absorption of oxygen. In some of the experiments the sea-water in which the eggs were con- tained was boiled, in others this was not done. The result was, however, not very different in the two cases. Lack of proper laboratory facilities did not allow me, however, to study how rapidly and how completely the. oxygen is absorbed by the pyrogallol in these experiments. But even if the absorption of oxygen was not complete in these experiments, it was nevertheless equally incomplete in all the experiments. Since we are interested in our experiments only in the relative sensitiveness to lack of oxygen, a quantitative determination of the oxygen absorbed was not absolutely necessary, if the lack of oxygen was only always the same. For the sake of brevity, I will designate the apparatus used for the absorp- tion of oxygen as an "oxygen vacuum." The temperature was usually 22-24° C. Fundulus eggs require a relatively high temperature for their development. 2. Some eggs of Fundulus were introduced, one-half hour after (artificial) fertilization, into a large number of test-tubes from which the oxygen was absorbed by the method given above. One of these sealed test-tubes was opened at different intervals, and the eggs were compared with eggs from the same culture which had remained in normal sea-water as a control. It was found that cleavage occurred in the oxygen vacuum, and at first even a little more rapidly than in normal sea-water. The latter was, however, probably only the result of the rise in temperature brought about in melting the glass for the purpose of sealing the 312 STUDIES IN GENERAL PHYSIOLOGY tubes. After twenty-four hours a blastoderm was formed in all the eggs contained in the oxygen vacuum. Then, how- ever, development stopped entirely, while it continued, of course, in normal sea-water. Never was the beginning of an embryo formed in such a series of experiments in the oxygen vacuum. Development continued in the oxygen vacuum only to a point which would have been reached in normal sea- water in about fifteen hours. An egg that ceased to develop in an oxygen vacuum had not necessarily lost its power of development. When brought back into normal sea-water, it could continue its development; only it was necessary that the egg had not been left too long in the oxygen vacuum. Eggs which were introduced into the oxygen vacuum immediately after fer- tilization could continue their development after they had lain for four days in such a vacuum at a temperature of 22° C. If they remained in the oxygen vacuum longer than this, they lost their power of development for all time. In these experiments the eggs were contained in only 2-3 c.c. of sea-water. One might think that this circumstance had affected the result. I therefore made control experi- ments in which the eggs were kept in just as little sea- water, but in the presence of an abundance of oxygen. In these experiments the eggs developed in an entirely normal way. 3. In the second series of experiments the eggs remained in normal sea-water for the first twenty -four hours after fer- tilization, and were then introduced into the oxygen vacuum. At this time a blastoderm, but no embryo, was formed. On the next morning an embryo with optic vesicles had formed in nearly all these eggs. The development of those eggs which remained in the oxygen vacuum then came to a stand- still, however. Development, therefore, again continued in the oxygen vacuum about as far as a fifteen-hour develop- KELATIVE SENSITIVENESS OF FISH EMBRYOS 313 ment would have gone in normal sea-water. These eggs O OO lost their power of development in the oxygen vacuum usually after about forty-eight hours. The maximum that I observed in one case was a continuation of development after a residence of fifty-five hours in the oxygen vacuum. An egg which is introduced into the oxygen vacuum twenty- four hours after fertilization loses its power of develop- ment much more quickly than an egg which is exposed to the same degree of lack of oxygen immediately after fertilization. If embryos are introduced into the oxygen vacuum forty- eight hours after fertilization, the retardation of develop- ment becomes more apparent. The formation of the embryo has already begun in these eggs, and the next elements in the further development of the egg would be the formation of pigment and the beginning of the circulation. Pigment is indeed formed, though less than normally, but no circula- tion is established. After remaining for thirty-two hours in the oxygen vacuum, these eggs had lost their power of development also. Seventy-two hours after fertilization the circulation is fully developed in an embryo, when it is left in normally aerated sea-water ; when at that time the embryo is placed into the oxygen vacuum, in about seven hours the heart- beat becomes very weak. When such an egg was returned to normal sea-water, however, the heart of the embryo began to beat again vigorously almost immediately, at first slowly, but then with so rapid an increase in the rate that the num- ber of beats became relatively great in a few minutes. These eggs lost their power of development in lack of oxygen in about twenty-four hours after introduction into the oxygen vacuum. In no case did such an egg recover when it had remained for forty-eight hours in the oxygen vacuum. The older the embryo, therefore, the more sensitive it is to lack 314 STUDIES IN GENERAL PHYSIOLOGY of oxygen. The young fish which had just hatched from the egg were still less resistant than the embryos. These experiments show that the sensitiveness of the embryo to lack of oxygen increases with development. This increase is very great at the beginning of development, so that a four-day old embryo, for example, has just as good, or even a better chance, for continuing its development when it has remained for all this time in an oxygen vacuum than when it has spent only the last forty-eight hours in the vacuum. This apparently paradoxical result is readily explained when we assume that the cells which are formed from the egg-cell during the first stages of cleavage are different chemically from the cells of the embryo which are formed later, so that the latter go to pieces more easily in lack of oxygen than the former. For this reason an egg which has just been fertilized may still be capable of develop- ment when it has spent four days in an oxygen vacuum, because it has developed only to the point of the formation of a blastoderm ; while an egg which is introduced into the vacuum after the formation of the embryo dies after forty- eight hours. We saw also that development can go 011 for about fifteen hours in the oxygen vacuum, especially in the first twenty-four hours after fertilization. Whether the conclusion can be drawn from this that cleavage can go on without oxygen, or that the oxygen was not completely absorbed in our experiments, must be determined by further experiments. II. THE RELATIVE SENSITIVENESS OF FUNDULUS EMBRYOS TO LOSS OF WATER1 The development of the form of an embryo is a function of processes of cell -division and growth. Both classes of processes are, as in plants, so also in animals, probably a function of osmotic processes.2 An accurate knowledge of 1 These experiments were made in the summer of 1892, at Woods Hole. 2 See Part I, p. 191. EELATIVE SENSITIVENESS OF FISH EMBRYOS 315 the dependence of processes of development upon the amount of water contained in the cells is therefore desirable. After some preliminary experiments with sea- water, whose concentration differed comparatively little from that of nor- mal sea-water, had shown that the Fundulus egg is most remarkably independent of the concentration of the sea- water, experiments were made with sea-water to which had been added 5, 7.5, 10, and 20 g. of NaCl to each 100 c.c. of sea-water. Experiments were also made with nor- mal sea-water and fresh water as controls. Freshly fer- tilized eggs of Fundulus develop perfectly normally in fresh water as well as in sea- water to which 5 g. of NaCl have been added to each 100 c.c. (that is, 50 g. per liter). When 7.5 g. of NaCl were added to each 100 c.c. of sea- water, a blastoderm was still formed; but only rarely an embryo, and if so the embryo had dwarf dimensions, and its development stopped before the optic vesicles were formed. In solutions to which 10 per cent. NaCl was added no embryo was formed. The segmentation of the eggs started in this case also, and occurred at first almost as rapidly as in normal sea-water, but it usually ceased at about the thirty-two-cell stage. I expected that these eggs would retain their power of development for some time after this, similarly to those kept in an oxygen vacuum. This was, however, not the case. Freshly fertilized eggs of Fundulus lost their power of devel- opment permanently after six to ten hours at a temperature of about 24° C. in sea-water to which 10 g. of NaCl had been added to each 100 c.c. When 20 g. of NaCl are added to each 100 c.c. of sea-water, the power of development of freshly fertilized eggs was annihilated in about three to four hours. The first segmentations, however, took place in such eggs. When Fundulus eggs are introduced into a 13.5 per cent. NaCl solution (or, more accurately, sea-water to which 10 g. 316 STUDIES IN GENERAL PHYSIOLOGY of NaCl have been added to each 100 c.c.), twenty-four hours after fertilization (before the beginning of the forma- tion of the embryo) not only an embryo is formed in these eggs, but the embryo develops every individual organ; the heart-beats and the circulation are established, and the embryo shows motions after several days. It lives in this concentrated salt solution for ten to fourteen days. Only in three points did the development of such embryos differ from that in normal sea- water: the embryos grew much more slowly in the concentrated sea-water than in normal sea- water; secondly, the development of the individual organs was somewhat delayed; and, finally, the yolk shrunk much more rapidly than under normal conditions. Only a relatively small percentage of the eggs which were introduced into concentrated sea-water twenty-four hours after fertilization were able to develop. When eggs were introduced, however, into the 13.5 per cent, solution forty- eight hours after fertilization, a larger percentage developed. After the third or fourth day the eggs could be transferred from normal sea-water directly into a 27.5 per cent. NaCl solution without interrupting development! Development continued for about three or four days. The circulation was not interrupted, even though the beat of the heart had become somewhat slower. The velocity of development and growth was the less the higher the concentration.1 We see, therefore, that the sensitiveness to loss of water is incomparably greater during the early period of segmenta- tion of the Fundulus embryo than later, and that even then the sensitiveness decreases somewhat with progressive devel- opment. The following experiment is suited to show the great difference in the sensitiveness before and after the for- 1 These experiments may perhaps find their explanation on the assumption that after twenty-four hours the permeability of the egg or the germ-cells is diminished, and hence the NaCl cannot longer enter fast enough in sufficient concentration to do harm. [1903J RELATIVE SENSITIVENESS or FISH EMBRYOS 317 rnation of the blastoderm. Some eggs were introduced into the 13.5 percent, solution one hour after fertilization; after five hours9 cleavage had gone on in these eggs to about the thirty-two-cell stage. Half of these eggs were then returned to normal sea-water, while the remainder were left in the concentrated solutions. Three hours later the latter had lost their power of development, with the exception of three among hundreds of specimens. The other portion of the eggs, after having remained for eighteen hours in normal sea- water, were returned again to the 13. 5 per cent. NaCl solu- tion. A large percentage of these eggs formed embryos which remained alive in this concentrated solution for more than a week. A remarkable phenomenon was observed in the embryos which developed in the very concentrated solutions; namely, that the maximum concentration of the sea- water in which the embryo is able to develop completely is much higher than the maximum of the solution in which the fully developed embryo is able to hatch from the egg. When normal Fundulus em- bryos are introduced in the second week of their development, into sea-water to which 10 per cent. NaCl was added they con- tinue their development, but they do not hatch, but die within the egg. This phenomenon shows itself still more defi- nitely when more dilute solutions are used. If 5 g. of NaCl are added to 100 c.c. of sea-water, the embryos develop in a normal way in this solution, and remain alive for more than five weeks without hatching. If, however, the eggs are intro- duced into normal sea-water as soon as the embryo is fully developed (after about two or three weeks), the embryo hatches in one or two days. This fact may be connected in some way with the fact that the fish which has just hatched is more sensitive to the concentrated sea-water than the two- day-old embryo. The former dies in a 13.5 per cent. NaCl solution in less than twenty-four hours. It can live, how- 318 STUDIES IN GENERAL PHYSIOLOGY ever, in a 6.5 per cent, solution. In fresh water the embryos hatch just as rapidly as in normal sea-water. The fish is able to live in fresh water. I know of no other animal which is able to stand such great and sudden variations in the concentration of the sea- water as the Fundulus embryo. The question might arise whether this depends upon a peculiar characteristic of the egg membrane or the protoplasm. It can be shown that the protoplasm is relatively insensitive to variations in concen- tration, while diffusion through the egg membrane occurs very promptly. In order to prove the latter I have to call attention to my earlier experiments on the action of KC1 upon the Fundulus embryo.1 The addition of 3 g. of KC1 to 100 c.c. of sea-water brought the heart of one of the older Fundulus embryos to a standstill in a minute. A consider- able amount of this salt must, therefore, diffuse through the egg membrane in a very short time. That the protoplasm is very insensitive to variations in concentration can be shown on the spermatozoa. I convinced myself, first of all, of the fact that the unfertilized Fundulus egg can form neither an embryo nor segment. I then introduced eggs, under bac- teriological precautions, into sea-water to which 5 g. of NaCl had been added to 100 c.c. When I added spermatozoa to so concentrated a solution, the eggs developed as in normal sea- water. The movements and the power of fertilization of the spermatozoa must, therefore, be retained in such a solu- tion. The spermatozoa also retain their power of fertiliza- tion in fresh water. I believe also that they still penetrated the egg in a 13.5 per cent, solution. I neglected, however, to follow this experiment more accurately, and so must leave this fact undecided for the time being. The fact that the spermatozoon retains its functions in an 8.5 per cent. NaCl solution is sufficient, however, to show that the independence 1 Part I, pp. 297, 298. RELATIVE SENSITIVENESS OF FISH EMBRYOS 319 of the Fundulus embryo to rapid and great variations in the concentration of the sea-water must rest (at least partly [1903]) upon properties of the germ-plasm. I will not try to say, however, what condition determines that the sensitiveness to loss of water is much greater during the process of cleavage than during the formation of the embryo. III. CONCLUDING REMARKS The experiments which have just been detailed were made with a view to obtaining further data for a theory of embry- onal organization, or — as this is ordinarily termed — for a theory of heredity. It seems to me that some of these theories — for example, Weissmann's theory of determinants — assume more for the germ-plasm than it contains. According to this theory, things are already definitely determined in the germ-plasm which, to my mind, are functions of circum- stances that first make their appearance in much later stages of development. To give an example: According to the theory of determinants, one would have to imagine that the marking of the yolk-sac of Fundulus is already prearranged by the spatial grouping of the determinants in the germ- plasm which are responsible for the marking, while I found that the marking is produced by the protoplasm of the chro- matophores being compelled through its "chemotropism" to spread over the surface of the blood-vessels. The formation of blood-vessels, as well as the formation of pigment-cells, may perhaps be traced back to the original germ, but the spatial arrangement of the pigment-cell is, as we have seen, the effect of a stimulus which the fully developed vessels, or rather the blood which they contain, exercises upon the fully developed chromatophores. These facts led me to the idea that the chemical circumstances determining organization do not all exist ready-formed in the germ-plasm, but arise gradually in the different stages of development. The 320 STUDIES IN GENERAL PHYSIOLOGY development of an embryo would, according to this, be in a physiologico-chemical sense an epigenesis and no evolution. In order to test this idea, I made experiments on the rela- tive sensitiveness of the embryo in the various stages of its development. I thought that sudden changes in the sensi- tiveness during the transition from one developmental stage to another would speak for epigenesis. Such a change in reaction was, indeed, found in the experiments on loss of water. I have begun experiments similar to those on Fundulus on Perca fluviatilis. The physiological reactions of the embryo of Perca are, however, very different from those of Fundulus, as was to be expected from the beginning XII ON THE LIMITS OF DIVISIBILITY OF LIVING MATTER1 1. THE progress which has been made in physics and chemistry as the result of our modern conceptions of mole- cules and atoms suggests the possibility that a more definite insight into the limits of divisibility of living matter might also be of importance for the development of physiology. As a criterion for "living matter" we might use the irrita- bility or spontaneity. But as the "spontaneity" of living matter is in its simplest form (in Amoebae) apparently not different from the physical phenomenon of spreading, for this criterion the limits of divisibility of living matter coin- cide with the limits of this purely physical phenomenon. But spontaneity is neither the deepest nor the most essential life-phenomenon; development — or, in other words, growth, organization, and reproduction — -occupies this place. If we ask how the ultimate elements of living matter are con- stituted which still possess the specific morphogenetic prop- erties, the excellent papers of Nussbaum give us a qualitative answer. This investigator found in experiments on the divisibility of an Infusorian, Gastrostyla, that only such pieces are able to regenerate into a complete animal as con- tain nuclear material. For the preservation of an Infusorian it is immaterial whether it is divided longitudinally, transversely, or obliquely; if only a portion of the nucleus is retained, the fragment regenerates into its original form in less than twenty-four hours, depending upon the temperature. As soon as twenty minutes after division the cut edges form new cilia, and upon the following day each of the pieces containing nuclear material possesses from four to six nuclei and nucleoli, and all the ciliary appendages characteristic of the species.2 1 Pflugers Archiv, Vol. LIX (1894) , p. 379. 2 NUSSBAUM, Archiv fiir mikroskopische Anatomie, Vol. XXVI, p. 514. 321 322 STUDIES IN GENEBAL PHYSIOLOGY A piece without the nucleus "cannot develop into the form characteristic of the species, and growth does not take place." Yet a piece without the nucleus, as Nuss- baum found, can move for a long time ; nuclear substance is, therefore, not necessary for "activity" or "spontaneity." Nussbaum draws the following conclusions from his ex- periments: (1) Nucleus and protoplasm can live only when united; isolated they die after a longer or shorter time. (2) For the maintenance of the morphogenetic energy of a cell the nucleus is indispensable. (3) Each of the energies produced by a cell depends upon a sub- stratum that can be divided. If I understand Nussbaum correctly, the latter statement means that a part of the nucleus and of the protoplasm is sufficient to render possible all the life-phenomena of the cell. Finally, I wish to quote also the following from the work of Nussbaum: The cell is not the ultimate physiologic unit, even though it must remain such for the morphologist. We are, however, not able to tell how far the divisibility of a cell goes, and how we can determine the limit theoretically. Yet for the present it will be well not to apply to living matter the conceptions of atom and molecule, which are well defined in physics and chemistry. The notion micella introduced by Nageli might also lead to difficulties, as the prop- erties of living matter are based upon both nucleus and proto- plasm The cell consequently represents a multiple of individuals. (P. 522.) The conception which we must therefore form of the nature of the simplest elements of living matter capable of development is this, that it consists of a system pf at least two different substances, of which the one is con- tained only in the nucleus and the other only in the proto- plasm. The experiments of Nussbaum have been repeated and amplified by a large number of careful observers. LIMITS OF DIVISIBILITY OF LIVING MATTER 323 Nussbaum's observations and conclusions were, so far as I know, corroborated in every particular. 2. In all these experiments the answer as to the quanti- tative limits of the divisibility of living matter has not been obtained. It is of great importance, however, to have a clear idea of how large the smallest piece of nucleus and protoplasm is that is capable of development. Is it of the order of magnitude of two or more micellae, or is it of the order of magnitude of a considerable fragment of the cell? I have tried to obtain an answer to this question in the sea- urchin egg. Pfltiger has stated already that the egg which had been considered as a unit can give rise to many indi- viduals.1 The experiments of Driesch, which we shall men- tion immediately, and my own experiments on the produc- tion of double and multiple embryos from a single egg, correspond with the views of Pfltiger. The question, there- fore, naturally arose as to how many embryos can arise from an egg, and to determine in this way the limits of the divisi- bility for one kind of living matter; namely, the egg. The simplest way of determining what fraction of the substance of the sea-urchin egg is still able to develop into a normal embryo seems at first sight to be that in which one of the cells of the egg in various stages of cleavage is isolated, and the last stage in which a single cell is still able to develop into a pluteus is determined. (The eggs cannot usually be kept beyond the pluteus stage in an aquarium.) The cleavage cells become smaller as cleavage progresses and the number of cells increases into which the egg divides. In another connection Driesch has shown that one of the cells from the four-cell stage of the sea-urchin egg is still able to develop into a pluteus.2 For our purposes, however, the methods and results of Driesch cannot be un- 1 Pflugers Archiv, Vol. XXXII, p. 562. 2 DRIESCH, Zeitschrift fur wissenschaftliche Zoologie, Vol. LV, pp. 5ff. 324 STUDIES IN GENEEAL PHYSIOLOGY reservedly employed, as it is questionable whether one of the cells of the eight- or sixteen-cell stage can really be con- sidered as identical with the eighth or sixteenth part of the unsegmented egg. It is possible that through the process of segmentation the substance of the egg is sepa- rated into unhomogeneous parts. It is further possible that the metabolism during segmentation changes the material contained in the various cleavage cells unequally. This might result, for example, in this, that one of the cells of the eight-cell stage would no longer be able to develop into a complete embryo, while one-eighth of the same egg before cleavage might have been able to form a complete embryo. A short time ago I published a method which enables us to divide the unsegmented fertilized egg into small pieces capable of development.1 The method consists in bringing the sea-urchin eggs, after fertilization, into sea- water which has been diluted through the addition of 100 per cent, of its volume of distilled water. The contents of the egg absorb water rapidly, and the thin egg-membrane ruptures at one or more points; a portion of the proto- plasm flows out of these ruptures, which assumes a spheri- cal form, and usually remains connected with the egg. (Fig. 74.) When the eggs are returned to normal sea- water, they begin to segment, the extraovate as well as the protoplasm which has remained in the egg form separate blastulse, and twins result from the egg. These may either remain attached to each other or separate later on. The latter usually occurs. It is possible, however, that not only two but several protoplasmic drops may exude from the egg, and in this way more than two embryos may develop from one egg. Finally, in some cases where only one extraovate exists, a separation of groups of cells may occur during segmentation which leads to the production of i Pflilgers Archiv, Vol. LV; and Biological Lectures Delivered at Woods Hole. 1893 (Boston: Ginn & Co.). LIMITS OF DIVISIBILITY OF LIVING MATTER 325 more than two embryos from an egg. If the eggs are made to burst before segmentation begins, only one nucleus is present, and in consequence either the contents of the egg or the extraovate must be without a nucleus. I have already mentioned in my earlier papers that in the course of the segmentation nuclear material gets into that portion of the protoplasm which was originally free from it. Occasionally the extraovate is cut off before nuclear division occurs. Nevertheless, cleavage occurs. From the observations of O. and R. Hertwig and of Boveri we may assume that in these cases a spermatozoon has entered the protoplasm. The nuclear material which is introduced in this way suffices to inaugurate the process of cleavage. 3. In these experiments it is natural, of course, that the extraovate, as a rule, does not contain exactly one-half the mass of the egg. Those cases in which the extraovate and the contents of the egg differ greatly in size are well adapted to decide how large an amount of the egg-substance is just sufficient to give rise to a normal pluteus. I followed the development of selected individual eggs with their extra- ovates in a drop of water contained in a moist chamber. I also examined very carefully from day to day cultures of such eggs kept in large vessels, and determined the size of the smallest plutei. Finally, I studied also from day to day the fate of these small fragments. The results of these ob- servations, which I pursued uninterruptedly for two months last year and again for two months this year, were very defi- nite, and may be expressed as follows: a) The smallest normal plutei which arose from fragments of an egg were linearly about half the size of the plutei aris- ing from a whole egg of the same culture. Their volume — their density being considered the same — was therefore about one-eighth of that of a normal pluteus. 6) Smaller fragments of an egg than these developed into 326 STUDIES IN GENERAL PHYSIOLOGY blastulse and reached the gastrula stage later than the larvae formed from entire eggs. In the most favorable case irregu- lar precipitates of calcium salts were formed, but the changes of form characteristic of the pluteus stage did not occur. These monsters did not develop beyond the spherical form of the FIG. 75 FIG. 74 gastrula. They lived, however, as long as the normal plutei, and, so far as motility was concerned, were comparable to normal embryos. 4. I will now enter upon these observations in somewhat greater detail, and will use for this purpose a series of draw- ings, all of which were made with the camera lucida at about the same magnification, Fig. 73 gives the form of a normal fertilized egg with its membrane. Fig. 74 gives the form of an egg with the extraovate E which has burst in sea-water. I have described, in a former paper, the first processes of cleavage which occur in such an egg. We will now follow the development of multiple embryos from such an egg. Fig. 75 shows a bursted egg in the twelve-cell stage. It can be seen that the micromeres form a bridge between the extrao- LIMITS OF DIVISIBILITY OF LIVING MATTER 327 vate and the egg. Four hours later this egg was in the condition shown in Fig. 76. The cleavage cells lying within the egg in Fig. 73 have developed into the blastula c. The micromeres and the large cells inclosed between them in Fig. 75 have developed into a misshapen mass of cells d in Fig. 76 ; each two of the four large cells of the extraovate have led to the development of the separate blastulsB a and &, so that we obtained from this one egg three blastulse and a mass of FIG. 77 FIG. 78 shapeless cells.1 Fig. 77 shows the same egg twenty-four hours later. The largest of the blastulse c, which has remained within the egg-membrane, has developed into a gastrula, while the two smaller blastulse a and 6, which have remained outside the egg, have developed no further. A short time after this drawing was made all four pieces began to swim about in the drop. The formation of the blastula therefore occurred at the same rate in the smaller masses as in the larger one. I may add that it also occurred at the same rate as in the eggs whose membrane had not burst. One notices, of course, that eggs which are placed into dilute sea-water, and consequently go into "water rigor," do not all recover and begin to segment at the same time after they are returned to normal sea-water. Under these conditions an extraovate i It often happened that the cells of the extraovate formed not one, but two or more, blastulse. The sliding motions of the cells are not restricted in the extraovate, and can therefore lead to various groupings of the cell-masses. Inside of the egg- membrane this origin of twins also occurs, but more rarely. The membrane restricts the sliding motions of the cells. I shall discuss this question in greater detail later in a paper on the formation of double embryos. 328 STUDIES IN GENERAL PHYSIOLOGY may begin to segment somewhat later than the rest of the egg. The influence of the quantity of the egg-substance upon the formation of the gastrula is, however, an unques- tionable one. The substance which remained within the egg has developed in Fig, 77, twenty-four hours after fertiliza- tion, to the gastrula stage. According to Fig. 75? its mass is about twice as great as that of each of the two blastulse which arose from the extraovate, and which at this time had FIG. 79 FIG. 81 FIG. 82 FIG. 83 not yet reached the gastrula stage. I have observed this over and over again; as, for example, in Fig. 78, where the small extraovate b has formed a blastula, while the rest of the egg a has formed a gastrula. The cultures kept in a drop of water always died in the course of the second or third day. The large pieces reached the pluteus stage dur- ing this time, while the smaller pieces remained in the blas- tula or gastrula stage. In order to follow the further fate of these small fragments of the egg after the second or third day, I had to rely on the material from the cultures kept in the larger dishes. Figs. 79-83 represent the relation between mass and development in a culture two days old. Fig. 73 shows the size of a fertilized but undeveloped egg of this culture two days old ; Fig. 79, the size of a pluteus that had developed from an entire egg, which, however, had been LIMITS OF DIVISIBILITY or LIVING MATTER 329 subjected to the same treatment with dilute sea- water as the pieces which are about to be described. Fig. 80 is a double embryo which had arisen from a bursted egg. The two embryos are unequal in size, and the larger is ahead of the smaller in development in so far as a deposition of needles of calcium carbonate has begun in the latter. Both, however, are less developed than the pluteus which has arisen from a whole egg. Fig. 81 arose from a fragment which was smaller than half the egg of Fig. 73. It is an early gastrula stage. Figs. 82 and 83 are still smaller frag- ments of an egg, which have, however, reached only the bias- tula stage. These ex- amples are not especially selected, but they represent only what the observ- er will find in any sample from such a culture. Do these small pieces develop to the pluteus stage? Two days later I found the conditions in this culture as shown in Figs. 84-87. Fig. 84 is one of the smallest fragments living at this time. It is only a blastula. Fig. 85 represents a larger piece in the gastrula stage, but without any evidence of a skeleton. Fig. 86 shows the smallest pluteus; Fig. 87, 330 STUDIES IN GENERAL PHYSIOLOGY a medium-sized pluteus which arose from a whole egg. If the linear dimensions of both plutei are compared, we find that they are about in the relation of one to two, which at equal density would correspond to a relation of their masses of one to eight. The smaller fragments generally do not go into a normal pluteus stage, but form only irregular needles of calcium salts, retaining, however, the spherical shape of the young gastrula. Figs. 88 and 89, for example, are such gastrulaB five days old from the culture under discussion. Fig. 90 is a smallest pluteus of the same age. These gastrulse with skeletal needles may grow, but their exter- FIG. 88 FIG. 89 FIG. 90 nal form usually remains unchanged, and the skeleton remains abnormal. Finally, I should like to add a few words regard- ing the fate of such misshapen heaps of cells as shown in Fig. 76d Their outer surface forms cilia, like that of nor- mal embryos, and like the latter they move with great rapidity in the aquarium; they seem to live as long as the plutei. These groups of cells represent free-swimming tumors, teratomas, which have arisen because of sliding motions of cells which could occur on a larger scale on account of the lack of a membrane. 5. We have seen, therefore, that the smallest pluteus with normal shape which arose from a fragment of an egg had about one-eighth the volume of the normal average-sized pluteus which sprang from an entire egg. I may add that I have never observed smaller normal plutei than this. In LIMITS OF DIVISIBILITY or LIVING MATTER 331 determining the limits of divisibility of living matter, it is of importance to decide whether such a pluteus arises from a piece of an egg the mass of which amounts also to one- eighth of that of the total egg. We must therefore know whether the embryos originating from fragments of an egg grow more rapidly or more slowly than those which arise from an entire egg. Now, as I have mentioned before, it is a general rule that the smallest pieces after having attained the blastula stage develop less rapidly than those which are formed from an entire egg. In my earlier experiments on growth and regeneration in Tubularia I found that develop- ment and growth are functions of the same variables ; there can be no doubt that to a certain extent development is only a function of growth. It is therefore probable that the embryo arising from a small fragment of an egg grows less rapidly than that arising from an entire egg. It is therefore also probable that a pluteus the mass of which amounts to only one-eighth of that of a normal pluteus has developed from a fragment of an egg which contained more than one- eighth of the substance of the entire egg. I will, however, not deny the possibility that a later observer may perhaps find a still smaller pluteus, even though the large number of my experiments renders this scarcely probable.1 But I believe that even in such a case the limit which I have given will suffer no great reduction. The divisibility of the egg is therefore very limited, if one demands that the fragment shall develop into a pluteus. 6. So far as my present experiments are concerned, I am not yet able to say where the limits of divisibility lie, when it is only required that the piece develop into a blastula. The tiniest pieces of isolated egg-protoplasm still divided if i Boveri has since stated that he found a pluteus whose linear dimensions were only one-third of those of a normal pluteus of the same culture. But as a slight retardation in the growth of the arms may easily lead to such a result, I think that, on the whole, the limits observed by me in many experiments will be nearer the tr th than the one exceptional observation made by Boveri. [1903] 332 STUDIES IN GENERAL PHYSIOLOGY they contained nuclear substance. So far as I was able to see, very small pieces developed into blastulsB. It is prob- able, therefore, that the egg can be divided much further when we wish only blastulse to develop than when we wish the fragments to reach the pluteus stage. It seems prob- able, however, from my observations, that the blastula must have attained a certain size before it is able to develop into a gastrula, so that the limit for the amount of egg material necessary for development into the gastrula stage is probably greater than that necessary for the development of a blastula. 7. We can here, in passing, answer a question which we have touched upon before, but which does not necessarily belong to our subject, namely: Do qualitative changes of a sufficiently profound nature occur in the cells during seg- mentation (besides the mere increase in their number), which interfere with the divisibility of the egg? As has been said, Driesch was able to convince himself that an iso- lated cell of the four-cell stage could still develop into a pluteus ; but this seems not to have been possible when he isolated one of the cells from the eight-cell stage. He attributed this, very correctly, to the amount of substance present. Others, however, were inclined to conclude from such experiments that in the eight-cell stage the differentia- tion in the individual cells had progressed so far that they could give rise only to individual tissues, but not to entire embryos. But it is apparent that the limits of divisibility of the egg in Driesch's experiments harmonize with those found in my own. If, therefore, one of the cells from the eight-cell stage is no longer able to develop into an entire embryo, this is to be attributed to the fact, as shown by my experiments, that the amount of material present in one cell in this stage does not suffice to form a pluteus. Besides this, of course, a differentiation might have occurred in the cells. I have made experiments which show that this cannot LIMITS OF DIVISIBILITY OF LIVING MATTER 333 be the case to a sufficient extent to interfere with the devel- opment. Fertilized eggs of Arbacia were kept in normal sea-water until they had reached the eight-, sixteen-, and thirty-two-cell stages, when the eggs were put into water which had been diluted to a sufficient extent. The mem- brane burst, and a portion of the egg contents flowed out, as in the case of the unfertilized egg, only with this difference, that in this case the extraovate consisted of a larger number of cells. Nevertheless, the result was the same as in eggs 7 oo which were made to burst before cleavage had begun. When the separated pieces were sufficiently large — greater than one-eighth of the entire mass of the egg — they developed into a pluteus; when they were smaller than this, they reached only the gastrula or blastula stage. If a differen- tiation had occurred in the eight-cell stage, so that the individual cells could give rise only to certain tissues, or- gans, or regions of the body, we would expect that, when only about eight of the cells from an egg in the thirty- two- cell stage were separated from the main mass, we should obtain only a conglomeration of different fragments of tis- sues, organs, etc., but not an entire embryo, which is, how- ever, not the case. It may also happen, as I have just mentioned in the case of the uncleaved eggs, that the cells in their sliding motions distribute themselves so that they do not form a blastula ; but this does not occur of tener in these eggs than in those which were ruptured before cleav- age began. 8. After this digression we will return to our main theme, and ask the question whether it makes any difference what portion of the protoplasm is cut out of an egg. It might be that the egg is not completely isotropic. The protoplasm always begins to flow out at the point at which the membrane is ruptured. It can easily be shown, how- ever, that the seat of the rupture may be at any point in the 334 STUDIES IN GENERAL PHYSIOLOGY cell-membrane, and that it certainly has no relation to the orientation of the first cleavage plane. For when we first allow the eggs to develop in ordinary sea-water into the two- cell stage before bringing them into dilute sea-water, it can be noticed that the first cleavage plane may lie in any posi- tion with regard to the point of rup- ture; the material lying nearest the rupture will be that which flows out. Figs. 91-94 are drawings of eggs the membranes of which were made to rupture in the two-cell stage, and which illustrate what has been said more clearly than words. Since the extraovate develops in all cases, if it is only sufficiently large, we must conclude that, so far as the question of divisibility is concerned, the protoplasm must be considered an isotropic substance. 9. What conceptions can we form of the nature of the smallest elements of living matter which are capable of development? As Nussbaum has shown, every attempt that has been made of assuming as the ulti- mate elements of living matter some- thing analogous to the atom and the molecule has failed, for the simple rea- son that two different substances, nucleus and protoplasm, are necessary. One might assume that a combination of two different "micellae "-—one composed of nuclear material, the other of proto- FIG- 92 plasm — might represent the smallest living element. Our experiments show that such an idea would be entirely wrong, when full capacity for development is taken as the criterion of living matter, inasmuch as a very considerable quantity of substance is necessary for full development — an amount LIMITS OF DIVISIBILITY OF LIVING MATTER 335 FIG. 93 which, according to our experiments, is not far removed from the limits of macroscopic visibility; I have empha- sized the fact that for geometrical reasons alone a certain amount of substance must be present before it is possible to form a pluteus. But the lowest limit actually found is reached very much earlier than that re- quired from geometrical considerations alone. Since it has been demonstrated that the ultimate source' of all energy for life-phenomena is of a chemical nature, we must conclude from our experiments that the ultimate unit of living matter is such a quantity of substance as is capable of developing that amount of energy which is necessary for that life- phenomenon which is used as a criterion. In this we find a natural explanation of why the amount of substance necessary for the formation of a pluteus must be much larger than the amount of substance which is sufficient for the formation of a blastula, inasmuch as a larger amount of living matter repre- sents also a larger amount of energy. It follows from this, also, that when one is satisfied with spontaneity or irritability as the criterion of living matter, the ulti- mate unit of living matter is not only much smaller quantitatively, but also different qualitatively, as the protoplasm alone suf- fices for this. In the case of ultimate units we not only deal with masses which represent a definite amount of chemical energy, but we have every reason for assuming that the mode of liberation of this energy follows a definite order, which is possibly the same for all life-processes. Our fur- FIG. 94 336 STUDIES IN GENERAL PHYSIOLOGY ther insight into the nature of these ultimate elements will consequently be dependent upon a knowledge of this order. This significance of the quantity of living substance as the carrier of a definite amount of energy is also apparent in the regeneration of multicellular animals. According to the experiments of Nussbaum, "at least one ectoderm, one entoderm, and one cell from the intermediary germinal layer is necessary" for the regeneration of a Hydra capable of reproduction.1 But this minimum gives us only a qualita- tive limit, in so far as the three qualitatively different ele- ments are necessary. So far as the quantity is concerned, it must be said that a very large multiple of each of these three elements is necessary for regeneration. In experi- ments on Tubularia which Miss Bickford made in my labo- ratory two years ago2 it was found that pieces 1 mm. long from the stem of this Hydroid are no longer able to regener- ate into complete Tubularise; either only a simple polyp without stem and root is formed, or a peculiar heteromor- phous formation, a sort of Janus head, occurs, consisting of two polyps connected with each other by their aboral ends, while the stem and root are missing between them. That the smallest amount of matter capable of develop- ment must have a different absolute size in different organ- isms— that, for example, it must be smaller for a coccus than for an Arbacia egg — need not be specially mentioned. 10. The results of our observations are briefly as follows : a) The limits of divisibility of living matter must vary according to the character of the life-phenomena used as a criterion of life. Each quantity of living matter is the bearer of a definite quantity of energy. 6) The smallest fraction of an unsegmented egg of Arbacia necessary for the formation of a pluteus is about 1 Archiv filr mikroskopische Anatomic, Vol. XXXV (1890). 2 E. BICKFOKD, Journal of Morphology, 1894. LIMITS OF DIVISIBILITY OF LIVING MATTER 337 one-eighth of the mass of the entire egg (nucleus plus pro- toplasm). c) The amount of substance necessary for the formation of a blastula is much smaller than that necessary for the formation of a pluteus ; for the formation of a gastrula more substance is probably required than for the formation of a blastula. d) It does not matter which position the fragments of an egg of Arbacia occupied in the intact egg; so far as divisi- bility is concerned, the protoplasm of the Arbacia egg can certainly be considered as isotropic. e) Since the limits of divisibility are almost the same in the unsegmented egg as in the first stages of segmentation (the thirty -two-cell stage included), it follows that (a) no qualitative changes occur in the egg during the early stages of segmentation which restrict the development of fragments of the egg, and that (ft) the individual cleavage cells, so far as the limits of divisibility of the substance of the egg are concerned, may be considered as equal. (In other respects, however, differences may exist between the individual cleav- age cells.) xm REMARKS OX REGENERATION ' I. ON THE REGENERATION OF THE BODY IN PANTOPODS 1. So FAR as I know, it is generally held that in Arthro- pods regeneration is possible only in the appendages, while segments of the trunk are not regenerated. According to experiments which I made at Woods Hole last year, Panto- pods, or at least one form of this group, Phoxichilidium maxillara, form an exception to this rule. The trunk of Phoxichilidium maxillara (Fig. 95) is about 1cm. long. The animal remains alive for weeks in a dish of sea-water. It is, like most of the free-moving inhabitants of the surface of the sea, positively heliotropic.2 If the body of one of these animals is cut in two by a transverse incision (at a, Fig. 95), each of the pieces is still capable of locomotion. If the pieces are exposed to the light, it is found that the oral piece continues to be heliotropic, while the aboral piece moves about independently of the light. The latter do not show much tendency to progressive motion. That injured Pantopods can remain alive has already been observed by Dohrn. The latter writes : I have observed that individuals continued to live for days even when all-of the extremities have been cut off. I have even cut a female specimen of Barana castelli in two, dissected the anterior portion of the body, and kept the posterior portion, carry- ing the extremities V to VII, alive for fully four weeks.3 1 ArchivfUr Entwickelungsmechanik der Orgamsmen, Vol. II (1895), p. 250. 2 Part I, p. 1. 3 A. DOHRX, Fauna und Flora des Golfes von Neapel; III, " Pantopoda " (Leip- zig, 1881), p. 81. 338 KEMARKS ON ^REGENERATION 339 I was unacquainted with this observation of Dohrn's, and so made more exhaustive experiments on the subject. I suc- ceeded not only in keeping alive for weeks pieces consisting of several segments, as did Dohrn, but even single segments of the body with only one pair of legs. Animals which I had cut longitudinally did not remain alive, yet I con- sider it possible that further experiments in this direc- tion may be accompanied by better results. 2. In animals in which the body was divided be- tween the second and third pair of legs (at «, Fig. 95) the segments of the body which were cut off were re- generated. This regenera- tion was especially marked in the oral halves of the animals which had to regen- erate the posterior segments (Figs. 96, 97). In the aboral halves of the animals I could discover only a swelling of the anterior end. In some cases a complicated regeneration was recognizable in this swelling. I had to discontinue my observations before the process of regeneration was completed. We will now discuss the regeneration of the posterior segments of the body in somewhat greater detail. The regenerated piece ab in Fig. 96 consists of three segments; yet the constriction at c is not as deep as is usual at one of these joints ; in Fig. 97 the regenerated piece ab consists of four segments instead of the three which were expected. The FIG. 95 340 STUDIES IN GENERAL PHYSIOLOGY presence of a supernumerary segment leads one to suspect that the regenerated piece might perhaps have developed into a leg in the course of time ; that, in other words, we might have had to do with a heteromorphosis, namely, the forma- tion of a leg in the place of the body segment which had been cut off. Hoek,1 however, states that the abdomen of Ammothea3 not infrequently shows traces of a segmentation. I had the animal shown in Fig. 97 cut into serial sections, which I examined microscopically. The intestine had grown into the anterior portion of the regenerated piece. The tissues were, however, but little differentiated. 3. The regenerated pieces ab of Figs. 96 and 97 did not make their appearance gradually and then grow steadily larger, but they suddenly appeared with the size and differ- entiation shown in the picture, while on the previous day no regeneration had been visible. As each operated animal was kept in a separate dish upon which its history had been written, and as each animal was examined daily, we must conclude that the regeneration and growth of the new pieces occur slowly under the skin, and that at the next molting the regenerated piece becomes suddenly visible. I have 1 HOEK, Archive de zoGlogie exp6rimentelle, Vol. IX (1881). REMARKS ON REGENERATION 341 observed that the injured Pycnogonides continue to molt. I expect to resume these experiments and to fill out the gaps left in this study. II. ON THE THEORY OF REGENERATION 1. Systematists have found it difficult to place the Panto- pods in the natural system upon the basis of their morpho- logical and developmental characteristics. It might be thought that under such conditions the consideration of the physiological behavior of the animals might offer advantages. On the basis of our observations on the regeneration of Pantopods it might appear as if the Pantopods were closely related to the Annelids ; but this conclusion would be inac- curate, for, as is well known, a whole group of Annelids, the leeches, do not regenerate body segments which have been lost, even though transverse pieces cut from the leech may remain alive for more than a year. One would therefore have to reason that the Pantopods are more closely related to the Chsetopods than the Hirudinese, which would be absurd. 2. One frequently encounters the statement that the capacity for regeneration in animals decreases the higher -animals stand in the natural system. This idea has been stated in a very definite form by Nussbaum: "The capacity of regeneration of organisms is proportional to their sys- tematic position, as determined by their characteristics, and decreases from below upward."1 This generalization goes too far, as can be seen from the facts mentioned above. Usually we find in every large group in the animal kingdom certain species with a greater, and others with a smaller power of regeneration. The salamander regenerates an amputated tail, inclusive of the spinal column and spinal cord, and this power of regeneration is almost as great as i NUSSBAUM, Sitzungsberichte der N iederrheinischen Gesellschaftfiir Natur-u. Heilkunde. Bonn, November 5, 1894. 342 STUDIES IN GENERAL PHYSIOLOGY that observed among the Chaetopods or Pantopods. On the other hand, the leech is not much more capable of regenera- tion after an injury to its body than the human being; both can only cover the amputated stump with skin. If one wishes to utilize the power of regeneration of animals for phylo- genetic purposes, this can be done only for members belong- ing to one and the same morphological group. According to Sachs, only "the forms of the same group may be con- sidered related to each other ; they have nothing in common phylogenetically with the members of another group ; every morphological group is, so to speak, a plant kingdom in itself."1 But whether even within the same phylogenetic group, in the sense in which Sachs uses the term, the power of regeneration of a species is a simple function of its position in the group can at present, from lack of facts, not be decided. My experiments on the functions of the brain in worms showed that no parallelism exists between these functions and the systematic position of each species. Much less does such a parallelism exist in regard to the tropisms which can be altered comparatively easily through external conditions. But though it is not correct to say that the power of regener- ation decreases the higher the animal stands in the system, it is perhaps true, that the number of species capable of complete regeneration is relatively greater in the groups of the Ccelenterates and worms than in the groups of Arthro- pods and vertebrates. 3. In general, it will also be found correct that the power of regeneration is greater in the embryo than in the adult animal. The young larva of the frog regenerates an ampu- tated leg, while this is not possible in the adult animal. Those who assume that the power of regeneration is the greater the lower the position of the animal in the natural 1 J. VON SACHS, Flora, 1894, p. 219. REMARKS ON REGENERATION 343 system will find this behavior of the embryo in harmony with the "biogenetic law." The explanation of the fact that the embryo possesses a greater capacity for regeneration than does the adult animal follows in a simple way from Sachs's theory of organization. Sachs assumes that the form of organs is determined by spe- cific substances, and that we have just as many specific mor- phogenetic substances in a plant as there are different organs present in it; but these substances are by no means all preformed in the germ; they originate through chemical changes from the germ substance during the process of development. If at first only those substances which lead to the formation of stems and roots are present, these, under the influence of external conditions, finally give rise little by little to another category of substances, which finally present themselves, in their purest form, in the male and female sexual cells. We can imagine this process as similar to the processes which follow one another in a chemical factory, where from the original raw material chemical compounds of the most varied kind gradually result, until finally the most val- uable product, perhaps in an exceedingly small amount, is obtained in a pure form.1 In the sense of this theory, we must assume that there are at first present in the animal egg only specific ectoderm and entoderm substances, from which, through chemical changes during the process of development, such compounds originate as are specific for epithelial cells, liver cells, periost cells, etc. If now we ask how, according to this theory, e. g.9 an animal which shows complete power of regeneration (Pla- naria torva), differs from one which is able only to cover the amputation stump with skin (Hirudo), the answer will be, it seems to me, as follows: In the leech all the original embryonic material (the egg substance) is used up in the production of the specific organogenetic substances, while in 1J. VON SACHS, Arbeiten des Botanischen Institute in Wiirzburg, Vol. II, p. 457. 344 STUDIES IN GENERAL PHYS-IOLOGY Planaria not all the "raw material" is consumed; yet enough substance for the formation of epithelium is still present in the leech for the repair of defects in the epithelium. During embryonal development the more "raw material" is changed into the specific organogenetic substances of the different organs, the larger the number of organs that are formed; in other words, the further the embryo progresses in the process of development. It is, therefore, easily intel- ligible why in some animals in which the "raw material" is present in only limited amounts regeneration is more com- plete in the earlier stages of the embryo than in the later stages. If the power of regeneration is dependent in this way upon the difference between the formation and consump- tion of the "raw material," it is not to be expected that the power of regeneration should be a function of the position of a species in the natural system. We can easily understand that this difference may be relatively great in one form, while in another form of the same group it drops to zero sooner or later during the embryonic development. We may expect, however, that we shall meet the latter state of affairs rela- tively more frequently in the more highly differentiated groups of Arthropods and vertebrates than in lower groups ; which, indeed, seems to be the case. An idea of the role of the organogenetic substances in regeneration has been given by Sachs in his paper on "Stoff und Form der Pflanzenor- gane." I have shown in several papers that the processes of regeneration, heteromorphosis and ontogenesis in animals, agree with the ideas of Sachs.1 1 Part I, p. 115 ; and On Some Facts and Principles of Physiological Morphology (Lectures Delivered at Woods Hole, XIV CONTRIBUTIONS TO THE BRAIN PHYSIOLOGY OF WORMS1 IN his well-known work Ueber Entwicklungsgeschichte der Thiere K. E. von Baer asks how the anterior pair of ganglia of segmented animals should be designated. Whether the first pair of ganglia of the segmented animals shall be called a brain or not depends entirely upon the significance which the word brain is given. It is certainly not the organ which we call the brain in vertebrates, for in them it is the anterior extremity of the neural tube, and this is lacking in the segmented animals. It is rather the foremost pair of the series of ganglia, and as the latter is to be compared with the spinal ganglia of the vertebrates the so-called brain of the segmented animals seems to correspond to the Gasserian ganglion of the vertebrates,2 for the latter also receives sensory impulses If , however, one wishes to designate by the term brain not a definite organ, but .... that mass of nervous tissue which receives the sensory impulses, then one can, of course, say that insects possess a brain. Only one must keep the meaning of this term in mind. He who seeks a definition for the word brain will find the necessary directions in von Baer's remarks. Steiner3 considers it the problem of the physiologist to find such a definition, and has come, apparently without knowledge of the remarks of von Baer, to a different definition which reads as follows : The brain is characterized by being the general center of loco- motion in connection with the activities of at least one of the higher sensory nerves Besides its simplicity this definition 1 Pflilgers Archiv, Vol. LVI (1894), p. 247. 2 Vol. I (1828), pp. 234 ff. 3 Sitzungsberichte der Berliner Akademie der Wissenschaften, 1890. 345 346 STUDIES IN GENERAL PHYSIOLOGY has the further advantage of being satisfied by a single experiment, in so far as, of the two elements of which the definition is com- posed, the one element is always given anatomically. This is the higher sensory nerve, the presence of which guarantees its func- tion. The only experiment which it is necessary to perform has to prove that the general center of locomotion is also present beside the sensory apparatus. This proof is furnished when the unilateral removal of the central nervous portion in question so alters the direction of the movement of the animal that it no longer moves in a straight line but in a circle — a phenomenon which is generally designated by the term " forced movements." This definition of Steiner leads to two new conclusions: first, that the cerebrum in human beings does not belong to the brain since, as is well known, unilateral removal does not bring about forced movements. Steiner himself realizes this, for he has found, aided by his definition, that the octo- pus "has a cerebrum but no brain." "To have no brain and yet a cerebrum seems strange and even paradoxical, probably only, however, because we have not until now encoun- tered such a case." The second necessary conclusion from Steiner' s definition is that the ear is a brain. This conclu- sion has not been drawn by Steiner himself, but is unavoid- able. For, first, unilateral extirpation of the ear brings about forced movements, and, secondly, the auditory nerve is one of the higher sensory nerves. Steiner further points out that bilateral destruction of an organ, the unilateral destruction of which brings about circus motions, renders impossible spontaneous progressive movements. Forced movements can be brought about in the shark from the medulla oblongata, and here is located also, according to Steiner, "the general center of locomotion" of this animal. Steiner has himself shown, however, that a shark still moves spontaneously after the loss of the entire medulla oblongafa.1 This "center of locomotion" is said by Steiner to be located in the medulla oblongata in the frog also, but 1 STEINER, Die Functionen des Cenlralnervensystems, Vol. II (1883), pp. 56 if. BRAIN PHYSIOLOGY OF WORMS 347 FIG. 98 Schrader has found that a frog is possessed of an irresistible impulse to move after losing this center.1 The simplest facts of comparative physiology show more- over that the power of progressive movement is possessed also by such organisms which have no brain whatever, i. e., the swarm spores of Algse. It is, in my opinion, not the problem of physi- ology to find a definition for an organ but to discover the functions of a given organ. From this standpoint I wish to make in the following pages some contribu- tions to the brain physiology of worms. I understand in this paper by the term brain, as is customary, the ganglia lying at the oral end of these animals. Brain physiology has shown that for the higher animals the biologi- cal character of a species, that is, the sum total of those reactions of a species which are determined by the external surroundings, depends chiefly upon the brain. I was espe- cially interested in determining whether the rudimentary brain of such low animals as the worms has a similar signifi- cance. The experiments which I wish to report have been made at long intervals, some in Naples in 1889, some in Woods Hole in 1893. II I. EXPERIMENTS ON THYSANOZOON BROCCHII 1. Thysanozoon is an elliptically shaped marine Planarian (Fig. 98, according to Lang), which is from one to three cm. long, and almost as broad. The brain g of the animal, an unpaired organ, is situated at the anterior extremity of the body, which latter can be recognized without difficulty by i SCHEADER, Pflugers Archiv, Vol. XLI. 348 STUDIES IN GENEKAL PHYSIOLOGY its possession of two tentacles (Fig. 98). From the posterior side of the brain arise the two large nerves (•/?, Fig. 98), which traverse the entire length of the animal. A number of other nerves also run to the brain. The nerves contain separate ganglion cells, a characteristic which, as is well known, is found also in certain peripheral nerves in the higher animals. The nerves form a plexus at the periphery.1 The central nervous system of this animal therefore con- sists mainly of the ganglion situated at the anterior extremity of the animal. As is the case with all Planarians, Thysano- zoon creeps over the walls of the aquarium or along the surface of the water. It differs in its movements from the fresh-water Planarians only in so far as it is able to execute actual swimming movements. In swimming the animal executes pendulum-like movements with the lateral portions of its body as does a butterfly with its wings. If a Thysanozoon is cut across transversely into two halves, while the animal is moving at the surface of the water, the posterior aboral half at once falls to the bottom like a dead mass, while the oral piece which contains the brain continues to move along quietly. If the cut is made rapidly and with a sharp pair of scissors, the behavior of the oral piece gives no indications of such reactions as accompany the sensation of pain in higher animals. If the animal is divided with a sharp knife while it is creeping over a glass plate, we notice the same phenomenon: the oral piece continues its motion undisturbed while the progressive movements of the posterior piece cease at once. It happens occasionally that the transverse division of a Thysanozoon causes the oral piece to execute more lively and rapid move- ments. That the progressive movements of this animal are indeed a function of the brain is shown particularly well when only the exceedingly small piece which contains the 1 Mittheilungen aus der zoologischen Station zu Neajel, Vol. I. BRAIN PHYSIOLOGY OF WORMS 349 FIG. 99 brain (Fig. 99) at the anterior end of a large (about 3 cm. long) Thysanozoon is cut off. The very small piece lying anteriorly then continues to creep or swim while the comparatively enormously large body executes no further progressive movements. The spontaneity of progressive movement in Thysanozoon is therefore a function of the brain. Both pieces of a divided Thysano- zoon continue to live and regenerate the missing portions. Only the oral piece regenerates more rapidly than the aboral, which has to form a head. I have not studied whether the latter forms a new brain. I kept such pieces alive for four months. The spontaneity of the aboral piece never returned, while the spontaneity of the oral piece persisted. 2. The beheaded frog will remain upon its back when the cut lies behind the medulla oblongata; when the cut lies in front of the medulla the frog will not remain upon its back. We assume in this case that the geotropic func- tions of the ear compel the frog to resume the normal orientation. It is probable that the tactile stimuli also act in such a way that they compel the frog to bring the soles °f its feet in contact with the surface of solid bodies, or to allow the weight of its body to press upon the nerve endings in these portions of the skin. I designated the fact that an animal is compelled to orient its body in a definite way toward the surface of solid bodies as stereotropism. Geotropism cannot be demonstrated in Thysanozoon as the animal assumes any orientation toward the center of gravity for a long time. Stereotropism is, however, present as the animal is compelled to bring its ventral surface in contact with solid bodies, or to allow its 350 STUDIES IN GENERAL PHYSIOLOGY body to come to rest upon its ventral surface. We cannot compel the animal to bring its back in contact with solid bodies, and at the same time expose its ventral surface to the water.1 The question now arises whether these phenomena of orientation are a function of the brain as are the spontaneous progressive movements. Strange to say this is not the case. The brainless Thysano- zoon returns to the ventral position when it is laid upon its back, only the reaction occurs more slowly than in the normal animal, or in that portion of the animal containing the brain. Reactions to light could not be demon- strated. 3. If, instead of making a complete transverse section of the animal, only the longitudinal nerves are cut, and the two pieces are left united with each other by a very thin bridge of protoplasm at one side (Fig. 100), the aboral piece is not innervated directly by the nerves from the brain. A conduction of the impulse by way of the lateral nerve plexus is, of course, still possible. When, immediately, after the operation, I laid such an animal upon the bottom of the aquarium the oral piece at once began to move, while the aboral piece tried to attach itself to the bottom. It responded, however, to the pull which the oral piece exerted upon it, and took part in a per- fectly co-ordinated manner in the progressive movements, as if no interruption had occurred. After some time the oral piece turned about, crept over the back of the aboral piece whereby the latter was dragged along passively, and laid 1 The righting of starfish which have been laid upon their backs is also only a case of stereotropism, and has nothing to do with the effects of gravity. BRAIN PHYSIOLOGY OF WORMS 351 upon its back. The posterior piece at once resumed the ventral position, however, and then moved actively in the same direction as the oral piece. Changes in movement, therefore, were inaugurated only by the oral piece which contained the brain and were never communicated directly to the aboral piece. When, however, the oral piece moved for some time in the same direction and with the same velocity, the same movement soon took place in the aboral piece also. The aboral piece therefore, behaved not entirely as a piece without the brain, as it was still capable of progressive movement, but not as a normal Thysanozoon either, as it had lost its spontaneity. This becomes still clearer from the following observations : I threw the animal into a basin of water. Both pieces executed energetic, synchronous swimming movements. The oral piece soon reached the vertical glass wall of the aqua- rium. In consequence of a change in the direction of the movement of the anterior piece, the connecting bridge between the two halves of the animal was twisted and the aboral piece came in contact with the glass wall with its back, while the ventral surface of the same was turned toward the water. The posterior piece now executed swim- ming motions and so followed the creeping movements of the oral piece. ' That when movement is constant the posterior piece takes an active part in the progressive move- ment, and is not merely dragged along passively, is further shown by the fact that it often crept with its free edge upon the back of the oral animal, especially when the latter sud- denly moved more slowly. The experiments detailed thus far show that a brainless piece of Thysanozoon no longer moves spontaneously, that is to say without an appreciable external stimulus. I did not succeed in bringing about progressive movements in a brain- less Thysanozoon even by stimulating it. If the animal is 352 STUDIES IN GENERAL PHYSIOLOGY touched a local contraction occurs at the stimulated point, but no progressive movements. 4. If a lateral piece ab is cut from an animal parallel to the median plane (Fig. 101), a contraction of the wound results which may be so great tha%t the piece rolls up into a spiral. (This contraction of the wound-edges occurs no matter what the position of the cut.) The lateral piece afr, which contains no brain, executes no pro- gressive movements. If, however, a contraction of the wound-edge occurs in the other piece, so that it is rolled into a spiral, the progressive move- ments no longer occur in a straight line but in a circle. I never succeeded in bringing about circular movements through unilateral destruc- tion of the brain in Thy- FIG. 101 sanozoon. II. EXPERIMENTS ON PLANARIA TORVA 1. The brain and nervous system of the fresh water Planarian (Fig. 102, according to Jijima) are so analogous to those of the marine Planarian, that it is unnecessary for our purposes to give a separate description of them. The most important difference exists perhaps in the fact that the two longitudinal nerves contain certain collections of ganglion cells. One might think that the brain function of the fresh -water Planarians might also be analogous to those of the Polyclads. That is, however, not the case. We experience here again what I have pointed out repeatedly in my papers on the lower animals: that animals which are very closely related morphologically may show the greatest FIG. 102 BRAIN PHYSIOLOGY OF WORMS 353 differences in their physiological reactions. If a fresh-water Planarian is cut in two the aboral piece which contains no brain creeps about in just as lively a manner as the oral half. The spontaneity of the progressive movements in Planaria torva is therefore in no way a function of the brain. Every piece of the animal, not too small, possesses spontaneity. The decapitated animals creep with the oral end directed forward, as do normal animals. 2. In a previous paper I have described the behavior of Planaria torva toward light. The animals are chiefly photokinetic, that is to say, changes in the intensity of the light alter their movements. If the animals are suddenly brought from darkness into light they begin to move. During the first few moments the direction of the move- ments is also influenced by the light. The animals move as do negatively heliotropic animals to the room side of the vessel, but they do not collect here as do negatively helio- tropic animals, but distribute themselves in all directions, and now begin to move in every direction, to come to rest finally in that region of the vessel which is more weakly illuminated than its surrounding. One receives the impression therefore that an increase in the intensity of the light causes the animals to move, while a decrease in the intensity of the light causes them to come to rest. For this reason one finds the animals through the day collected in relatively dark places in the vessel, or on the under-surface of stones. I suspect that the animals begin to move anew at night, and then at the approach of day again collect in relatively dark places. I repeatedly covered one- half of the glass vessel in which I kept the Planarians with black paper in the morning. No change occurred through the day. On the next morning, however, I found all the animals under the covered portion of the aquarium. This could be interpreted only as showing that the animals crept 354 STUDIES IN GENERAL PHYSIOLOGY about in the vessel during the night, and came to rest in the morning in the darkest regions. These animals possess at the oral pole not only a brain but also comparatively well-developed eyes. I decided to test whether a decapitated Planarian still shows the same reactions toward light as normal Planarians, in spite of the loss of brain and eyes. This is true in a most surprising way. About sixty specimens of Planaria torva were cut across transversely, close behind the brain and the eyes, in the evening. All the pieces were put into a vessel having vertical sides and half covered with black paper. On the next morning nearly all the posterior pieces as well as the oral pieces were found in the covered part of the aquarium. They were fairly uniformly distributed here. A few were found in the uncovered portion of the vessel, but among these there were head pieces as well as aboral pieces crowded together in one corner in the room side of the dish. In this corner the intensity of the light was relatively low. In repeating this experiment with normal animals I obtained the same results as with the decapitated animals. When the decapitated animals had collected in the covered portion of the vessel and had become perfectly quiet, their quiet was quickly disturbed when the dark paper was suddenly removed without jarring the vessel. The animals began to move, crept at first toward the room side, and finally collected again in regions where the light was least intense. This reaction also occurred as in normal animals, with this difference, however, that the reaction time of the brainless animals to changes in the intensity of the light was greater than in normal animals. In the animals possessing a brain and eyes the reaction began about one minute after light struck them; in the brainless pieces after about five minutes. In these experiments only diffuse daylight was used as a stimulus. I have pointed out before that when uninjured Planarians BRAIN PHYSIOLOGY OF WORMS 355 are put into a vessel having a cylindrical form, they do not collect as do purely heliotropic animals on the window or room side of the vessel but on the right and left sides of the same. The Planarians from which the head together with the brain and the eyes have been removed behave in exactly the same way. All these experiments are successful as early as the day after the operation. Only perfectly fresh material must be used for these experiments. In warm weather the posterior pieces regenerate a new head with eyes and probably a brain after a week. 3. It was formerly the custom to conclude that an animal possessed eyes when it reacted to light. And since no one doubted that the reaction to light was a reflex movement it was assumed that such animals possessed a central nervous system also. When I found that the orientation of animals toward light is determined by the same conditions as the orientation of plant organs toward the same stimulus I drew the self-evident conclusion that the orientation of the animals toward the light could not possibly rest upon characteristics which are possessed only by the eyes or only by the brain, as plants do not possess such organs. The sensitiveness of the eye to light must rather rest upon the fact that the eyes have a condition in common with heliotropic plants, namely, elements which suffer some change under the influence of the light. For the rest, however, these elements need not be either morphologically or chemically identical in the dif- ferent organisms. An analogous conclusion may be drawn for the brain. When an animal with a brain shows the same reaction as a plant which possesses no brain it follows that the reaction toward light is not the result of a "specific energy" of the brain, but that the brain in that case only performs a function which is possible in plants without nerves. This function need be nothing else than the con- duction of the light stimulus, which, as is well known, occurs 356 STUDIES IN GENEEAL PHYSIOLOGY in plants also, but in the latter case through different mechanisms. These facts which I considered as self-evident have been misunderstood by one author and have been falsely represented, as though I had asserted that brain and eyes are superfluous. From what I said it also followed that in animals which possess eyes the sensitiveness to light need not by any means be limited to the eyes. Photosensitive elements may be found also in other portions of the surface of the body. Graber has in fact already reported experiments on Tritons and angleworms which he claims point in the same direction. It could, moreover, be foreseen 'that an effect of light might perhaps be possible in many animals without the existence of a true reflex arc, for I found in Ciona intestinalis that a typical reflex phenomenon continues to exist in this animal after destruction of the central nervous system. There was finally the possibility also that both conditions might be found together in the same animal. The latter is the case in Planaria torva, determined to be such by Dr. Wheeler. 4. After what has been said it is scarcely necessary to emphasize the fact that brainless pieces of Planaria torva will not assume the dorsal position any more than will uninjured animals. All my experiments to bring about forced move- ments in these animals through unilateral destruction of the brain remain as fruitless as in Thysanozoon. III. EXPERIMENTS IN CEREBRATULUS MARGINATUS The Nemertines possess a more highly developed brain than do the Planarians. The brain is larger and shows also a greater subdivision into smaller parts. It also is continued into two lateral nervous cords. The latter contain "a super- ficial covering of ganglion cells which may give rise to ganglion-like swellings at the points where the nerves branch."1 1 CLAUS, Lehrbuch der Zoologie, 4. Aufl., p. 336. BRAIN PHYSIOLOGY OF WORMS 357 The specimens of Cerebratulus which I used in my experiments were over 50 cm. long and almost as thick as a finger. The layman would readily have believed that he was dealing with an eel instead of with a worm. The ani- mal lives in the sand. If it is laid upon the sand-covered bottom of the aquarium it soon buries itself in the sand. If the head is amputated from such an animal the head-piece continues to bury itself in the sand when it is not too short. The body, on the other hand, does not make a single attempt to bury itself in the sand. IV. EXPERIMENTS ON ANNELIDS The Annelids possess besides a fairly complex brain a chain of ganglia which transverse the entire length of the body. We have the experiments of B. Friedlander and of Graber on the function of the central nervous system of the Annelids. Friedlander1 amputated the anterior and posterior seg- ments from angleworms. The latter conduct themselves, to put it shortly, as normal ani- mals : they soon bury themselves in the earth. Not so the beheaded worms. Immediately after the operation they execute violent winding motions, perhaps creep about for some time, but usually come to rest after a short time, and can now remain quietly upon moist earth covered with moist filter paper, for days and weeks without, apparently, making any autonomous movements after the wounds have healed. Every stimulus however soon awakens them from their passive condition. They then move about energetically, even creep some distance, but soon fall back into their original lethargy. The second series of experiments of Friedlander consisted in excising a small (5 to 10 mm. long) piece from the abdomi- nal nerve-cord of angleworms. Friedlander had expected that the two portions of the worm lying anterior and posterior to the operated point would conduct themselves physiologically during i Biolofjisches Centralblatt, Vol. VIII. 358 STUDIES IN GENERAL PHYSIOLOGY locomotion as two individuals, or perhaps that the posterior portion would simply be dragged along passively. Neither, however, occurs. The animals which lack entirely a portion of the central nervous system creep about as normal animals, except for a slight variation which will be described later. A contraction of the longitudinal muscles begins in the anterior segments which passes posteriorly, reaches the operated point, jumps over it, and continues behind it, so that the movements of both parts are co-ordinated in exactly the same way as in the normal animal. To explain this remarkable phenomenon Friedlander assumes that a "longitudinal pull" is exerted upon the poste- rior segments through the contraction of the anterior seg- ments. "This acts as a stimulus to the stretched portions of the abdominal nerve cord, and the reflex brought about thereby consists of a contraction of the longitudinal muscles of the stretched segments." The longitudinal stretching of the skin (not of the abdominal nerve cord) would therefore lib- erate reflexly a longitudinal contraction. The correctness of this idea was proved by the following experiment. An angle- worm is cut in two in the middle and both pieces are sewed together in such a way that they are connected by a thread about 1 cm. long. The pieces when connected in this way by means of a thread execute co-ordinated movements. Graber tested the statement of Hofmeister and Darwin that the anterior end of the body of the angleworm is sensi- tive to light.1 He amputated the anterior segments of angle- worms, and found that the brainless pieces were still sensi- tive to light. The reaction of the angleworm to light is therefore no function of the brain alone. My own experi- ments consisted in amplifying these facts in some directions. V. EXPERIMENTS ON NEREIS If a Nereis is cut into several pieces, only the oral piece retains the power of burying itself in the sand. Earlier i Grundlinien zur Erforschung des Helligkeits- und Farbensinnes der Thiere, Prague (1884;, p. 290. BRAIN PHYSIOLOGY OF WORMS 359 experiments had led me to suspect that this "spontaneous" or "instinctive" burial was only a reflex called forth through the stimulus of contact with sand. I tried whether it would not be possible under certain conditions to demonstrate the same reflex, even in brainless pieces. I laid such a brainless piece upon the sand; as usual it remained quiet. I now carefully covered the anterior end of this piece with sand. The rest of the animal soon began to execute the typical movements which are necessary to bring about the burial of the animal in the sand. At the same time the glands at the foot end of the animal began to secrete the sticky substance which cements together the grains of sand and renders solid the wall of the tube in which the animal lives. This secre- tion is a constant accompaniment of the burrowing of the animal. It is the same secretion which in other worms leads to the formation of a tube. The animal did not however succeed in burying itself, and so it soon began its movements anew. Spontaneous progressive movements were executed only by that piece which contained the brain. Yet weak stimuli, such as the shaking of the aquarium in passing through the room, sufficed to bring about progressive movements in the posterior piece. None of the pieces would remain on their backs when turned over. When I attempted to keep a brainless piece forcibly in this position it made great attempts to return to the ventral position. In all the experiments which have been described the size of the piece is not imma- terial: the more segments it contains the more definite are its reactions. I tried finally to determine whether all the ganglia even the most posterior are able to bring about the same winding and bending which is characteristic of the injured worm. That is indeed the case. VI. EXPERIMENTS ON LUMBRICUS FCETIDUS I wished to determine whether angleworms are heliotropic or photokinetic, and whether the decapitated animals show 360 STUDIES IN GENEKAL PHYSIOLOGY the same relation to light in every particular as do the animals with the brain. If Lumbricus foetidus is introduced into a transparent closed vessel it is noticed first of all that the animals are strongly stereotropic. As soon as they reach the concavity of a corner or a groove they crawl along it and do not leave it. Secondly it can be shown that they are photokinetic. They come to rest in those regions which are more weakly illuminated than the surrounding areas. The direction of the rays of light is of little consequence. It seems also as if, when one or more animals have come to rest at any point, the others also come to rest at the same place more readily. This looks as though the animals were "social." This may be attributed to a chemotropic irritability. It is noteworthy that the less refrangible rays which pass through red glass are less active for photokinetic animals than the more strongly refrangible rays which go through blue glass. The angleworms come to rest sooner under red glass than under blue glass. (We can speak of a "prefer- ence" for red light for this case just as little as in the case of heliotropic animals.) Decapitated Lumbrici foetidi all show the same stereotropism as normal animals. When they reached the concave side of the corner of a vessel they did not readily leave the corner again. The decapitated animals also came to rest as did the normal animals in those regions where the light was least intense, while an increase in the intensity of .the light stimulated them to movement. It could also be shown that the light which passes through blue glass acted in this regard as light of a greater intensity than that which passes through red glass. In all these experiments the decapitated pieces crept about with their tail end forward as well as with the oral end forward. I noticed repeatedly a fact which shows that even the extreme caudal end of the angleworm is sensitive to light. For when the caudal end suddenly entered an illuminated BRAIN PHYSIOLOGY OF WORMS 361 area from one which was protected from the light it at once turned about. Strange to say the reaction-time to light is not markedly greater in decapitated angleworms than in normal animals. The experimental animals were contained in a dark box in which they could without being jarred be suddenly exposed to diffuse daylight. Three to eighteen seconds after the entrance of the light the decapitated angleworms first began to move. It took about the same time in normal worms. Lumbricus foetidus lives in decaying straw and manure, and it can readily be assumed that the chemical nature of certain substances contained in the straw and manure holds the animals — that in other words they are positively chemo- tropic to the substances. I could readily show that when one-half of the bottom of a box was covered with white moist filter-paper and the other half with a thin layer of decaying straw, the normal worms which were laid upon the filter- paper were soon all collected upon the manure. The poste- rior pieces of transversely severed worms behave in exactly the same way. When they were laid upon the filter-paper they were not directly attracted by the odorous substances contained in the manure. As soon, however, as they came in contact with the manure in their progressive movements they crept upon it, and once upon it they did not leave it again. In this way it soon happened that all the brainless worms were collected, without exception, upon the manure. When they were laid upon a heap of decaying straw, they soon buried themselves in it. That was not alone the effect of the light, for the reaction occurred also in the dark. VII. EXPERIMENTS ON LEECHES If a leech is cut in two transversely the two pieces show entirely different reactions. The wound soon heals and the pieces may live a year or more without however, as is well 362 STUDIES IN GENERAL PHYSIOLOGY known, any regeneration occurring. Both pieces can execute swimming motions: in the case of the posterior piece the oral end leads. Both pieces attach themselves by means of their suction disks. At times I also saw the posterior piece exe- cute lively progressive movements, without however being able to discover an external stimulus for this movement. Both pieces returned to the ventral position when they were laid upon their backs. For the rest however the two pieces behaved differently. While I frequently found the anterior piece attached to the vertical glass sides of the aquarium, I found the posterior piece attached to the bottom. I sus- pected that the attachment of the suction disk came to pass reflexively, in consequence of the pressure or the friction to which the skin of the suction disk is exposed when it is allowed to come in contact with the bottom. It was therefore to bo expected that it should be possible to compel the animal to attach itself to any desired point by pressing the suction disk against it. The experiment succeeds very readily with the posterior piece of a transversely divided animal, when it is gently pressed against the desired spot by means of a brush. The anterior piece however behaves entirely differ- ently when the same experiment is made with it. It turns itself in an apparently aggressive manner against the brush so that it is impossibe to press the suction disk against the wall. If the posterior piece is laid upon its back, it begins to execute swimming motions by means of which it brings itself back into the ventral position. The anterior piece returns to the ventral position by turning over. The tendency to execute swimming motions is much more marked in the posterior than in the anterior piece. If the leech is divided in such a way that the two pieces are still connected by a small piece of skin, they at times execute co-ordinated pro- gressive movements^ as la x riediander's experiment. More BKAIN PHYSIOLOGY OF WORMS 363 frequently however one observes that the posterior piece apparently begins its swimming motions spontaneously, and pushes before it the anterior piece which contracts and ruffles itself. Ill 1. Our observations therefore show that when a worm is cut through transversely that piece which contains the brain retains to a greater degree, generally speaking, the biological or psychic character of the species than the brainless piece, even when the latter far exceeds the anterior piece in mass. The difference which the oral and aboral pieces show in this regard is different in different species of worms. In Thysano- zoon this difference is marked, also in leeches and inCerebratu- lus, while in Lumbricus and especially in Planariatorva1 it is less. It is however questionable whether this difference is chiefly determined by the brain. For we do not know how far the specific irritability of the individual peripheral elements of the oral pole has to do with it. 2. The latter thought may go too far for many readers. But it seems to me that we are too much inclined to seek the " irritable structure" which determines the reaction of an animal exclusively in the central nervous system, while fre- quently a more careful analysis of the phenomena by no means compels such a conclusion. The Ascidians are the simplest reflex animals. The central nervous system is reduced to a single ganglion which receives sepsory fibers from the surface and sends motor fibers to the muscles. If the skin of the animal is touched, the muscles contract, and the oral and aboral openings of the animal close. Under these circumstances the stimulus passes from the touched spot to the ganglion and from here to the muscles. The ganglion can be readily extirpated in transparent forms such i In this form Bardeen has recently found that the so-called longitudinal nerves resemble more closely the oral ganglion in their histological structure. [1903] 364 STUDIES IN GENERAL PHYSIOLOGY as Ciona intestinalis. After the loss of the ganglion, how- ever, the animal reacts in the same way when touched as before. Only one condition is different. The intensity of the stimulus which is necessary to bring about a reaction in the brainless Ciona is much stronger than in the intact animal. This indicates that the mechanism is different in the two cases. In the normal animals the sensory nerves are stimulated and the stimulus passes through the ganglion to the muscles. In the brainless animal the muscles at the irritated point are possibly stimulated directly. The con- traction of these muscles is then the cause of the contraction of the neighboring muscles and so on.1 Under these circum- stances it is more than a mere possibility that in the normal Ciona also the characteristic reaction when touched is deter- mined not by the brain, but that the brain serves the purpose in this case of a better and more rapid conductor of the stimu- lus. The nature of the reaction might rather be chiefly determined by the arrangement of the muscles. The heliotropic phenomena of animals are identical in all respects with those of plants. The latter have no central nervous system and the remarkable nature of these reactions is therefore not necessarily determined in animals also by the specific characteristics of their reflex centers. It is much more probable that the nervous system plays in this only the role of a conductor of stimuli, while the actual char- acter of the process is determined by the following condi- tions: (1) The shape of the body and the topographical distribution of irritability corresponding with it. (2) The changes brought about by the light in the illuminated tissues : (3) The conduction of the stimulus to the contractile tissues. (4) The arrangement and structure of the latter. Examples of the same sort are the observations described i It is however possible that the stimulus is conducted to the individual muscle fibers through nerves. [1903] BRAIN PHYSIOLOGY OF WORMS 365 in this paper on stereotropism in brainless Planarians, and the reaction of brainless animals toward light. 3. Even though the brain, or more accurately that part of the body containing the brain, determines in the main the bio- logical or psychic reactions of worms which are typical for each species in the same way as in higher animals, there never- theless exists a specific difference between the brain of worms and that of many of the higher animals. The worms lack asso- ciative memory and consequently also consciousness, which is only a function of the former. By associative memory we understand that arrangement of the brain by virtue of which a stimulus brings about not only the effects corresponding with its nature and the specific structure of the irritable tissue, but also the stimulating effects of other causes which at a previous time once affected the organism at the same or almost the same time with the stimulus. No trace of such an associative effect of stimuli can be proved to exist in worms, and consequently also no trace of consciousness. One might go farther and speak of memory, when the effect of a stimulus depends upon previoSis^ffscte-of stimula- tion at all. Under these circumstances, for example, we would have to designate it as memory when a plant which was originally cultivated in the tropics does not withstand low temperatures as well as a plant of the same species culti- vated in the North. We would also have to call it memory when a Medusa of the temperate zone is moved to the regions of the midnight sun and here continues its depth migrations in a period corresponding with the changes of day and night in its home. No objection could be made to this, only I do not believe that this kind of memory can be looked upon as a lower form of associative memory. And I am convinced that it is something entirely different from associative mem- ory. When the plant which has been cultivated in the South does not do as well in the North as the same species of plant 366 STUDIES IN GENERAL PHYSIOLOGY which has been cultivated in the North it is to be attributed to a difference in the constitution of the tissues of the two plants, possibly to a difference in the amount of water contained in the two tissues. The periodic depth migra- tion of the Medusa are, as I believe, brought about through a periodic change resulting from internal causes in the amount of water contained in the animal, which, in the temperate zones, corresponds in its period with the change of day and night. (I suspect that the light brings about a change in the amount of water contained in the animal in one sense, while darkness brings it about in the oppo- site sense.) When the Medusa is then transferred to the North there is no occasion for a change in the period. In associative memory, on the other hand, we have to deal, it seems to me, with a definite mechanical arrangement which, from experiment and pathological experience, has to be sought in the brain and which is present only in certain animals, while it is missing in others. Correspondingly consciousness is present also only in certain animals, and in these only after a certain stage in embryonal development has been reached. To claim, as does one English author, that a "subconsciousness" exists in the egg, I consider just as wrong as though one would say that a subphonograph exists in a drop of water. The Darwinian habit of seeing transitions everywhere becomes erroneous when it attempts to take into consideration machines which yield qualified energy. And we have to do with such machines in the case of associative memory, as well as in many other physiological apparatus.1 4. If, therefore, a decided difference exists between many vertebrates and the worms (and other lower animals) so far as associated memory and consciousness are concerned, 1 In many respects my views coincide with those expressed by DRIESCH in his excellent booklet Die Biologic als selbststdndige Wissenschaft. BRAIN PHYSIOLOGY OF WORMS 367 only a quantitative difference exists so far as spontaneity is concerned ; when we designate as spontaneous those changes in an animal which are the result of internal, or more cor- rectly, which occur without demonstrable external stimuli. As a matter of fact, many changes brought about through external stimuli will seem spontaneous to us because the external stimulus escapes our observation. We have to dis- tinguish between conscious spontaneous changes (the true will-action in which the idea of the coming change precedes the latter) and simple spontaneous changes in which internal causes determine the latter without processes of conscious- ness being present. In the case of worms, of course, we can speak only of the latter forms of spontaneity. In Thysano- zoon this spontaneity seems to be exclusively a function of the brain. In Planaria torva, on the other hand, this is not so distinctly the case. When compared with the number of reactions to external stimuli the number of the spontaneous movements of worms is small. Only where associative memory is present do the spontaneous changes step into the foreground numerically. 5. Whether the sensations of pleasure and pain are pos- sible without consciousness cannot be decided absolutely. If it is permissible to consider the reactions of a frog when its skin is touched with acid, or the bending of a worm when one steps on it, as an expression of a sensation of pain,1 our experiments show that all pieces of a worm are capable of the sensation of pain. It is worthy of notice, how- ever, that the reactions pointing to sensations of pain are weaker in Planarians than in Annelids, or may be lacking altogether. 6. One might be led to believe that the reflex motions in higher animals depend to a higher degree upon the i W. W. NORMAN has since shown that this is not permissible. The problem is more fully discussed in my book on the Comparative Physiology of the Brain. [1903] 368 STUDIES IN GENERAL PHYSIOLOGY structure of the central nervous system than in lower ani- mals. The fact that Planarians continue to react upon light if their brain is removed, or that Ciona continues to show its characteristic reaction after the loss of its ganglion seems to suggest such an idea. But it would be false, never- theless. The contraction of the iris of the eye, if stronger light falls into the latter, is a typical reflex action, called forth through the effect of the light upon the retina. If, however, the reflex center is destroyed or the iris cut out, an increase of the intensity of the light which strikes the iris continues to cause a contraction of the iris. This fact is known for frogs and eels, and I have observed it in sharks. It is probably true for mammalians also. The reflex act therefore may serve here, as in Planarians, for the greater conduction of stimuli. When a dog, whose spinal cord has been cut, is lifted so that its body hangs down vertically, a peculiar fact can be observed, as Goltz has shown. The legs are thrown into pendulum-like motions resembling walking motions. These motions are produced by the passive stretching to which the skin on the ventral side of the hip- joints is subjected through the weight of the legs. These pendulum-like motions are comparable to the reflectory contraction of the longitudinal muscles of the earthworm when its skin is stretched. This reflex would suffice to call forth co-ordinated walking motions in the dog whose spinal cord is severed, if such a dog were only able to stand on its legs. The walking motions of the anterior legs would produce periodically the stretching of the skin which is required for the locomotion of the posterior legs. The difference in the behavior of a dog with severed spinal cord and an earthworm with severed ganglionic chain in regard to co-ordinated locomotion is therefore less deter- mined by the differences in the function of their central nervous system than by differences in the structure of their BRAIN PHYSIOLOGY OF WORMS 369 peripheral organs of locomotion. If the dog possessed, in the place of its high legs, short bristles like the earthworm, the dog with severed spinal cord would continue to make co-ordinated progressive motions as the earthworm in Friedlander's experiment. XV THE PHYSIOLOGICAL EFFECTS OF LACK OF OXYGEN l I. INTRODUCTION MORE than a century ago Spallanzani published his ob- servations on the effect of stagnant air upon animals and plants. Spallanzani introduced organisms into hermetically sealed vessels of various sizes (which were, however, filled with air), and found that the smaller the vessel, the earlier did all life cease to exist in it. He also observed that differ- ent organisms show an unequal resistance to lack of air. As the most remarkable case he cites Anguillula aceti: "They live and multiply prodigiously where the volume of air does not exceed three inches; and die in several days only, when confined in a tube where the vacuum is less than an inch."2 The further experiments on the same subject have fully confirmed the observations of Spallanzani. Bunge, in his well-known treatise on the respiration of mud-dwelling organisms, concludes that apparently "all transitional stages exist in the animal kingdom from the anaerobic unicellular organisms up to the most highly organized animals with a most energetic demand for oxygen." 3 A series of brilliant observations have served to elucidate the chemical side of these phenomena. It is an established fact that carbon dioxide can be produced in an organism without the presence of oxygen. Hermann has shown that the excised muscle of the frog is able to do work and to 1 Pfliloers Archiv, Vol. LXII (1895), p. 249. 2 SPALLANZANI, Tracts of the Natural History of Animals and Vegetables. 3 Zeifchrift filr physiologische Chemie, Vol. XII 370 PHYSIOLOGICAL EFFECTS OF LACK OF OXYGEN * 371 — r~ produce carbon dioxide without containing oxygen that can be exhausted by an air pump.1 Pfluger2 and Aubert3 have shown the same to be true for the living frog. No one doubts that we are dealing with phenomena of splitting in these cases, which are at the same time the source of energy and the source for the motions, and the other physi- ological functions that go on in the vacuum. The differences observed by Spallanzani and Bunge in the length of time that animals live without oxygen may therefore be explained by assuming that different forms of animals contain different amounts of hydrolyzable substances. As soon as this material is used up "the clock stands still." Oxygen plays the r6le of replacing the substances capable of undergoing splitting which have been used up. By calculating the energy obtainable by the hydrolysis and by the oxidation of carbohydrates Bunge has rendered it probable that in the higher animals the production of powerful work is not caused by hydrolysis alone, but by hydrolysis and oxidation. According to this, lack of oxygen could at once reduce the capacity for work of an animal by limiting it to the energy which can be obtained from hydrolysis. Hoppe-Seyler was the first to suggest a chemical theory for the processes of oxidation which occur in the living organ- ism.4 He believes that, as in the process of putrefaction, reducing substances (such as hydrogen in the nascent state) are formed in all living cells through hydrolysis, and that these substances, when atmospheric oxygen is present, tear apart the oxygen molecule. The free oxygen atom is then in the condition in which it is able -to bring about the oxida- 1 Untersuchungen iiber den Stoffwechsel der Muskeln (Berlin, 1867). 2 Pfliigers Archiv, Vol. X. 3 Ibid., Vol. XXVI. * At the time I wrote this paper I was not familiar with the papers of Traube on the subject, which seem to give a more adequate presentation of the ti feet than Hoppe-Seyler's hypothesis. [1903] 372 STUDIES IN GENERAL PHYSIOLOGY tions which are characteristic of living matter.1 If. how- ever, oxygen is absent, the chemical changes will be of an entirely different character. Compounds may be formed which are poisonous for the organism. Araki, for example, has shown that when oxygen is lacking considerable amounts of lactic acid and sugar appear in the urine. If the oxygen supply is normal, these compounds may also be formed from glycogen as intermediate products, but they are soon oxidized further. To the immediate consequences of lack of oxygen are therefore added those which are determined through the presence of lactic acid in the body — for ex- ample, a diminution in the alkalinity of the blood. The cells of the kidney are also altered, as evidenced by the albumi- nuria which results when oxygen is lacking.2 While there exists a comparatively large number of in- vestigations concerning the chemical side of the effects of lack of oxygen (of which we have mentioned only a few), we have very few biological observations on the same sub- ject. Even careful search of the literature discloses little more than the fact that all animal life-phenomena cease sooner or later in the absence of oxygen, that in higher animals the phenomena of dyspnoea precede death, and that, as Ktihne showed3 the protoplasm becomes vacuolated and opaque and disintegrates. The observations made upon mountain disease may also be mentioned under this head- ing. It did not seem possible to me that these facts ex- hausted all the biological effects of lack of oxygen. The fundamental importance of oxygen for all life-phenomena rendered it probable, a priori, that an accurate investigation would yield a series of qualitative and quantitative changes in life-phenomena. Such an investigation is, of course, 1 Physiologische Chemie, Vol. I, p. 126. 2 Zeitschrift filr physiologische Chemie, Vol. XIX. 3 Untersuchungen iiber das Protoplasma, 1864. PHYSIOLOGICAL EFFECTS OF LACK OF OXYGEN 373 rendered difficult by the fact that many animals soon die without oxygen, and I believe that this fact explains why this field has not thus far been the subject of more study. Still the duration of life is in most cases sufficiently long to demonstrate a series of such changes. Copepods are, for example, exceedingly sensitive to lack of oxygen ; I know no other cold-blooded animals that die so rapidly without oxygen. Yet it is possible, as we shall see, to show definitely even in these animals that lack of oxygen affects their helio- tropic sense in a most remarkable way; when deprived of oxygen negatively heliotropic Copepods become positively heliotropic. Another consideration shows the importance of such inves- tigation. Physiological chemistry alone may suffice to dis- close the general sources of energy in animals. But the question as to how chemical energy is converted into the physiological activities of muscles, glands, etc., can of course not be answered by purely chemical researches. Molecular physiology must here bridge the chasm between the chemi- cal changes and the outwardly manifested physiological activities of the organs. A complete understanding of the energetics of animals is not possible so long as we have no conception of the molecular changes which are brought about through processes of oxidation. It therefore seemed of importance to see whether such changes manifest them- selves when oxygen is taken away. In this way arose the experiments detailed here on cleavage without oxygen, which I began three years ago, and which I discussed in a short note which appeared in P fingers Archiv two years ago.1 I directed my attention to processes of segmentation because I considered these phenomena especially favorable for obtain- ing facts for a molecular physiology. When we find that a physiological function is impossible i Pfliigers Archiv, Vol. LV, p. 530. 374 STUDIES IN GENERAL PHYSIOLOGY without oxygen, we are inclined to imagine that the given organ or organism lacks the necessary energy for performing this function. If we remember, however, that the conver- sion of chemical into the physiological function depends upon definite molecular conditions in the cells (state of mat- ter, osmotic pressure, surface tension, phenomena of spread- ing, etc.), another explanation is possible, a priori: the cells are still able to produce the energy necessary for the physio- logical functions of the organ, but lack of oxygen led to molecular changes in the cells which prevented the conver- sion of the chemical energy into mechanical or other forms of energy. So far as I know, there are as yet no facts at hand to support such a view. Yet the following experi- ments, I believe, have led to positive results in this direction ; for we can show that the eggs of Ctenolabrus and sea-urchin cannot segment without oxygen, and that, moreover, the already formed cleavage-cells of these eggs, especially those of Ctenolabrus, undergo certain structural changes when deprived of oxygen, which cause the cells to fuse together. It is possible that in this case we deal with a liquefaction of the membrane or the specific surface film of the cleavage-cells, and furthermore that the impossibility of the formation of a membrane in the absence of oxgen is why no cleavage occurs under these conditions. But no matter what one may assume regarding the formation of a membrane, it is clear that when separate cells fuse in the absence of oxygen, it is not to be expected that the unsegmented egg will be able to divide under these conditions. But those cells which have fused are by no means dead. When they are again supplied with air, segmentation sets in anew. We therefore see that the structural changes resulting from the absence of oxygen suffice to explain the failure of segmentation, and that it is not necessary in this case to attribute the latter to a failure of the source of chemical energy. This conclusion is further PHYSIOLOGICAL EFFECTS OF LACK OF OXYGEN 375 supported by the fact that the eggs of Funduhis in which such structural changes do not occur in the absence of oxygen even after twenty-four or more hours, continue to seg- ment for more than twelve hours in the absence of oxygen. What we are going to show for cleavage holds also for other life-phenomena, for example, the activity of the heart. We find that the heart of the Ctenolabrus embryo comes to a standstill so quickly after the oxgen has been withdrawn that it is impossible to think that the source of the energy for the beating of the heart has given out, while the heart of the Fundulus embryo continues to beat for many hours under similar conditions. It is possible that in this case also structural changes occur which are similar to those which we are able to observe directly in the cleavage-cells. These changes render impossible the transformation of chemical energy into mechanical energy in the contraction of the heart of the Ctenolabrus embryo. Such molecular changes as manifest themselves by struc- tural changes can be brought about just as well through processes of reduction due to the lack of oxygen as through the injurious compounds which may be formed in the absence of oxygen. The facts which we obtained by the biological study of the effects of lack of oxygen may also again be of importance to the physiological chemist. We shall meet with some facts in this paper which will serve to illustrate the view of Bunge that in all probability the greater part of the energy necessary for considerable work of the muscles is furnished through processes of oxidation (and not through processes of hydrolysis). The frequency of the heart-beat of the Fundu- lus embryo decreases steadily during the period during which oxygen is withdrawn, until it reaches the minimum (of about twenty beats per minute), when all the oxygen has dis- appeared. But the heart can beat for ten hours at this rate 376 STUDIES IN GENERAL PHYSIOLOGY at a temperature of 22° C., in the entire absence of oxygen. This looks, indeed, as if the energy for a certain small num- ber of beats could be furnished through hydrolysis alone, but that for a larger number of heart-beats (about 120) the energy is obtained, in the presence of oxygen, through oxidation. II. ON THE METHOD OF THE EXPERIMENTS In the following experiments on the effects of lack of oxygen, the oxygen was always displaced by hydrogen ; the latter was freshly prepared from zinc and sulphuric acid for every experiment. The gas was washed thoroughly through two bottles of potassium hydroxide, one of potassium per- manganate, and one of water. The hydrogen obtained in this way was entirely odorless. Before the experiment was begun, all air was driven out of the apparatus. The animals ex- perimented upon were kept in an Engelmann chamber which permitted of direct microscopic observation. In this method one fact is to be emphasized, which I have not seen mentioned elsewhere. When the living tissue upon which we wish to test the effects of lack of oxygen is placed in the Engelmann chamber, our judgment of the results is subject to the criticism that we do not know the exact moment at which the oxygen is all driven out of the liv- ing specimen. The resistances to the diffusion of oxygen out of the protoplasm are very great, and it may take con- siderable time before all the oxygen is removed. As long as we find that the phenomena whose dependence upon oxygen we wish to study cease immediately after the hydrogen is passed through the gas-chamber, this diffi- culty is less marked; for even though in this case not all of the oxygen had been driven out of the organism, it would show only that the particular function which is being studied already ceases at a diminution of the oxygen supply, and still more so in the total absence of oxygen. It is PHYSIOLOGICAL EFFECTS OF LACK OF OXYGEN 377 different, however, when the function which we are study- ing does not cease immediately. We are then unable to say whether the transitory persistence of the function shows that not all of the oxygen has been driven out, or that the given function is not directly dependent upon oxygen. Things of this sort confront us when we attempt to decide whether cleavage is possible without oxygen. We find that when sea-urchin eggs are placed in an Engelmann chamber immediately after fertilization, and we begin to pass hydrogen through it, the eggs not only cleave into two, but often even into four, cells ; but this cleavage occurs within the first fifty or eighty minutes after fertilization, and it might be thought that it takes a longer time than this to drive out all the oxygen. To be certain on this point in such cases, I made use of the following procedure: I in- troduced the eggs in which I wished to study the depend- ence of cleavage upon oxygen into an Engelmann chamber which was kept on ice. Hydrogen was then passed through the apparatus. The low temperature inhibits cleavage. In order to discover when I could take the eggs off the ice, and know that the objection could no longer be raised that the eggs still contain oxygen, I introduced a second gas-chamber into the circuit. This contained eggs of the same culture, and through it I passed the same current of gas as that which went through the first. The second, control, chamber was not put on ice. As long as a trace of cleavage continued in this control chamber, there was reason to suspect that not all the oxygen was driven out. As soon, however, as cleavage ceased, it seemed reasonable to assume that, even though not all the oxygen had been driven out of this chamber, the por- tion which remained behind was no longer sufficient to start cleavage. It must be remembered, however, that the eggs kept on ice had not lost as much oxygen as the control eggs during this time, for oxidation did not occur as rapidly in the 378 STUDIES IN GENERAL PHYSIOLOGY former as in the control eggs. It was therefore necessary, after cleavage had ceased in the control eggs, to pour the current of hydrogen through the Engelmann chamber for some time before the eggs to be experimented upon were removed from the ice, and the real experiment was begun. The objection might be raised that the prevention of cleavage through the ice injured the eggs. I guarded against this objection by the following control experiments: First of all the eggs were again exposed to the air after the completion of the experiment, and their cleavage observed. If this took place in a normal way, its absence in the lack of oxygen could not have been the effect of the cooling. Secondly, another portion of the eggs of the same cul- ture were put upon the ice at the same time and for the same length of time as the eggs used in the experiment, only they remained exposed to the air. I will state at once that these eggs always segmented when brought back to room temperature. During the entire time of the experi- ment hydrogen passed uninterruptedly through the Engel- mann chamber, not only to guard against possible leaks in the apparatus, but also to remove the carbon dioxide formed. The latter is absolutely necessary. III. RESEGMENTATION OF THE CTENOLABRUS EGG WITHOUT OXYGEN The older experiments of Spallanzani, Dutrochet, Saus- sure, and Schwann had already established the fact that in permanent lack of oxygen the development of plants and of animal eggs is impossible. Paul Bert added to these ob- servations the fact that when the air contains only 3.4 per cent, of oxygen certain plants cease to germinate. In these experiments, however, the question as to whether cell- division is at all possible without oxygen was not touched upon. Three years ago I began experiments on the eggs of PHYSIOLOGICAL EFFECTS OF LACK OF OXYGEN 379 Fundulus which showed that the egg not only segments when oxygen is removed by pyrogallol, but even continues to develop for about sixteen hours. Demoor, who was un- familiar with my experiments, began experiments on Trades- cantia cells in which he found that a cell-division which had already started at the time that the oxygen was being removed continues to the completion of nuclear division, but that the subsequent cell-division does not occur. He concludes from this, first, that cell-division is impossible without oxygen, and especially that without oxygen the cell- membrane cannot be formed; and, secondly, that the nucleus may divide without oxygen, that it is anaerobic.1 I have in a previous paper pointed out the incorrectness of the second conclusion. My own experiments, which I will give here, were made on fish eggs (Ctenolabrus and Fundulus) and sea-urchin The egg of Ctenolabrus, a marine Teleost, is perfectly transparent and free from pigment, and the changes which are described in the following pages can be studied with great accuracy under the microscope. The eggs which were used in the following experiments were always fertilized artificially in the laboratory. If the freshly fertilized eggs of Ctenolabrus are intro- duced into an Engelmann chamber, and care is taken that all the air is driven out of the apparatus before the experiment is begun, and the stream of gas is maintained, the eggs cleave, without exception, into two cells, and in most cases even into four cells. Occasionally they even go into the eight-cell stage. If the eggs are introduced into the gas- chamber not immediately after fertilization, but in one of the later stages of cleavage, two or three divisions of all the cells still occur. i Archive de biologic, Vol. XIII. 380 STUDIES IN GENERAL PHYSIOLOGY Must we now assume that Ctenolabrus is able to divide two or three times without oxygen? The first cleavage of the Ctenolabrus egg occurs in from fifty to seventy minutes after fertilization, according to temperature; the second, about fifteen to thirty minutes later. It is entirely possible that even in a strong current of hydrogen all of the oxygen is not driven out of the eggs in so short a time. In order to settle this point, I made a long series of experiments in the manner described above, in that I introduced the eggs immediately after fertilization into two gas chambers, one of which was kept on ice, while the other was exposed to room temperature, and passed the same stream of hydrogen through both. I will describe a few of these experiments here. In order to be brief, I will call the eggs upon the ice the experimental eggs, the .others the control eggs. In one experiment the control eggs divided into two cells fifty minutes after fertilization, when the experimental eggs were removed from the ice, while the stream of hydrogen was kept up uninterruptedly. In thirty minutes the first cleavage occurred in the experimental eggs. At the same time the control eggs went into the four-cell stage, and twenty-five minutes later the experimental eggs also went into the four-cell stage. Cleavage then ceased in both chambers. Even though hydrogen had been conducted through the chamber containing the experimental eggs, which had been kept on the ice for a long time before the beginning of cleavage, and the oxygen had probably been driven out more thoroughly than in the control eggs, cleavage nevertheless occurred in the same way in both. There was only one difference ; all the control eggs reached the four-cell stage, while about 25 per cent, of the ex- perimental eggs remained in the two-cell stage. In another experiment the experimental eggs remained for one hour and forty minutes on the ice. During the first PHYSIOLOGICAL EFFECTS OF LACK OF OXYGEN 381 hour all of the control eggs had developed into the two-cell stage, but did not go any farther. When exposed to room temperature, an incomplete division occurred in a small number of the experimental eggs after thirty minutes. Only a suggestion of a membrane dividing the two cells was formed ; the peripheral cell-membranes were not formed. The process then came to a standstill. That this result was attributable to the lack of oxygen, and not to the prolonged stay in the cold, was shown by the fact that when, after some time, the gas-chamber was opened, vigorous cleavage set in in all the eggs after thirty minutes. The experimental eggs in a third experiment remained on the ice for two hours. The control eggs reached the four- cell stage within the first eighty minutes. Cleavage then ceased. When the experimental eggs were taken from the ice, not a suggestion of cleavage set in during the following eighty minutes. Air was then admitted. All the eggs be- gan to divide in thirty minutes. I obtained the same result in more than ten further ex- periments. With the exception of the fact that in an occa- sional egg among hundreds an intimation of a dividing membrane was visible, no cleavage whatsoever occurred when a vigorous stream of hydrogen was led for two hours or longer before the beginning of the experiment through the gas-chamber which contained the experimental eggs and was kept on the ice. Yet the same eggs all divided within half an hour when later exposed to the air. It might be thought that lack of oxygen only markedly retards cleavage, but does not bring it to a complete stand- still. Yet it did not matter how long one waited — cleavage never occurred in the gas-chamber in the case of lack of oxygen, when all the oxygen had been driven out. Furthermore, I ascertained that when any segmentation whatsoever occurred in a weak stream of hydrogen, it always 382 STUDIES IN GENEKAL PHYSIOLOGY occurred at the same time as (or earlier than) in the eggs from the same culture kept in oxygen. If the minimum amount of oxygen necessary for cleavage is present, the velocity of the cleavage is a function of the temperature and not of the amount of oxygen present. It is important to emphasize the fact that lack of oxygen at room temperature does not retard cleavage, as does a reduction in the tempera- ture. Finally, I convinced myself of the fact, through a special series of experiments, that prolonged exposure to cold does not diminish the power of the egg to divide. I allowed a weak current of hydrogen to pass through a gas-chamber which remained on ice for four hours. When I then exposed the eggs to room temperature and continued to pass the same weak current of gas through the chamber, all the eggs divided. The majority reached the four-cell stage, and a few even the eight-cell stage. Cleavage then ceased. If not all the oxygen is driven -out, cleavage proportionate to the amount of oxygen present still occurs in spite of the prolonged cooling. We are, therefore, justified in conclud- ing that when all the oxygen which it is possible to 'remove from the Ctenolabrus egg is driven out, no complete cell- division can occur. The question now arises in how far a division of the nucleus is possible in such an egg. At the surface of the Ctenolabrus egg a series of visible changes occurs before the first cleavage. In the center of the nucleus several droplets of a strongly refractive substance collect, which increase in number and size, then coalesce, to resolve again into a large number of minute droplets just before cell-division. These droplets probably play a role, as we shall see later, in the process of cell-division. It is possible, but not proved, that their formation is a function of karyokinesis. These changes in the strongly refractive substance also occur PHYSIOLOGICAL EFFECTS OF LACK OF OXYGEN 383 when all the exhaustible oxygen has been removed so that cell-division is no longer possible. The energy necessary for these changes must probably therefore be obtained from processes of hydrolysis. One might be tempted to believe that the nucleus could continue to divide without oxygen, while the cell remains undivided — a phenomenon I discovered in sea-urchin eggs, when they are brought into sea-water of a certain concentra- tion. In such eggs the number of nuclei steadily increases, but no cell-division occurs. But such phenomena certainly do not take place in the absence of oxygen. Eggs were freed from oxygen by passing hydrogen over them for two hours while on ice. They were then exposed for one hour to room temperature, while the flow of hydrogen was not interrupted. No cleavage had occurred. The eggs were then killed and sectioned. It was impossible to find more than one nucleus in these eggs; this was, however, in a number of instances undergoing mitosis. The experiment was repeated with the same result. One mitotic division may therefore occur without oxygen, but no more. If eggs which have been freed from oxygen for a suffi- ciently long time while on ice, and which have shown no evidence of cell-division when exposed to hydrogen for an hour at room temperature, are exposed to the air, they all divide in the course of thirty to fifty minutes. But they do not then first divide into two cells and later into four, but immediately into four — occasionally into three or five cells. This also occurs when a strong stream of hydrogen is sent through the gas-chamber for three and a half hours at a low temperature before the experiment is begun — under condi- tions therefore when, in all probability, all of the free oxygen has been removed from the eggs. Two divisions of the nucleus therefore always occur — one in the hydrogen, and one after the admission of air — before the first cell-division 384 STUDIES IN GENERAL PHYSIOLOGY is inaugurated. This marked retardation of cell-division has its basis in peculiar molecular changes, which we will discuss in detail in the following sections of this paper. IV. THE FUSION OF CLEAVAGE -CELLS THROUGH LACK OF OXYGEN The fact that the egg of Ctenolabrus is not able to seg- ment without oxygen may be due to one of two causes: first, processes of oxidation might be the only source of energy for segmentation; second, it might be possible that, even though enough chemical energy for segmentation can be obtained from hydrolysis, yet this chemical energy cannot be connected with the chemical energy necessary for cleavage because of the structural changes brought about by the lack of oxygen. Demoor concludes from his experiments on Tradescantia that no cell-wall is formed without oxygen, and that in consequence no cell-division occurs without oxygen. Demoor brings no positive proofs for his view. In the case of the Ctenolabrus egg, however, we can show that structural changes occur in cleavage-cells, in con- sequence of which these cells fuse together. It is conceiv- able that the same structural changes must also hinder the segmentation of the freshly fertilized egg. The sketches, Figs. 103-8 were made with the camera lucida and were all taken from the same egg. The egg was fertilized at 10 A. M., and immediately thereafter introduced into a gas-chamber and kept in a current of hydrogen. Cleavage took place in the normal way, and since the current of hydrogen was not very strong, even the eight-cell stage was reached (Figs. 103-5). A series of degenerative changes then set in. At first a gathering of the strongly refractive droplets, which we have described already, was formed in the two main furrows (Figs. 104 and 105) and some furrows began to be- come indistinct. Fifteen minutes later the greater portion PHYSIOLOGICAL EFFECTS OF LACK OF OXYGEN 385 FIG. 103 FIG. 104 of the peripheral cell-limits had already become invisible (Fig. 106), and after another fifteen minutes nothing could be recognized of the entire blastoderm except a collection of droplets which had fused into larger drops (Fig. 107). In the next two hours the latter only became more spherical, but otherwise under- went no change (Fig. 108). The germdisk was optically still less visible than in the unfertilized egg. It therefore required only thirty-five min- utes after cleavage came to a stop, for the complete liquefaction of the cleavage-cells of an eight- celled blastoderm. It is scarcely necessary to mention that the same process in various experiments took a little more or a little less time. What we observe here is found in every such experiment upon Ctenolabrus eggs, the only difference being in the form and the arrangement of the droplets of the strongly refractive material, which at times may form a more perfect cast of the old lines of cleavage than in the experiment described. Even when the oxygen is driven out so slowly that the egg has time to reach the sixteen- or the thirty-two-cell stage in the stream of hydrogen, the same series of degenerative changes occurs as soon as cleavage has come to a standstill. FIG. 105 FIG. 106 386 STUDIES IN GENERAL PHYSIOLOGY The question now arises whether this liquefaction or fusion of the cleavage-cells occurs equally well and with the same rapidity in every stage of development. As soon as the eggs have reached the sixty-four- or the one-hundred-and-twenty- eight-cell stage, the behavior toward lack of oxygen is some- what different. While in an egg a** • r* which is in the eight-cell stage the • •" :../•'•}// * j-.r cells fuse in about one hour in the absence of oxygen, a liquefaction of A •"••' the cleavage-cells also occurs in the y eggs i*1 the sixty-four- and one- FIG. 107 hundred-and-twenty-eight-cell stage, but only at the periphery of the blastoderm, and even here more slowly than in the cells in the earlier stages of segmentation. The droplets of the refractive substance appear in the furrows, but they are] smaller than in the eggs in an earlier stage of development, and it is for this ,-,,-, ., , FIG. 108 reason perhaps that large oil drops are formed less easily. Figs. 109-111 illustrate the process of liquefaction in such an egg. The egg was put into the gas-chamber at 2:25 o'clock while in the sixty-four-cell stage. At this time its shape was sketched with the camera lucida (Fig. 109). The outlines of the cells within the blastoderm are not shown. The segmentation at first continued. FIG. 109 At 4 o'clock liquefaction was very distinct at the periphery. It occurred in this way that in individual cells at the periphery of the blastoderm the outline at first becomes invisible, after which the entire cell gradually disappears. Through this disappearance of the PHYSIOLOGICAL EFFECTS OF LACK OF OXYGEN 387 cells at the periphery the blastoderm becomes smaller (Fig. 110). At 6:35 o'clock the liquefaction of the cells at the periphery had progressed much farther (Fig. 111). The diameter of the blastoderm was only a little more than three- fifths of the diameter the egg possessed four hours earlier when it was in the sixty-four-cell stage. Around the blastoderm lay granular masses, which were in all probability the remains of the lique- fied cells. Soon thereafter a change (shrinking?) occurred in the yolk, which served to close the experiment. The disappearance of the cleavage- cells occurs more slowly therefore in the later stages of development than in the earlier stages of development. It may be of interest to raise the question: In what do these peculiar structural changes consist which lead to the fusion of the cleavage-cells in the absence of oxygen ? If we wish to answer this question, we must acquaint ourselves more fully with the history and the significance of those peculiar refractive substances which appear in droplets. Soon after fertilization, before the union of the pro-nuclei, one observes in the center and upon the surface of the germ the appearance of several strongly refractive droplets. These undergo, as has already been said, a series of changes, of which the most remarkable is this, that shortly before the first cleavage a single system FIG- m of radiations coming from a common center is formed, which looks very much like the radiations about a centrosome. These radiations might be a process of emulsion, for the radii break up very rapidly into small droplets which are strongly 388 STUDIES IN GENERAL PHYSIOLOGY refractive, and soon thereafter the center also breaks up into such droplets. As soon as the blastoderm divides, these drop- lets are found distributed over the entire surface, collected especially thickly along the furrow between the two cells. Before the next cleavage occurs, these droplets again arrange themselves in a line which corresponds to the next furrow (Fig. 103, a). In the process of segmentation part of these droplets disappear. This fact, in conjunction with a series of other facts, with which we shall become acquainted in the next section of this paper, leads me to suspect that this strongly refractive substance serves for the formation of the membrane of the cleavage-cells of the Ctenolabrus egg.1 That such a membrane, or at least a solid surface layer, covers the cleavage-cells of the Ctenolabrus egg immediately after a cleavage is completed I have observed directly; for folds are often formed on the surface of the cells, which are especially distinct immediately after a cell-division in the furrow (Figs. 104 and 105, /). On the assumption of the existence of a membrane our observa- tions can be expressed in a simple way. In the absence of oxygen the membranes of the cleavage-cells are liquefied and this brings about the fusion of the latter. The material of which the cell- walls were formed flows together in droplets which coalesce into larger drops in the center of the germ- disk. This liquefaction of the material of which the mem- brane is formed also renders cell-division impossible in the case of lack of oxygen. The assumption of the existence of a membrane, or at least of a specific surface film, in animal cleavage-cells also brings the mechanics of cell-division in animals and in plants into better harmony. The fact that in the process of cleavage the droplets always collect along the plane in which cleavage is to occur later is, as I would suggest in passing, a corroboration of 1 It is now generally assumed that the surface film of cells is formed by lipoids. The optical appearance of the droplets mentioned in the text is indeed that of a fatty substance. [1903] PHYSIOLOGICAL EFFECTS OF LACK OF OXYGEN 389 the view I expressed a short time ago on the mechanics of cell-division.1 I imagine that as soon as the nucleus divides, vortex motions take place about each of the two daughter- nuclei, which leads to a tearing apart of the cell-contents ; in other words, to cell-division.2 If this assumption is correct, movable particles must collect where the two vortex motions meet — that is, along the lines in which cleavage is to occur later. We indeed find this to be the case in the Ctenolabrus egg, and also in such eggs as carry pigment at their surfaces. These vortex motions carry the droplets to the place where the next cleavage is to occur, and where they are necessary for the formation of a membrane — a remarkable example of that "purposeful" interaction between mechanical conditions which we meet so often in processes of development/ We see, therefore, that molecular changes — apparently a liquefaction and an emulsion of the membrane or the surface film of the cleavage-cells — occur in the case of lack of oxygen which gives an adequate explanation of the fact that no cleavage occurs in Ctenolabrus eggs without oxygen. But the fact that nuclear division also soon comes to a standstill indicates that changes corresponding to those in the mem- brane must also occur inside the cells. V. EEVERSAL OF THE EFFECT OF LACK OF OXYGEN UPON ADMISSION OF AIR When an egg whose entire blastoderm has become invisible in hydrogen is again exposed to air, the changes which ensue differ according to the length of time during which the eggs have been exposed to the current of hydrogen. If the egg remains too long without oxygen at room tempera- 1 Archiv fur Entwicklungsmechanik, Vol. I. 2 1 find in this case, as in that of all hypotheses, that they do not gain in attract- iveness with growing age. Conklin, though, has accepted the hypothesis of vortex motions and Butschli has justly claimed priority for it. [1903] 390 STUDIES IN GENERAL PHYSIOLOGY FIG. 112 ture, it dies. If the experiment is interrupted early — that is, when the cell-walls have just begun to become indistinct — all, or at least a part, of the cell-membranes again become visible upon admission of air. Under these circumstances, however, every cell usually divides, not into two, but into four cells which cor- responds with what has been said before. When we wait a little longer before admitting air, a circular blastoderm is at first formed in which no trace of cleav- age is visible. The blastoderm then sud- denly breaks up into a large number of cells at once, but curiously enough this cleavage is confined, in most cases, to the periphery of the blastoderm. In this case also the refractive sub- stance which has been described plays a peculiar r6le. Figs. 112- 17 represent the various stages of the renewed cleavage of the same blastoderm in which we studied the disappearance of the lines of cleav- age in hydrogen (Figs. 103-8). Fig. 108 shows the condition of the blasto- derm in hydrogen at 2 : 10 o'clock. Only four large drops of the refractive sub- stance, surrounded by droplets of smaller size, permit one to recognize the place At 2:18 pure oxygen was sent through the gas-chamber. At first the smaller droplets separated from the surface of the large droplets and moved toward what had been the periphery of the blastoderm. (Previously, FIG. 113 FIG. 114 of the blastoderm. PHYSIOLOGICAL EFFECTS OF LACK OF OXYGEN 391 during the liquefaction of the cell-walls, they had moved toward the center of the blastoderm.) At 2:35 a unicellular spherical blastoderm became visible (Fig. 112). The larger central drops gradually broke up into minute droplets, which often arranged themselves in a ring about the periphery of the blasto- derm (Figs. 113, 114). Then forty- five minutes after the admission of oxygen, cleavage began. It occur- red only at such places where the FIG. 115 tiny droplets have collected, namely, at the periphery. The periphery broke up into about eighteen cells at once (Fig. 115). These cells about corre- FIG. 116 spond in size with those e found normally in the thirty-two to sixty-four- cell stages. The center of the blastoderm did not segment, with the excep- tion of one spot where the outlines of two cells be- came visible. I had noticed previously a small collection of the refractive droplets at this point. Still another relation between the distribution of the droplets and segmentation is noticeable. If the reader will FIG. 117 392 STUDIES IN GENERAL PHYSIOLOGY compare Figs. 114 and 115 he will notice that more cleavage- cells are formed in the four sectors which lie between the four large drops, and which contain the larger number of small droplets at the periphery, than in the four sectors which contain the large drops and a smaller number of small drop- lets. This relation may be accidental, but I may be allowed to state that when Fig. 114 was formed I expected from my earlier observations precisely that type of cleavage which is shown in Fig. 115. The cells formed at the periphery continued to divide, while the center remained undivided. The two cells which had first been perceived there disappeared again. The four central drops became steadily smaller, and one of them broke up entirely into small droplets, as if a slow emulsion took place (Fig. 117, e). In this way the blastoderm changed, within fifty minutes, from the condition shown in Fig. 115 into that shown in Figs. 116 and 117. Development then ceased. The long exposure to lack of oxygen at a relatively high temperature led to an early death of the germ. The phenomena shown in Figs. 112-17 are typical. The exceptions which one encounters are connected with differ- ences in the behavior of the small droplets of the strongly refractive substance. In this connection I must mention the fact that a small percentage of the eggs also showed segmen- tation in the middle of the blastoderm. In these cases, however, I usually (if not always) found that not only the periphery, but the entire blastoderm, was studded with very minute droplets immediately before cleavage. The large cen- tral drops were then found to dissolve rapidly (emulsion ?). I also found in rare cases, though, that only one sector of the blas- toderm divided, while the rest remained undivided. In these cases also the small refractive droplets were usually collected in this sector. These facts all support the idea that the refract- ive substance forms the membranes of the cleavage spheres. PHYSIOLOGICAL EFFECTS OF LACK OF OXYGEN 393 VI. THE EFFECT OF CARBON DIOXIDE UPON THE PROCESSES OF CLEAVAGE IN THE CTENOLABRUS EGG If eggs are introduced into a current of pure carbon dioxide (which has been carefully washed), we must expect to obtain, besides the effects of mere lack of oxygen, the specific chemi- cal effects of the CO3. Even though everything indicates that the action of CO2 is qualitatively different from the action of simple lack of oxygen, such differences have only rarely to my knowledge, been demonstrated directly in the cell.1 In the egg of Ctenolabrus, however, these differences are very striking. If freshly fertilized eggs are introduced into a stream of pure CO2, no trace of cleavage occurs, even though the eggs are not kept on ice. Under similar external conditions the eggs kept in hydrogen divided two or even three times. The germs also die much more rapidly in CO 2 than in hydrogen. This constitutes, however, only a quantitative difference. A qualitative difference evidences itself, however, immediately that the air is replaced by a current of CO2 in eggs in the two- or four-cell stage. In these experiments the eggs were kept in a drop of sea-water in an Engelmann gas-chamber. Amoeboid movements (which were first noticed at the periphery of the drop) took place on the surface of the eggs in some ten to fifteen minutes, when a current of carbon dioxide was passed through the cham- ber. Whether the whole protoplasm or only the superficial layer of the protoplasm takes part in these changes could not be determined. I have made a series of camera draw- ings of these movements, which I will reproduce here. Fig. 118 shows the outlines of the four cells of an egg at the beginning of the experiment. Fourteen minutes later this cell had the appearance shown in Fig. 119. One of the four cells, that which was directed toward the periphery of the drop and first struck by the stream of carbon dioxide, sent out amce- i See LOEB AND HARDEST?, Pflilgers Archiv, Vol. LXI. 394 STUDIES IN GENERAL PHYSIOLOGY bold pseudopodia. A few minutes later all of the cells sent out such pseudopodia, which soon became shorter, however, as if the substance of the pseudopodia had been torn, e. #., through an emulsion (Fig. 120). The outlines of the germ then again became smooth, but not entirely so (Fig. 121), and finally the blastoderm gradually disappeared (Fig. 122). The entire series of changes shown in Figs. 118-22 took about forty- five minutes. Besides these changes, another series took place in FIG-118 the blasto- derm and the yolk, which, however, I am not as yet able to interpret^ and which I therefore do not de- scribe, as their description would take up much room without at present being of any use. If eggs in an advanced state of division are introduced into CO 2 , a solution of the cleavage-cells occurs at the periphery just as in hydrogen. VII. THE EFFECT OF PURE OXYGEN UPON CLEAVAGE In embryological literature one at times encounters the statement that the processes of development in pure oxygen at atmospheric pressure go on differently from those in air. Demoor also states that nuclear division is accelerated in pure oxygen. Now, it is one of the established facts of physiology that FIG. 119 FIG. 120 PHYSIOLOGICAL EFFECTS OF LACK OF OXYGEN 395 the consumption of oxygen is, within wide limits, independent of the partial pressure of the oxygen, and that it makes nor- mally little difference for the processes of oxidation whether we breathe air or pure oxygen. Still, in order to determine experimentally the action of pure oxygen upon cleavage, I made the following experiments. An inverted ten -liter bottle A (Fig. 123) was filled with pure oxygen. A long glass tube a and a short one b passed through the rubber stopper in the bottle. The glass tube a was connected with an Engelmann gas- FIG. 121 chamber 7. The short glass tube b was connected with a longer tube c, and the bottle B was filled at the beginning of the experiment with water. A second short glass tube passed through the stopper of the latter and was connected to the Engelmann gas-chamber 77. The connecting rubber tube between A and B was filled at the beginning of the experiment with water and closed by a pinch-cock. As soon as the pinch-cock was opened the oxygen was driven out of A through the gas- I chamber 7 by the flow of the water I out of B, and the same amount of air was suctioned through the gas- / chamber 77 into t1'^ bottle B. In ; this way the effecx of pure oxygen ..-•*• could be compared with that of atmos- FIG ^' '' "' pheric air. A few important but self- evident details in the arrangement of the experiment have been omitted in the drawing. In one experiment eggs which had been in the eight-cell stage, but the cleavage-cells of which had been fused by ex- 396 STUDIES IN GENERAL PHYSIOLOGY posure to a current of hydrogen, were introduced into both gas-chambers. I wished to determine whether the renewal of segmentation would occur more rapidly and differently in pure oxygen than in air. The result was that after fifty minutes cleavage occurred almost simultaneously in both gas-chambers and in exactly the same way. Cleavage occurred only at the periphery and the cells which were formed were about the size of those found in the thirty- two- or sixty-four-cell stage. In a second experi- ment cleavage occurred even a little more rapidly in the air than in pure oxy- gen. For the rest things were about the same. Under these circumstances I saw no reason for con- tinuing these experiments ; they showed clearly enough that it does not 'matter, so far as the renewal of cleavage of liquefied Cten- olabrus eggs is concerned, whether air or FIG. us PHYSIOLOGICAL EFFECTS OF LACK OF OXYGEN 397 supplied to them. We found before that a lack of oxygen does not retard cleavage as long as cleavage is at all possible. In the same way an excess of oxygen does not accelerate the process. VIII. EFFECT OF LACK OF OXYGEN ON THE CLEAVAGE OP THE FUNDULUS EGG The eggs of Ctenolabrus have a lower specific gravity than sea-water, and therefore float at the surface of the water. Here they find the oxygen necessary for their devel- opment. If the eggs of Ctenolabrus with their great need for oxygen had a specific gravity large enough to cause them to sink to the bottom, they could scarcely develop in many places, since at the bottom of the ocean where processes of putrefaction are going on, the tension of oxygen is much less than at the surface. We may therefore expect, in gen- eral, that fish eggs which sink to the bottom of the ocean and develop there are much more independent of oxygen than the egg of Ctenolabrus. This is really often the case. The egg of Fundulus has a greater specific gravity than sea- water and develops at the bottom of the ocean. I have shown that the egg of Fundulus can develop for some time in the absence of oxygen. In these experiments the eggs were introduced with a few drops of sea-water into a small glass tube sealed at its lower end, and this tube was put into a test-tube containing several cubic centimeters of an alkaline pyrogallol solution. The test-tube then was sealed at the top. The pyrogallol solution was prepared according to Hempel's directions, and the oxygen must have been ab- sorbed in a short time. Nevertheless, the eggs not only segmented, but they developed as far as normal eggs do in about fifteen hours after fertilization. A large blastoderm was formed which spread over a great part of the surface of the egg. 398 STUDIES IN GENERAL PHYSIOLOGY In order to be able to compare these results with those obtained on the Ctenolabrus egg, I repeated the experiments on Funduhis, using the same method of replacing oxygen by hydrogen, and the same apparatus which had been used in the case of the Ctenolabrus egg. The results obtained were in entire harmony with our earlier findings. When freshly fertilized eggs of Fundulus are introduced into the Engelmann chamber, and a vigorous stream of hydrogen is passed through it, the eggs divide not only once, but continue to do so for fifteen to twenty hours, until a blastoderm is formed which extends over the sur- face of the egg. The result was the same when the eggs were put in an Engelmann chamber and kept for two and one-half or three hours on ice, during which time they were exposed to a vigorous stream of hydrogen. When the eggs were then exposed to room temperature, segmentation at once began and continued in a regular manner. During the entire course of the experiment hydrogen was permitted to pass through the chamber. As long as the number of the cleavage-cells was so small that they could be counted, it could be seen that develop- ment without oxygen occurred as rapidly as in oxygen. Whether this holds also for later stages when cleavage approaches the standstill cannot be determined, as the cells are then too small to allow one to count them. Not only cleavage, but also growth, of the blastoderm, that is to say, increase in area (at the expense of the yolk (?), 1903) occurs in the absence of oxygen. The blastoderm grows from a small area to a large area on the surface of the yolk. If Fundulus eggs are allowed to remain more than twelve to fifteen hours in hydrogen, the cells nevertheless do not liquefy, as is the case in Ctenolabrus in the absence of oxygen. Even after twenty-four hours no such phenomena are observable in the Fundulus egg. I have shown in PHYSIOLOGICAL EFFECTS OF LACK OF OXYGEN 399 previous papers that such eggs do not lose their power of dividing even after being kept for three to four days with- out oxygen. On the other hand, I noticed a collection of the strongly refractive droplets in the furrows between the cells in Fundulus eggs also. Our observations on the mechanics of cell-division, therefore, seem to hold also for the Fundulus egg, only that the material for the surface layer of the Fundulus cells seems to be different chemically from that of the Ctenolabrus cells in that the latter in the lack of oxygen flows together into droplets, while the former undergoes no such structural changes. On the other hand, the Fundulus egg is very sen- sitive to carbon dioxide. If a current of carbon dioxide is passed through the gas-chamber in which are contained the freshly fertilized Fundulus eggs, not a single cleavage occurs. Furthermore, the eggs which have resided for only four hours in such a current of carbon dioxide have lost their power of development for all time. This is of great impor- tance in judging of the effects of lack of oxygen — it points to the possibility that the resistance of the protoplasm to lack of oxygen is not so very different in the Ctenolabrus egg from that in the Fundulus egg, and that only a second- ary molecular change — the disintegration of the surface layer of the cells into a number of droplets — brings about a rapid destruction of the Ctenolabrus cells. This possibility is supported by another fact. I have pointed out in an article, which I have already cited, the remarkable indifference of the Fundulus egg to the concen- tration of the sea-water. This year Professor W. W. Nor- man made similar experiments in my laboratory upon the Ctenolabrus egg. In these it was found that the Ctenolabrus egg is almost as insensitive to an increase in the concentra- tion of the sea-water as is the Fundulus egg. I should not like to conclude this section without adding 400 STUDIES IN GENERAL PHYSIOLOGY a word on the importance of comparative methods in physi- ology. If we had confined our experiments to the Cteno- labrus egg, a generalization of the facts observed would have been as follows: Cleavage is impossible without oxygen. Had we confined our experiments to the Fundulus egg, we should have come to the opposite conclusion. In reality, conditions are such that in some forms a cleavage is possible without oxygen, while in others it is impossible. The same may be said regarding protoplasmic motion. I do not as yet consider it as settled that every muscle is able to do a large amount of work without free oxygen. IX. THE EFFECT OF THE REMOVAL OF OXYGEN ON THE SEGMENTATION OF SEA-URCHIN EGGS If freshly fertilized sea-urchin eggs are introduced into a gas-chamber and a strong current of hydrogen is sent through it, one cleavage always occurs, and sometimes two. If, however, before beginning the actual experiment, all of the oxygen necessary for cleavage is driven out of the eggs and the gas-chamber (by placing the latter upon ice for two hours and sending a current of hydrogen through it), no cleavage occurs, even though we wait from three to four hours. If after this the eggs are again exposed to air, cleavage begins in about forty to fifty minutes. But all the eggs first divide into two cells, and only a few divide at once into three or four cells. The number of the latter is not greater in the experimental eggs than in the normal eggs of the same culture. Such phenomena are very probably attributable to polyspermia. These facts show that in sea- urchin eggs neither a division of the cell nor of the nucleus is possible without oxygen. In this particular they behave like the eggs of Ctenolabrus. We must now raise the ques- tion: Is the inability of cleavage in sea-urchin eggs also the consequence of molecular changes which are brought about by lack of oxygen ? This, indeed, seems to be the case. PHYSIOLOGICAL EFFECTS OF LACK OF OXYGEN 401 If the eggs have divided into two or four cells, and the oxygen is then removed completely from them, the cell-limits become indistinct in about three hours. The cells then absorb water in consequence of the effects of lack of oxygen. The volume of the eggs increases, and the space within the membrane is soon filled uniformly with the protoplasm of the cleavage-cells. The outlines of the cell then become invisible, and the egg looks as if it had never divided. If oxygen is readmitted, the eggs cleave anew, if too long a time is not allowed to elapse. In many cases the old lines of cleavage reappear, but this is by no means always the case. The changes remind one of those in the eggs of Ctenola- brus, only that they occur more rapidly and more distinctly in the latter than in the eggs of the sea-urchin. The surface of the cleavage-cells of the Arbacia is pig- mented, and the pigment granules move upon the surface of the egg during cleavage. I do not doubt that by more care- ful study phenomena similar to those observed in the cleavage of the Ctenolabrus and the Fundulus eggs will be observed in the case of Arbacia also. The fact has been mentioned that, in general, the cleavage of the Fundulus egg without oxygen occurs not only just as rapidly as under normal conditions, but even a little more rapidly, as stated in my paper on " The Relative Sensitiveness of the Fundulus Embryos in the Different Stages of Devel- opment against Lack of Oxygen." In that article, however, I attributed this difference in time to the increase in tempera- ture brought about in sealing up the test-tubes used in the experiments. Since I again noticed these changes this year, first in the Ctenolabrus egg, and later in the Arbacia egg, in an Engelmann chamber — where there was, therefore, no considerable increase in heat — I decided to determine by more careful experiments whether this difference in time is indeed dependent entirely upon differences in temperature, 402 STUDIES IN GENERAL PHYSIOLOGY or whether the altered metabolism in the initial lack of oxygen does not at first lead to a slight acceleration of cleavage. If the latter were correct, it would give a basis for the explanation of a very purposeful arrangement in organic nature, namely, the increase in respiratory activity in the lack of oxygen. For if lack of oxygen leads to such a universal change in metabolism that more energy is at first set free than under normal conditions, then the purposeful arrangement of the respiratory center is only a special case of a general property of protoplasm. Yet the acceleration of cleavage in the Engelmann chamber might also be dependent upon an increase in temperature. One source of this increase in temperature might be sought in these experiments in the heat produced in developing hydrogen from zinc and sulphuric acid. The gas was passed through four wash-bottles before reach- ing the gas-chamber, yet it might nevertheless have caused an increase in the temperature in the gas-chamber. To render this impossible or less possible the gas generator was packed in a vessel with ice before beginning the experiment. From this the hydrogen was led through a bottle filled with chipped ice which was in turn again packed in ice. The first three wash-bottles were also kept on ice. The temperature of the last wash-bottle through which the gas passed before reaching the gas-chamber was carefully watched before and during the experiment. No increase in temperature was noted when the hydrogen was passed through it. The same water was used for the eggs in the gas-chamber that was used for the control eggs. Every decrease in the temperature of the latter through evaporation of the water was carefully avoided, and their temperature carefully watched. Cleavage in the eggs kept in the gas-chamber neverthe- PHYSIOLOGICAL EFFECTS OF LACK OF OXYGEN 403 less preceded that in the normal eggs by three or four min- utes. The experimental eggs as well as the control eggs were fertilized at the same time and in the same dish with a large amount of sperm. The process of driving out the oxygen by hydrogen was begun some ten or fifteen minutes after fertilization. About half an hour later cleavage occurred, usually first in the gas-chamber. At this time all the oxygen was probably not yet driven out of the eggs, so that we were dealing only with a partial lack of oxygen. This partial lack of oxygen, therefore, often brought about an acceleration of cleavage equal to 6 to 10 per cent, of the time necessary for the first cleavage.1 These experiments give one the impression that when lack of oxygen has reached a certain stage, a transitory increase in the development of energy occurs within the egg at first (through the formation of poisonous substances?). This increase in the development of energy, which, in the case of the respiratory center, is of enormous practical importance, therefore seems to appear also in such cases where its appearance is entirely unimportant, as in cleavage. I will not yet commit myself definitely to the statement that in case of a partial lack of oxygen a transi- tory acceleration of cleavage occurs; but to trace back the purposefulness of organized nature to the general chem- ical and physical properties of protoplasm seems to me much more promising than the assumption of natural selection. If we summarize the results of these experiments on the effects of lack of oxygen on cleavage, we find that in the Fundulus egg, where in the absence of oxygen no dissolution of the cell-walls of the cleavage-spheres occurs, cleavage can continue for more than ten hours without oxygen; while in 1 1 am still inclined to believe that, in spite of all the precautions, the hydrogen had a slightly higher temperature than the air when it reached the egys. [ T'OSJ 404 STUDIES IN GENERAL PHYSIOLOGY the Ctenolabrus, and eggs1 which cannot cleave without oxygen, the surface layer of the cleavage-cells is liquefied and the cells fuse together. The latter fact seems to indi- cate that cleavage does not occur in certain eggs, because without oxygen profound molecular changes occur, which, among other things, prevent the formation of a membrane or a specific surface film. X. ON THE EFFECT OF LACK OF OXYGEN ON CARDIAC ACTIVITY IN FISH EMBRYOS The older experiments on the effect of lack of oxygen on the activity of the heart have in part led to strange results. Tiedemann, for example, found that when the heart of frogs or salamanders is excised and kept under the bell of an air-pump, it ceased to beat in less than one minute when the air is rarified.2 Castell3 came to more probable results. He found that when the heart is cut out of the body of a frog and kept in an indifferent medium in the absence of oxygen, it may continue to beat for an hour. In the experiments of Pfltiger and Aubert, which have already been mentioned, the heart continued to beat after all the spontaneous movements of the animal had long ceased. The older authors had discussed the question as to whether oxygen does not have a direct stimulating effect upon the heart. This would, of course, explain why the heart ceases to beat when oxygen is lacking. Castell, however, showed that a heart which has ceased to beat in an atmosphere free from oxygen will also not beat when stimulated by other means. The papers which have been cited in the introduc- tion give a more rational explanation of the role of oxygen 1 This phenomenon is less distinct, and therefore not so certain, in the egg of Arba- cia as in that of Ctenolabrus. Driesch questions it in the sea-urchin egg, but I am not certain that his experiments are identical with mine. [1903J 2 ArchivfUr Anatomic und Physiologic, 1847, p. 490. 3 Ibid., 1854, p. 226. PHYSIOLOGICAL EFFECTS OF LACK OF OXYGEN 405 than that furnished by the assumption that the oxygen "stimulates" the heart. I was especially interested in comparing the effects of lack of oxygen on the beat of the heart in Ctenolabrus and Fundulus embryos. Does the same difference in behavior toward lack of oxygen exist here as in regard to cleavage ? The heart begins to beat and the circulation is estab- lished in Ctenolabrus embryos as early as forty-eight hours after fertilization. If such forty-eight-hour-old embryos, which are still contained within the eggs, are introduced into a gas-chamber through which a current of hydrogen is passed, the heart usually comes to a standstill in from three to ten minutes after the current of gas is turned on. The activity of the heart does not, however, gradually fall to zero, but the heart comes to a standstill suddenly when the number of heart-beats has decreased but little or not at all. In one case the heart beat about 90 times a minute before the hydrogen was admitted. Hydrogen was then passed through the gas-chamber, and after four minutes the heart still beat 89 times; two minutes later it beat 78 times, and in the following minute 77 ; in the next minute the heart came to a sudden standstill. After hydrogen had been passed through the gas-chamber for only seven minutes, and when the number of heart-beats had fallen only from 90 to 77 — a slight de- crease only — the heart suddenly stood still; at that time blood was still circulating beautifully. In a second experiment the number of the heart-beats was 108 per minute at the beginning of the experiment. Two minutes after turning on the hydrogen gas the heart beat 105 times, and three minutes later 108 times a minute. During the next minute the heart stood still after having beaten 23 times in the first eighteen seconds of that minute. The heart stood absolutely still for four minutes, after which it gave a few weak pulsations. For the next three minutes 406 STUDIES IN GENERAL PHYSIOLOGY it again stood still, after which the heart beat rhythmically for one minute (38 beats in a minute), when it again ceased. A few irregular pulsations followed, and then everything was over. Sixteen minutes after turning on the current of hydrogen the heart had come to a complete standstill, but the embryo itself still moved at this time, and even five minutes after the heart and the circulation had ceased entirely the embryo still moved! In a third experiment the current of hydrogen was turned on at 11:26 A. M. The number of beats was 90 per minute; in the following minute it was 81, and in the third minute the heart came to a sudden and permanent standstill. In a fourth experiment the current of hydrogen was started at 10:03 A. M. The number of heart-beats was 100 per minute. The following table indicates the course of the experiment: 10 : 03 100 beats per minute 10:04 102 " " " 10:05 - 100 " " " 10:06 96 « « « 10:07 - 98 " " " 10:08 90 " " " 10:11 - 60 " " " 10:12 54 « « « 10:13 - 54 " « « The heart then came to a sudden standstill. Three min- utes later the heart again beat twice; shortly after this it beat regularly for one minute (39 times per minute). The heart then again stopped; a few scattered beats followed, and at 10:25 A. M. the heart came to a permanent standstill. When the embryos whose hearts had come to a standstill were returned, after not too long a time, to water containing oxygen, resuscitation of the heart followed, and this the earlier, the shorter the time the embryo had remained in the atmosphere free from oxygen. If the eggs remained for one to one and a half hours in the gas-chamber, they became PHYSIOLOGICAL EFFECTS OF LACK OF OXYGEN 407 opaque and sank to the bottom. Twenty-five minutes after turning on the hydrogen the changes which we have described in detail above — namely, the appearance of the strongly refractive droplets — were often clearly visible. If now we ask for the cause of the rapid and sudden standstill of the heart of Ctenolabrus embryos when deprived of oxygen, we must admit, first of all, that a failure of the energy which is supplied perhaps by processes of oxidation cannot be the cause. For, since the oxygen is replaced by hydrogen only gradually, the number of heart-beats should under these circumstances also decrease only gradually until a minimum is reached. The behavior of the heart was, how- ever, entirely different. The heart usually came to a stand- still without a noteworthy decrease in the number of heart- beats; sometimes a decrease was noted. For the same reasons the view that in three to ten minutes after iurning on the current of hydrogen all the potential energy present in the heart has been used up is also to be set aside. After the heart had ceased to beat, the entire animal still executed spontaneous movements, and the heart remained generally active in case of lack of oxygen longer than the rest of the body of an animal.1 The rapid and sudden standstill of the heart of Ctenolabrus is the consequence either of a sudden poisoning, or of a structural change in the heart brought about by the removal of oxygen. It might also be that the poisonous effect consists only in bringing about molecular changes. The experiments on the cleavage of the Ctenola- brus egg showed that a change occurs in the cell-walls in consequence of which they break up into droplets. We must assume that these changes are brought about by the beginning lack of oxygen, or the metabolic products formed in consequence of this lack of oxygen. Might it not be pos- sible that a liquefaction of solid elements arid the formation i Miss Moore has since found that in young fish whose respiratory and sponta- neous motions have ceased the heart still continues to beat for hours. [1903] 408 STUDIES IN GENERAL PHYSIOLOGY of droplets hinders the production or the transmission of molecular movements, and in this way brings about the sud- den standstill of the heart ? This idea would also harmonize very well with the fact that the heart comes to a standstill as suddenly and as unexpectedly as death ensues from embol- ism. It would also be in harmony with this idea that after the sudden standstill of the heart a few occasional heart- beats may yet appear. We will, however, not enter farther into the field of hypotheses, but rather attempt to see how the heart of Fundulus behaves in the lack of oxygen. Numerous experiments on embryos from four to ten days old (the embryos do not hatch until after the twelfth day) showed without exception the following behavior of the heart in the case of lack of oxygen: During the first ten to twenty minutes after the hydrogen is turned on through the gas-chamber, the number of heart- beats does not decrease. A transitory acceleration even occurred, which, however, was brought about through a rise in temperature caused by passing the hydrogen gas through the gas-chamber. This acceleration did not occur when I packed the hydrogen generator in ice. But the decrease in the amount of oxygen contained in the Fundulus egg, which occurs during the first twenty minutes and which causes the heart of the Ctenolabrus embryo to stand still, has no eft'ect upon the rate of the heart of the Fundulus embryo. Then follows a period of steady decrease in the number of heart-beats, which continues for about one and one-half hours. The decrease occurred most rapidly at first and then more slowly. During this period the number of heart-beats fell from about 120 or 100 a minute to about 20 per minute. This period corresponds, it seems to me (and we shall find further proofs for this idea later), to the period of progres- sive decrease in the oxygen necessary for the oxidations in the heart. PHYSIOLOGICAL EFFECTS OF LACK OF OXYGEN 409 When the number of the heart-beats has decreased to the minimum of about 20 per minute, the heart continues to beat at this rate for about eight to ten hours in an uninter- rupted and regular manner, until at the end of this time it comes to a standstill. Since our earlier experiments rendered it possible that after two hours all the exhaust- ible oxygen has certainly been driven out by the current of hydrogen, we are perhaps justified in assuming that the energy for this long- continued and regular, but slow, activity of the heart is derived from processes of hydrolysis. It seems as if we are able in the Fundulus heart to separate numerically the energy derived from hydrolytic processes from that derived from processes of oxidation, in that the former source of energy yields about 20, the latter the remaining, about 80 to 100, heart- beats per minute. I would especially emphasize the fact that during the entire time of the experiment the lo FIG. 124 current of hydrogen was passed through the gas-chamber uninterruptedly, and that in consequence every action of the carbon dioxide had been shut out in these experiments, as in those upon the Ctenolabrus embryo. We shall now describe a few of the individual experi- ments. In one case the hydrogen current was turned on at 8:42 A. M. The number of heart-beats was 108 to 114 per 410 STUDIES IN GENERAL PHYSIOLOGY minute. This number remained constant until about 9:08. (The current of hydrogen was not as vigorous as usual.) At 9 : 12 the number of heart-beats was 96 ; at 9 : 30 the number was 69; at 10 the number was 48; and at 11 it had fallen to 27. At 11:25 the heart beat 23 times per minute; at 11:40 it beat 20 times per minute ; after which the number of beats varied between 20 and 23 per minute, until 8:45 P. M. ; in other words, more than nine hours. The curve of Fig. 124 illustrates the condition of affairs better than description. The curve is typical and may be looked upon as representing any one of these experiments. Only the absolute values varied with different individuals and with the temperature. In another experiment the current of hydrogen was turned on at 3:06 A. M. The number of heart-beats was 120. At 3:17 the heart beat 126 times, after which the number decreased, as shown in the following table: 3:20 - 110 beats per minute 3:22 86(!)" " " 3:25 60 " " " 3:27 54 " " 3:31 - 50 " " " 3:34 .... 44. « « « 3:40 - 36 " « " 3:45 - 33 « « « 3:52 - - - 24 " " " 4:00 22 " " " 4:05 - . 20 " " " 4:12 - 19 " « " 4:20 - 16 " " " 4:30 - 14. « « « 4:55 - 12 " " « This rate continued unchanged until 9:50, when the experiment was brought to a close. It is readily seen how much more rapidly the decrease occurs at first than later. Fig. 125, which illustrates the beginning of this experiment, shows this very strikingly. PHYSIOLOGICAL EFFECTS OF LACK OF OXYGEN 411 I have repeated this experiment eight times, always with the same result. It was of importance now to determine whether the number of heart-beats increases, and how much it increases, when a heart which has attained its minimum rate in hydrogen is again exposed to the oxygen of the air. T In one such experiment the current of hydrogen was turned on at 9:10 A. M. The number of heart-beats was 120 per minute. At 11 the number of heart-beats had fallen to 42, and soon thereafter the minimum of 24 heart-beats was reached. At 2 : 40 the number of heart-beats was still 24. At 2 : 44 the embryo was taken out of the gas-chamber and brought into fresh water, and at 2:48 the num- ber of heart-beats was counted; it was then 30. The further course of the experi- ment is shown in the following table: FIG. 125 2:48 2:49 2:50 2:55 3:00 3:03 - 40 beats per minute 51 " fiO " " " 66 " " " - 66 " " 69 " " 412 STUDIES IN GENERAL PHYSIOLOGY *l£kQ ^"Gt T"l£*T* 1 minn i"f* 3:08 81 JtJdla ^Jt!l J U U LUJLUUM? 3:15 84 u u " 3:25 96 u a U 3:35 - 102 u u " 3:47 111 a a U 3:53 - 120 u a u This rate continued until 5 :10, when it increased to 132. ! This experiment, which I repeated several times with the same result, shows that the decrease in the number of heart- beats when the heart is deprived of oxygen is dependent chiefly upon the decrease in the energy furnished by oxida- tion and not upon the formation of poisonous substances. The fact that the minimum number of heart-beats continues a very long time without oxygen also speaks against the latter idea. We can make use of still another method to determine what proportion of the heart-beats in the Fundulus embryo depends upon oxidations, and what proportion upon pro- cesses of splitting. By placing the gas-chamber upon ice and passing a current of hydrogen through it, we are able to drive out the oxygen, while the processes of hydrolysis are at the same time reduced to a minimum through the lower- ing of the temperature. In one experiment I passed the hydrogen through the gas-chamber for two hours, while keeping it on ice. The hearts were then removed from the ice, but the current of hydrogen was maintained. At room temperature the number of heart-beats, which at the begin- ning of the experiment had been 117, rose to 87 (in twelve minutes), to descend again to 36 in the course of the next hour. Forty minutes later the minimum of 21 was attained, at which rate the heart continued to beat for seven hours. Toward the last a slight increase occurred i This experiment shows that the oxygen diffuses comparatively rapi'dly into the egg. [1903J PHYSIOLOGICAL EFFECTS OF LACK OF OXYGEN 413 in the number of beats. I had expected that the number of heart- beats would be only the minimal one after removing the gas-chamber from the ice. Possibly all the oxygen had not been driven out. I therefore repeated the same experi- ment, but allowed the gas-chamber to remain for three hours on the ice. This time I expected that at room temperature the number of heart-beats would only reach the minimum which corresponded to the temperature. But this time also the number of heart-beats rose in six minutes to 66, after which the rate decreased steadily. One hour later the heart beat 42 times, and after thirty-five minutes the minimum of 24 was reached. I do not doubt that after passing a vigor- ous current of hydrogen through the gas-chamber for three hours all the oxygen is exhausted from the egg. If this assumption is correct, these experiments can be made to har- monize theoretically with the results obtained earlier' only by assuming that the processes of hydrolysis do not occur with uniform intensity, but that they occur much more rap- idly at first when the oxygen is first withdrawn (or perhaps also under the ordinary conditions of oxygen supply) than in the continued lack of oxygen.1 It is, moreover, to be noted that the data necessary for calculating the work of the heart are lacking in these ex- periments. Only by assuming that these data are the same in the presence of oxygen as in its absence can conclusions be drawn as to the behavior of the two sources of energy. If we make this assumption, we come to the conclusion that of all the energy which is used up by the Fundulus embryo in normal heart-activity, that much at least which .cor- responds to the minimal number of heart-beats in the lack of oxygen is dependent upon processes of hydrolysis. This number is about one-sixth or one-fourth of the total number of heart-beats which occur under normal conditions of i Perhaps in this case the effects of poisonous substances are to be considered. [1903] 414 STUDIES IN GENERAL PHYSIOLOGY oxygen supply and at the same temperature. When, how- ever, we consider the results of the experiments carried on in the cold, we come to the conclusion that in the presence of oxygen the proportion of energy obtained through processes of splitting may be much greater than this; it may then amount to 50 or 70 per cent, of the work done by the heart. The behavior of the heart of a Fundulus embryo in car- bon dioxide is of interest in so far as it shows that carbon dioxide is just as poisonous in this case as upon the heart of the Ctenolabrus. While the Fundulus heart continues to beat for twelve hours, and even longer, when the oxygen is driven out by hydrogen, the ventricle ceases to beat as early as twelve minutes after passing carbon dioxide through the gas-chamber. Only the auricle continues to beat, and the circulation soon comes to a stop. The contractions become weaker and less numerous. In one experiment the heart beat 96 times per minute at the beginning of the experiment, 54 times after eight minutes, 45 times after ten minutes, and 42 times after twenty minutes. The heart then ceased to beat entirely for long periods of time, and thirty-two minutes after turning on the carbon dioxide the heart stood still. In other experiments the heart did not cease to beat until after one and one-half hours. When the heart is exposed to the poisonous effect of CO2 and ceases to beat even after one hour, the heart begins to beat again when the carbon dioxide is replaced by air. The resuscitation of the heart is as follows: The auricle recovers more rapidly than the ventricle, and the latter at first beats a less number of times than the former. In one ex- periment a heart which had come to a standstill was exposed to air at 10 A. M. At 10:06 the auricle beat 24 times per minute, while the ventricle was still quiet. The ventricle did not begin to contract until the next minute, and the number of auricular contractions was 33 a minute at this PHYSIOLOGICAL EFFECTS OF LACK OF OXYGEN 415 time. At 10:23 the auricle beat 72 times, while the ven- tricle beat 42 times per minute. The ventricle often con- tracted only once to every two, or even at times three, auric- ular contractions. At 10:35, however, the rate of the ventricular and the auricular contractions was the same, namely, 84 per minute, and from then on they continued the same. These phenomena, which are characteristic of all experiments with carbon dioxide, were never observed in replacing the air by pure hydrogen. In another series of experiments I permitted CO2 and hydrogen to pass alternately through the gas-chamber. In one case I turned on the hydrogen at 8:30 A. M. At 10:30 the heart beat 24 times per minute; at 10:31 the hydrogen current was interrupted and the CO2 current was turned on. (Through a simple T-tube connection and a pair of pinch cocks it was possible to pass either the hydrogen or the CO8 through the gas-chamber at will, without admitting air.) In a few minutes the ventricle ceased to beat and the circula- tion stopped. After an hour the current of carbon dioxide was interrupted, and hydrogen was again passed through the chamber. After forty minutes the ventricle again be- gan to beat ; the number of its beats was 24, and remained so until death. By replacing the C02 by hydrogen it is therefore possible to do away with the poisonous action of the former. This experiment demonstrates very nicely a fact which is perhaps doubted by no one: that carbon dioxide and lack of oxygen have entirely different effects, which in ordinary cases of asphyxia are added together.1 In this way it is possible by passing through the chamber a current of pure hydrogen gas to bring to life again a ventricle which has been asphyxiated in carbon dioxide. i It also demonstrates very nicely the possibility that other non-volatile poisonous substances may be formed, by lack of oxygen, which are destroyed again when oxygen is again admitted. The phenomena of fatigue may belong to category. [1903J 416 STUDIES IN GENERAL PHYSIOLOGY Finally, it was of interest to compare the resuscitating effect of air with the resuscitating effect of hydrogen. Fun- dulus embryos were introduced into two gas-chambers. At the beginning of the experiment the heart under observation in one of the chambers beat 90 times a minute; that in the other, 96 times a minute. Hydrogen was passed through the chambers, and after an hour and fifty minutes the frequency of the heart-beats had fallen in both cases to 18 per minute. In place of the hydrogen, carbon dioxide was then passed through the chambers. In fifteen minutes the ventricles stopped beating, and the pulsations of the auricles became much weaker. After 45 minutes one of the hearts was apparently dead, while the auricle of the other still beat 18 times a minute, though the beats were scarcely perceptible. One of the .gas-chambers was then opened and the embryo exposed to the air, while in the second cham- ber the CO 2 was replaced by hydrogen. After fifteen minutes the heart which had been apparently dead and which was exposed to the hydrogen beat 24 times a minute, but only the auricles contracted. Both the auricle and the ventricle of the heart which was exposed to the air beat 60 times. Two hours later the heart beat 30 times per minute in the hydrogen, but the contractions were still limited to the auricle, while the heart exposed to the air beat 72 times. When a little later I exposed the em- bryo kept in the hydrogen to air, the ventricle did not recover. The number of auricular contractions did rise within fifteen minutes from 18 to 54, but shortly there- after the entire heart ceased to beat. In resuscitating a heart poisoned by CO2, oxygen is therefore more effect- ive than the simple removal of the CO2 by hydrogen. We are not able to explain why the ventricle ceases to beat when exposed to carbon dioxide sooner than the auricle. We meet with an entirely different relation between car- PHYSIOLOGICAL EFFECTS OF LACK OF OXYGEN 417 diac activity and oxygen in the larvae of a fresh-water mussel (Cyclas). In this animal the frequency of the heart- beat steadily decreases from 50 heart-beats to 0 in the course of one and one-half hours in an atmosphere of hydrogen (at 24° C.). In this case, therefore, we have neither a sudden standstill of the heart without an appreciable decrease in the frequency, as in Ctenolabrus, nor a long-continued steady beat of low frequency, as in Fundulus, but a decrease in cardiac activity which runs parallel with the removal of oxygen, as if processes of oxidation are the sole source of energy for the activity of the heart. XI. ON THE TRANSFORMATION OF NEGATIVELY HELIOTROPIC ANIMALS INTO POSITIVELY HELIOTROPIC THROUGH LACK OF OXYGEN A series of papers have proved that it is possible to change the sign of heliotropism in certain animals at will through external conditions.1 It is an easy matter, for example, to render negatively heliotropic Copepods posi- tively heliotropic by cooling, and to keep them permanently positively heliotropic at a low temperature ; while it is also possible to render positively heliotropic Copepods negatively heliotropic by an increase in temperature. The same experiments can be made on larvae of Polygordius. In order to determine the cause of this change in the sign of heliotropism, and also the conditions upon which the latter depends, I tried to see whether other conditions could bring about similar changes. Groom and I had previously found that the positively heliotropic Nauplii of Balanus perforatus rapidly became negatively heliotropic when exposed to strong light. I also found that the same effect can be produced upon Copepods and Polygordius larvae by properly diluting the i GROOM TJND LOEB, Biologisches Centralblatt, Vol. X; LOEB, Vol. I, pp. 265 ff. 418 STUDIES IN GENERAL PHYSIOLOGY sea-water as by increasing the temperature, while a proper increase in the concentration of the sea-water brings about the same effect as cooling. The majority of Copepods were, immediately after being caught, positively heliotropic. It seemed as if the majority of the negatively heliotropic Copepods belonged to one and the same species. When the Copepods were allowed to remain for a long time in a vessel containing sea- water, the number of negatively heliotropic animals decreased, becom- ing positively heliotropic with time, while the reverse change occurred only rarely. The experiments on the effect of lack of oxygen were made under a small bell-jar, the contents of which were separated from the air on the outside by mer- cury. Two tubes extended into the bell-jar, one of which conducted the hydrogen into the bell, while the other con- ducted it away from the bell. Two vessels were placed under the bell-jar, of which the one contained freshly selected positively heliotropic Copepods, while the other contained negatively heliotropic Copepods. While the positively heliotropic animals remained positively heliotropic during the course of the experiment, the negatively helio- tropic Copepods within fifteen to twenty minutes after the hydrogen was turned on began, in part, to leave the room side of the vessel and to distribute themselves irregularly throughout the vessel, in part to collect at the window side of the vessel. The number of animals collected near the window steadily increased, while the number of Copepods at the room side of the vessel steadily decreased. In about thirty to forty-five minutes after the current of hydrogen had been turned on all the Copepods lay quietly on the bottom of the vessel. The Copepods which had from the beginning been positively heliotropic died at the side of the vessel nearest the source of light. Most (if not all) of the Copepods which had at first been negatively heliotropic PHYSIOLOGICAL EFFECTS OF LACK OF OXYGEN 419 were also found at the window side of the vessel. A second small collection occurred in the middle of the vessel, while the room side of the vessel was entirely vacated. The ani- mals usually did not become positively heliotropic until shortly before they became motionless. This explains why the conversion of the negatively heliotropic into the posi- tively heliotropic animals through lack of oxygen cannot be obtained with the precision and elegance with which the change can be obtained by cooling. In the latter case the animals retain their full power of movement; in the former the transformation does not occur until the animals have suffered from lack of oxygen. But even then the phenome- non is so striking that it might be used as a demonstration experiment. I have repeated the experiment eight times with the same result. At first it seemed to me as if the negatively heliotropic Copepods died more rapidly in the absence of oxygen than those which were positively helio- tropic from the beginning. This finding, however, was not borne out in every case. When the experiment was interrupted early, at a time when the animals first began to become positively heliotropic, and air was then admitted, the Copepods which had become positively heliotropic again became negatively heliotropic. The remarkable effects, which we have described here, of lack of oxygen on the sense of heliotropism, are, of course, not confined to Copepods. I made similar experiments upon the negatively heliotropic marine Isopods, the majority of which also become positively heliotropic in less than two hours when oxygen is withdrawn. These experiments will be continued. We see, therefore, that lack of oxygen has the same effect upon the sense of heliotropism as cooling or increas- ing the concentration of the sea-water. Araki has shown that by cooling the chemical effects of lack of oxygen can 420 STUDIES IN GENERAL PHYSIOLOGY be brought about, and it is therefore possible that the posi- tive heliotropism in both cases is determined by the same chemical conditions. It must be left for further experiment to decide this point. XII. ON CHANGES IN PIGMENT CELLS IN LACK OF OXYGEN It is a definitely established fact that the pigment cells in the skin of the frog become lighter after death. This lightening is brought about, as Biedermann has found,1 by the fact that the coloring matter collects into small clumps. A piece of the skin which has been deprived of its circu- lation shows the same changes. In the transparent portions of the skin which can be studied microscopically — such, for example, as the web of the amputated foot of Rana temporaria — it can easily be seen how the much- branched pigment cells which follow the course of the capillaries gradually change their form, in that the coloring matter moves toward the center of the cell until finally all the pigment is col- lected into clumps (p. 1 75). Increase in the carbon dioxide cannot be the cause of this change in the pigment cells, for Biedermann found that the skin does not become lighter when the frog is poisoned with CO2. Biedermann believes that the cause is probably to be found in the decrease in the amount of oxygen. The surface of the yolk-sac of the Fundulus embryo is studded with a large number of black and reddish-yellow pigment cells, which are at first distributed irregularly, but which later, as I have shown,2 are compelled to creep upon the blood-vessels and surround them. With this the first physiological cause was furnished for the marking of an animal. Since then other authors have also found that the course of the embryonic blood-vessels determines the mark- ing of the embryo. 1 PflUgers Archiv, Vol. LI. 2 Journal of Morphology, 1893. PHYSIOLOGICAL EFFECTS OF LACK OF OXYGEN 421 The black and red pigment cells can be distinguished from each other, not only by their color, but also by their form. The latter send out a large number of thin pseudo- podia-like processes which are never found in the black pig- ment cells. In the experiments on the effect of lack of oxygen on the cardiac activity of the Fundulus embryo, it was noticed that the originally dark yolk-sac gradually be- came lighter in color when exposed to hydrogen for a long time. The pigment cells can be observed very carefully with the microscope, and I expected to observe the same phenomena that Biedermann observed in frogs. This was, however, not the case. It was found in the course of a series of experiments that the dark pigment granules and the black cells gradually disappear the longer the current of hydrogen is kept up, and that the collection of the -pigment in the center of the cell does not occur. The changes in the red pigment cells in lack of oxygen are of a somewhat different nature. The lightening of the color often occurs in this case also. Besides this, however, the cells become smaller. The tips of the cell-processes break off, remaining visible at first as tiny droplets, which disappear later. As this process continues, the pigment cells gradually become smaller. These changes remind one of the fact that certain dyes become colorless when reduced. In our experiments it might also be possible that the discoloration of the black pigment is a result of a reduction, which does not occur in the presence of atmospheric oxygen. XIII. CONCLUDING REMARKS It seems to me that the most important result of the foregoing experiments consists in the proof which has been brought forward that in certain cases at first molecular, and later morphological, changes are brought about in cells 422 STUDIES IN GENERAL PHYSIOLOGY through lack of oxygen, which in their turn are the cause of the suspension of life-phenomena. This has been proved for the process of cleavage in the Ctenolabrus egg. The cleavage-cells of Ctenolabrus are dissolved again and fuse together when oxygen is removed. These changes are not, however, an evidence of death, for as soon as such a fused blastoderm is again exposed to air it begins to divide anew. On the other hand, these molecular changes are sufficient to hinder cleavage. The cleavage-cells of the Arbacia egg seem to suffer similarly in the lack of oxygen, although the changes are much less marked. We find that here also cleavage is impossible without oxygen. Yet lack of oxygen does not bring about the same sort of molecular changes in the Fundulus egg as in the Ctenolabrus egg, and corre- sponding with this difference cleavage may also go on with- out oxygen for many hours in Fundulus. It is also possible that such molecular changes as are brought about by the lack of oxygen in the cell are also the cause of the cessation of other life-phenomena; for example, the beat of the heart (and the activity of the respiratory center). We thus find that in Ctenolabrus, where the first cleavage-cells suffer such profound structural changes through lack of oxygen, the heart of the embryo comes to a standstill very rapidly and suddenly through lack of oxygen before a marked decrease has taken place in the frequency of the heart-beats; while the heart of the Fundulus, whose cells suffer no such structural changes, continues to beat for many hours without oxygen. Since the chemical energy set free in the cells must first be converted into molecular energy in order to bring about the physiological function, it is clear, a priori, that not only a decrease in. the supply of the chemical energy, but any structural change which ren- ders impossible the conversion of chemical energy into the molecular energy necessary for the activity of the tissue, 14 DAY USE RETURN TO DESK FROM WHICH BORROWED LOAN DEPT. This book is due on the last date stamped below, or on the date to which renewed. Renewed books are subject to immediate recall. AUG131962 JUN2 1971 9 REC'D 1 REC. CIR. FEB 1 6 1973 9196ZARR 61979 REC'D LD SEP 5 '69 -2PM LD 21A-50m-8,'61 (Cl795slO)476B