THE BIOLOGICAL BULLETIN PUBLISHED BY THE MARINE BIOLOGICAL LABORATORY Editorial Board p GARY N. CALKINS, Columbia University FRANK R. LlLLEE, University of Chicago E. G. CONKLIN, Princeton University CARL R. MOORE, University of Chicago E. N. HARVEY, Princeton University GEORGE T. MOORE, Missouri Botanical Garden SELIG HECHT, Columbia University T. H. MORGAN, California Institute of Technology M. H. JACOBS, University of Pennsylvania G. H. PARKER, Harvard University H. S. JENNINGS, Johns Hopkins University W. M. WHEELER, Harvard University E. E. JUST, Howard University EDMUND B. WILSON, Columbia University ALFRED C. REDFffiLD, Harvard University Managing Editor VOLUME LVIII FEBRUARY TO JUNE, 1930 Printed and Issued by LANCASTER PRESS, Inc. PRINCE 8C LEMON STS. LANCASTER, PA. 11 THE BIOLOGICAL BULLETIN is issued six times a year. Single numbers, $1.75. Subscription per volume (3 numbers), $4.50. Subscriptions and other matter should be addressed to the Biological Bulletin, Prince and Lemon Streets, Lancaster, Pa. Agent for Great Britain: Wheldon & Wesley, Limited, 2, 3 and 4 Arthur Street, New Oxford Street, London, W.C. 2. Communications relative to manuscripts should be sent to the Managing Editor, 240 Longwood Avenue, Boston, Mass. Entered October 10, 1902, at Lancaster, Pa., as second-class matter under Act of Congress of July 16, 1894. CONTENTS No. 1. FEBRUARY. HOBER, RUDOLPH The First Reynold A. Spaeth Memorial Lecture. The Pres- ent Conception of the Structure of the Plasma Memhrane ..... 1 MONTGOMERY, HUGH The Copper Content and the .Minimal Molecular Weight of the Hemocyanins of Busycon canalicnlatum and of Loli-o pealei ................................................. IS FAURE-FREMIET, K. (irowth and Differentiation of the Colonies of Xoothamninm alternans (Clap, and Lachm. ) ............................ JS I IAI.L, F. G., and ROOT. R. \\". The Influence of Humidity on the Bodv Temperature of Cer- tain Poikilotherms ...................................... 5_' MORGAN, T. H., and TVI.KR, AI.KKRT The Point of Entrance of the Spermato/oon in Relation to tin- Orientation of the Emhryo in E.i^s with Spiral Cleavage ..... 31' Lrrz. BKENTOX R. The Effect of Low < >xy^cn Tension on the Pulsations of the Isolated Holothurian Cloaca .............................. 74 Ar.pATOv, \Y. \\'. Phenotypical Variation in Body and Cell Si/.e of I )ro>ophila melanogaster ........................................... S3 JACOBS, M. H. Osmotic Properties of the Erythrocytc. I. Introduction. A Simple Method for Studying the Rate of I leniolysis ......... In4 No. 2. APRIL, I IOADLKV. LKK.H Some Effects of H^CL on Fertili/.ed and I'nfi-rtili/.ed I;.^.L:- of . \rhacia punctulata ...................................... 1 -'.^ \\'nriAKER. Dorci.A.s. and MORCAN. T. II. The Cleavage o! Polar and . \ntipolar MaKe^ nf the 1-".^^ of ( 'haeti i] >t<.-ru> ........................................... 145 iii iv CONTENTS PAGE REDFIELD, ALFRED C. The Absorption Spectra of Some Bloods and Solutions Con- taining Hemocyanin 150 CONKLIN, CECILE Anoplophrya marylandensis n. sp., a Ciliate from the Intes- tine of Earthworms of the Family Lmnbricidae 176 DEMPSTER, W. T. The Growth of Larvae of Ambystoma maculatum under Nat- ural Conditions 182 SMITH, DIETRICH C. The Effects of Temperature Changes upon the Chromato- phores of Crustaceans 193 No. 3. JUNE, 1930 COE, WESLEY R. Unusual Types of Nephridia in Nemerteans 203 GRAY, I. E., and HALL, F. G. Blood Sugar and Activity in Fishes with Notes on the Action of Insulin 217 BLUM, HAROLD F. Studies of Photodynamic Action. I. Hemolysis by Previously Irradiated Fluorescein Dyes 224 REDFIELD, ALFRED C. The Equilibrium of Oxygen with the Hemocyanin of Limulus polyphemus determined by a Spectrophotometric Alethod 238 HOADLEY, LEIGH Polocyte Formation and the Cleavage of the Polar Body in Loligo and Chaetopterus 256 PICKFORD, GRACE EVELYN The Distribution of Pigment and other Morphological Con- comitants of the Metabolic Gradient in Oligochaets 265 SIVICKIS, P. B. Distribution of Setae in the Earthworm, Pheretima bengueten- sis Beddard 274 JAHN, THEODORE L. Studies on the Physiology of the Euglenoid Flagellates. II. The Autocatalytic Equation and the Question of an Autocatalyst in Growth of Euglena 281 CONTENTS v PAGE HARVEY, ETHEL BROWNE The Effect of Lack of Oxygen on the Sperm and Unfertilized Eggs of Arbacia punctulata, and on Fertilization 288 RAFFEL, DANIEL The Effect of Conjugation within a Clone of Paramecium anrelia 293 SMITH, GEORGE MILTON A Mechanism of Intake and Expulsion of Colored Fluids by the Lateral Line Canals as Seen Experimentally in the Goldfish (Carassius auratus) 313 VATNA, SUP. Rat Vas Deferens Cytology as a Testis Hormone Indicator and the Prevention of Castration Changes by Testis Extract Injec- tions 322 STUNKARD, H. W., and NIGRELLI, R. F. On Distomum vibex Linton, with Special Reference to its Systematic Position 336 CHAMBERS, ROBERT . The Manner of Sperm Entry in the Starfish Egg 344 INDEX 370 Vol. LVIII, No. 1 FEBRUARY, 1930 THE BIOLOGICAL BULLETIN PUBLISHED BY THE MARINE BIOLOGICAL LABORATORY Woods Hole, Massachusetts THE FIRST REYNOLD A. SPAETH MEMORIAL LECTURE1 THE PRESENT CONCEPTION OF THE STRUCTURE OF THE PLASMA MEMBRANE RUDOLPH HOBER PHYSIOLOGISCHES INSTITUT, KIEL Ladies and gentlemen: I feel in this moment that more than ever since the beginning of my scientific life, I have sympathies with this country, where the modern view of general physiology, to which I myself have devoted my life's work, has been developed with perhaps greater success than anywhere else. More than ever here in Woods Hole I feel the genius of Jacques Loeb, who, as no one else since the days of Claude Bernard, taught us so impressively that the most important task of physiology lies in recognizing the general properties of living matter, and who spent here in this place the happiest days of his life doing research work. And sadly here too I remember at this hour my friend, Reynold Spaeth, in whose memory I have the honour to give you this lecture today. Sixteen years ago he came to Kiel with his young wife, an enthusiastic young scientist, eagerly longing to take up physical chemistry as his weapon with which to advance into the undiscovered land of science. And then after his return from Germany, like his great idol Jacques Loeb and many of Loeb's students, he learned to love above all the scientific atmosphere of Woods Hole, —he who was destined to leave us so early, disregarding in his intense eagerness for research the dangers of the tropics. The genius loci of Woods Hole, who apparently holds his protecting hand over general physiology with particular kindness, also moves me to take as the theme of this lecture the present conception of the structure of the plasma membrane. I am sure that with this theme 1 Delivered at the Marine Biological Laboratory, Woods Hole, on September 9, 1929. The announcement of the foundation of the Spaeth Memorial Lecture will be found in the Report of the Director of the Laboratory for 1928 (Biol. Bull., 1929, 57: 22). 1 1 RUDOLPH HOBER I shall enter the sphere of interest of many people who have performed and are still performing physiological studies at Woods Hole. At once, with this theme I recall to mind the investigations which Reynold Spaeth put forward with so much skill, perseverance and enthusiasm during his short residence in my laboratory. Ladies and gentlemen, many of you will agree with me that the problem of cell permeability belongs among the most urgent questions of general physiology, I daresay perhaps of special physiology too. For the working out of a theory of permeability is intimately joined with the understanding of many fundamental phenomena of life, such as nutrition, secretion, absorption, excretion, growth and irritability. Hence let us begin to follow a little the development of the doctrines of the permeability of cells. In 1855 Naegeli described the phenomenon of plasmolysis of the plant cell, consisting in a persisting retraction of the protoplast from the cell wall, if the cell is bathed in what we call today a hypertonic solution. Pfeffer in 1877 gave an explanation of the permanent plasmolysis, comparing the plant cell with the "Traubesche Zelle" made up by a precipitation membrane, for instance, by a copper ferrocyanide membrane. He suggested that the protoplast is sur- rounded by some limiting layer on the outer surface, the plasma membrane having, like the inorganic precipitation membrane, the property of semipermeability, that is, permeability to water but im- permeability to such dissolved substances as produce the plasmolysis. Some time later, Overton gave an unquestionable proof that Pfeffer's assumption was correct; making use of a series of organic compounds, he showed that, in full harmony with the theory of solutions based by van't Hoff upon the experiments of Pfeffer, all solutions which produce the beginning of plasmolysis have the same molecular concentration. Thus at first one was compelled to assume that the interior of the living cell was shut off from dissolved substances and that only water was able to enter. The next important step in recognizing the nature of the limiting membrane of the cell was the discovery of Klebs and de Vries that besides the dissolved substances which cause permanent plasmolysis, there exist some others, — for instance glycerol and urea, which given in hypertonic solution, instead of bringing about the permanent plasmolysis, only produce the initiation of shrinking, which is followed sooner or later by deplasmolysis. Furthermore, Overton and others have discovered a great many substances that do not plasmolyse at all. It was only a logical outcome of the theory of the plasma membrane to explain all this by the assumption that for such substances the mem- STRUCTURE OF THE PLASMA MEMBRANE brane is not impermeable, but allows them to pass faster or slower. These conclusions could be often established beyond doubt by chemical, optical and other forms of analysis of the contents of the cells. But the question now arises as to whether the inorganic precipi- tation membranes behave in perhaps the same manner, that is, whether or not they are permeable to the same substances to which the plant cells, as indicated by the plasmolysis experiments, are permeable. Curiously enough, this question, which is derived so easily from the experiments, has been answered only recently by the systematic experi- ments of Collander.2 This author showed that the copper ferrocyanide membrane behaves also in a very different manner in relation to a great number of organic non-electrolytes, allowing some to pass not at all, others to pass slowly, and still others quickly. But the laws governing the speed of permeation through the precipitation membrane differ widely from those which hold good in the case of the plasma membrane, as is illustrated by Table I. TABLE I Substance Relative per- meability of Rhcco discolor Permeability of Copper Ferrocyanide Molecular Volume Relative Solubility in Ether Methyl alcohol . 125 + + + + + 8 2 0 273 Ethyl alcohol . . 71 + + + + 12 8 1 86 Valeramide 69 + + 28 7 0 170 Ethyl urethane. . 59 0 637 Ethylene glycol 4.4 + + + + 144 0 0068 Diethylurea. . . . 2.0 0 0185 Glycerol . 1.3 + + + 206 00012 Methylurea 1.2 00012 Urea . .... 1.1 + + + + + 13 7 00005 Glucose . . 1.02 + 37 5 <0 0001 Glycocoll . 1.0 + + + + 17.1 < 00001 Saccharose. ... 1.0 + 704 <0 0001 The table contains data with respect to the behavior of twelve organic non-electrolytes. The first series of numbers shows the various speeds of permeation as related to the epidermis cells of Rhvo discolor, the second the speeds of permeation in relation to the copper ferro- cyanide membrane. It can be easily seen that between both there exists no parallelism at all. The third series indicates the molecular volumes calculated by Collander from the values of molecular re- fraction; the fourth series gives their relative solubilities in ether. Now comparing the second and third series, we recognize clearly that the velocity of permeation of the precipitation membrane is a function - Collander, Kolloidchem. Beihefte, 19, 72, 1924 and 20, 273, 1925. 4 RUDOLPH HOBER of molecular volume. This governing rule being established, the character of the membrane is immediately revealed. It behaves as a sieve for molecules so that the size of its pores determines whether or not the dissolved substance can permeate. Such a membrane is semipermeable as soon as the diameter of the molecules of the solution surpasses a certain size. The passage is then allowed only to water, for its molecules are characterized by an especially small volume. And since it is highly probable that the pores of the mem- brane are not all of the same size, the molecules with a diameter below the limiting value have a greater chance to slip through, as they are smaller. Furthermore, the fourth series of numbers shows that the permeability of the plasma membrane might depend upon quite another principle, that is, the principle of solubility in the sub- stance of which the membrane is composed or, briefly, the principle of selective solubility. Thus we come to speak of the first compre- hensive theory of cell permeation, the lipoid theory of Overton. It is a well-known fact that the lipoid theory has been supported by a large amount of powerful arguments, but it is also well-known that one has struggled sharply against it, very often, I believe, with insufficient arguments. But even today the theory cannot con- clusively be judged for the simple reason that the physico-chemical foundation is partly too narrow and partly too uncertain. Professor Jacobs was indeed completely right when he wrote a short time ago: "It may be emphasized that what is most needed in the field of cell permeability at the present day is facts." As everybody knows, Overton based his theory in the first place on the comparison between the speed of penetration of substances and their relative solubility in oil. Collander, recently reviewing most carefully the experiments of Overton on plant cells, advocated especially the correspondence between permeability and solubility in ether.3 Both these authors are quite clear about the limited value of their comparison, and the table also shows that the parallelism is fairly incomplete. Therefore the lipoid theory is still nowadays a petitio principii. However, the thesis that the permeability to the organic non-electrolytes is to be compared to the solubility in organic solvents agrees so often with the experimental data, that I myself have practically no doubt that it is only necessary to discover such solvents as might be still better suited to comparison with the material of the plasma membrane than oil or ether. It is really astonishing that since the lipoid theory was set up more than thirty years ago, so little systematic research work 3 Collander and Barlund, Soc. scient. fenn., 2, 9, 1926; Barlund, Ada botan.fenn., 5, 1929. STRUCTURE OF THE PLASMA MEMBRANE has been done on the relative solubility of organic compounds in different organic solvents comparable to the lipoids of Overton, in order to get a firm basis for the theory. It is well known that an interesting attempt to find a better model was made by Nirenstein some years ago.4 He showed that several exceptions to the rule previously given by Paul Ehrlich, that the vital colors which enter easily into the living cell dissolve in oil, could be removed by trying to imitate the plasma membrane with a mixture of an oil with a fatty acid and an organic fat-soluble base. Table II shows experiments by which I was able to compare the relative solubility of acid dyes in the above- mentioned oil mixture with the relative absorption of colors by red blood corpuscles.5 TABLE II Sulfonic Acid Dyes • Relative Solubility in Oil Mixture Relative Absorption by Blood Corpuscles Sulfonic Acid Dyes Relative Solubility in Oil Mixture Relative Absorption by Blood Corpuscles Wollgrun, Licht- Tropaeolin 1 30 1-3 grun 0 0 Tropseolin 2 30 4-8 Cyanol, EriocyaninJ Orange R . ... 85 7-16 Azofuchsin I . < 1 0.3-1 Brilliant orange R 71 7-16 Azofuchsin G < 1 <1 Metanil yellow 94 10-16 Bromophenol blue . . 27 1-3 It can be seen that there exists a parallelism between solubility and absorption, and it is especially noteworthy that this similarity means not only intensity of staining, but means permeability; for it follows from the table that dyes which are not dissolved in the oil mixture do not enter the blood corpuscles at all. However, such experiments with dyestuffs do not come out quite satisfactorily, as I can easily show. Therefore it is necessary to collect further experimental data to get a clear understanding, inasmuch as the cell permeability is a solution permeability. But there can be no doubt that the cell permeability is not only a solution permeability with regard to an oil-like solvent within the plasma membrane. In the first place in this connection it is a very striking fact that water enters the cell usually with remarkable speed, because it is impossible to reconcile this entrance with the supposition that the membrane consists entirely of an oily phase. Secondly there exists an apparent permeability of certain kinds of cells to inorganic anions, though the inorganic salts are generally not 4 Nirenstein, Pfliiger's Arch., 179, 233, 1920. 5 Unpublished experiments. 6 RUDOLPH HOBER in the least soluble in organic solvents. Thirdly, there are important nutritive materials which cannot get into the cell in any way, but belong also to those substances that are nearly or entirely insoluble in the organic solvents. Now we are able to interpret the first and second points by re- turning to the already-mentioned sieve theory of permeability of the precipitation membranes, and we will see that on that account the comprehension of the structure of the plasma membrane receives a very important supplement. More than thirty years ago Koeppe, Giirber and Hamburger made the discovery, — which has often been verified since, — that the red blood corpuscles have a selective permeability for anions. It is well known that this property has the greatest importance for the buffer capacity of the blood; but it seemed for a long time to be a strange unicum, for which there existed hardly any physical parallel. Otherwise it might have been possible to construct a model to imitate the peculiarities of the membrane of the blood corpuscles. Today the matter is practically clear; the well-known experiments of Michaelis and Collander with artificial membranes, especially with dried collodion membranes, enabled us to understand the singular phenomenon. Michaelis proved that these membranes are, under certain circumstances, the seat of a great potential difference, whose direction and amount is an obvious sign that the membranes are exclusively cation-permeable.6 Therefore there is an analogy between the cation-permeable collodion membrane and the anion- permeable blood corpuscle membrane. At the collodion membrane the anion plays no role, whereas a cation gives rise to an electromotive force which increases as its migration velocity increases or as its diameter decreases. On account of these facts Michaelis has proposed the following hypothesis: the membrane allows only the cation to pass through it as through a sieve; the ions with the smallest diameter pass with the greatest speed, and the entrance of ions into the pores of the membrane is prevented if their diameter exceeds a certain value. That is apparently the reason why, for instance, the earth-alkali ions are unable to pass some collodion membranes characterized by rather narrow pores. In this way we may understand that a quantity of an ion sufficient to'be detected by chemical methods can penetrate only if there is present another cation on the other side of the membrane, so that an exchange can take place. That is exactly the same as with the red blood corpuscles, where from the beginning the demon- stration of the selective anion-permeability depended upon the fact that as long as there exist differences of concentration in the proper 6 Michaelis, Naturwissenschaften 1926, 14: 33. STRUCTURE OF THE PLASMA MEMBRANE direction, the anions of the surrounding solution can be exchanged against the anions of the interior of the cell. And now the question arises, how it is to be understood, that in the case of the collodion membrane the pore permeability is limited to the cations and, in the case of the red blood corpuscles, to the anions. Michaelis had already turned his mind to the fact that the substance of the cation-permeable membrane itself is negatively charged, and he connected this idea with the well-known membrane studies of Bethe and Toropoff 7 and the experiments on the reversal of membrane potentials in gelatine discs, which have been established by Matsuo in my own laboratory.8 As a matter of fact it can be proved that this idea is right. There exists a relation between the electric charge of the membrane material and the faculty of the ions with opposite electric charge to pass. In my laboratory Mond succeeded in demonstrating that if the negative charge of the collodion is changed to a positive charge by addition of a basic dye, for instance by rhodamin, the membrane thus formed, instead of being exclusively cation-permeable is changed into a membrane of selective anion- permeability.9 Table III illustrates the resulting conditions. TABLE III Membrane Potentials in Rhodamin-collodion Membranes — Cl 0.1M NaCl ^~ 0.1M NaCl 0.1M NaCl 0.1M NaCl 0.1M NaCl 0.1M NaCl SCN > I > Br > Cl > SO4 Cations without effect The dotted arrows show the direction of the movement of the chlorine ions; their length is a measure of the potential dependent on the velocity of the ions. The arrows drawn refer in a corresponding manner to the anions of the opposite side of the membrane. The electromotive forces decrease from -f- 60 millivolts to - - 3.8 millivolts along the series of anions: thiocyanate, iodine, bromine, chlorine, sulfate. The cations are without any effect. The membrane potential is therefore approximately zero if there is sodium chloride and 7 Bethe and Toropoff, Zeitschr. f. physik. Chemie, 88, 686, 1914 and 89, 597, 0.1M NaSCX + 60 millivolts 0.1M Nal + 33 0.1M NaBr + 20 0.1M NaCl 0 0.1M Na2SO4 - 3.8 0.1M KC1 + 2 1915. 8 Matsuo, Pfliiger's Arch., 200, 232, 1923. 9 Mond and Hoffmann, Pfliiger's Arch., 220, 194, 1928. RUDOLPH HOBER potassium chloride in the same concentration on each side of the membrane. There can be no doubt that these experiments demonstrate on the one hand in a very conclusive manner the existence of anion- permeability on a membrane originally cation-permeable, but they reveal on the other hand some difficulties in our understanding of these and, as we shall see, of other alterations of the ion-permeable membranes. The membrane potential of the rhodamin collodion membrane does not increase with increasing migration velocity of the effective ion, as has been found by Michaelis with the cation- permeable collodion membrane, but the potential changes according to the lyotropic series. This seems to point to some kind of relation of ion-permeability to the colloidal state of the membrane, which is known to depend in an especially characteristic manner on the lyotropic properties of the ions. However, before discussing this question more amply, we will look at a remarkable consequence of the membrane experiments just described. Mond, supposing that the membrane material of the red blood corpuscles is electropositive, suggested that their natural anion- permeability might be turned into cation-permeability, if one succeeds in giving the membrane substance a negative charge.10 This actually happens by the addition of a suitable amount of hydroxyl ions. As soon as the reaction in the surrounding medium of the blood corpuscles is made more alkaline than pH 8, the usual selective anion-permeability is displaced by selective cation-permeability, so that now an exchange between the potassium ions of the interior with the sodium ions of the environmental solution begins, while the chlorine and bicarbonate ions present in both serum and corpuscles, which were up to this point able to pass through, are now fixed. Mond has advocated the view that the decisive constituent of the plasma membrane, to which the opposite charge is to be attributed, has ampholyte character and might be globine, that is, a protein body, because the reaction by which this reversal of anion-permeability into cation-permeability takes place conforms with the isoelectric point of the globine, which is pH 8.1. In this way we come to a conception, similar to the well-known hypothesis of Nathansohn, that the cell surface is comparable to a mosaic of both lipoids and proteins. Apparently the plasma mem- brane of the red blood corpuscles consists of at least two constituents, a lipoid phase, whose existence enables the lipoid-soluble substances to enter, and a protein phase, which is pore-permeable, so that water 10 Mond, Pflilger's Arch., 217, 618, 1927. STRUCTURE OF THE PLASMA MEMBRANE as well as dissolved substances, whose molecular size is small enough, can pass through. As to the character of the structure of the cation- permeable membranes, which we will now discuss, the opinion is not yet substantiated enough. TABLE IV Resting Potentials of the Sciatic Nerve Time Potential Solution Time Potential Solution millivolts millii'ctis 3:21 20.6 Ringer 3:20 27 .4 Ringer 3:53 20.5 Ringer 3:52 27.5 Ringer 3:55 Ringer with 0.08 % KC1 3:54 Ringer with 0.08% KC1 + 0.1%CaCl2 4:03 19.0 4:02 27.7 ( i 4:28 17.2 4:27 28.5 i ( 5:18 16.5 5:17 28.6 1 1 It seems that the cation-permeable membranes exist more fre- quently than anion-permeable membranes. As Bernstein and I have pointed out twenty-five years ago, the hypothesis of selective cation- permeability gives a good explanation of the electro-negativity that results from injury and of negativity resulting from activity. But if we produce a difference of potential in uninjured tissues such as muscles, liver and apple by joining their surfaces at two different points with two different salt solutions, then we see that, although the cations are better enabled to produce an electric current (ap- parently in connection with the negative character of the membrane colloids), the anions have an action too, the strength of which is correlated with their position in the lyotropic series: sulfate, chlorine, bromine, nitrate, iodine, thiocyanate. But the cations also do not act exactly according to the series of the size of the ions: csesium, rubidium, potassium, sodium, lithium, but succeed each other as potassium, rubidium, csesium, sodium, lithium, — a series met with rather often, as I have found in relation to changing the hydrophilic colloidal state, and characterized by the peculiar dislocation of caesium. And finally, regarding the membrane potentials of muscle and nerve, we encounter a cation antagonism, for example that between potassium and calcium, which might be explained by assuming an influence on the state of membrane colloids, whereas it is difficult to explain it by supposing the existence of a sieve-like membrane, the size of whose pores remains unchangeable. Table IV gives as an example the behavior of the sciatic nerve of a frog.11 11 Hober and Strobe, P finger's Arch., 222, 71, 1929. 10 RUDOLPH HOBER As the experiment on the left side demonstrates, the resting potential falls if the uninjured surface of the nerve is brought into contact with a Ringer's solution in which the percentage of potassium chloride is raised to more than 0.08. If we increase not only the con- centration of potassium chloride but also of calcium chloride to 0.01, the alteration of the initial potential, as is to be seen in the experiment on the right side of the figure, does not take place. So we notice again in regard to the membrane potentials the well-known antagonism between potassium and calcium, and since there is hardly any doubt that the permeability of the plasma membrane due to its porous structure plays a significant role, we concluded that this permeability is, according to the nature of the composing material, much more variable than the pore-permeability of the artificial ion-permeable membranes, especially of the collodion membranes. Before leaving the interesting question of ion-permeability, I wish to direct your attention to a membrane with very curious qualities. Last year I set up and examined a membrane which was a patch- work of cation-permeable pieces of collodion and anion-permeable pieces of rhodamin collodion.12 Figure 1 gives the scheme. This membrane must have the following qualities, and in fact it does have them. If we place a salt solution on one side of it, for instance, a solution of potassium chloride, and on the other side water, the salt cannot diffuse into the water, although the membrane is as permeable for the potassium ions as for the chlorine ions, because a passage in chemically detectable quantity would be possible only if it could happen at just the same place in equivalent amounts of cation and anion, or in other words, because one ion can move only at an infinitesimal distance from the opposite. However, the passage of the potassium chloride is rendered possible as soon as a salt, whose ions can interchange through the membrane with the potassium and the chlorine ions, is placed on the other side of the membrane. It seems to me that membranes of this kind, which, in spite of their permeability for anion and cation, are able to entirely prevent the escape of salts, have been realized by nature and play an important role. Now the question arises as to whether, in addition to the water, only inorganic ions take the way through the pores of the plasma mem- brane. Logically the answer is no. For if there are molecules whose volume is of the same order as that of the permeating ions, then they naturally must take the way through the pores, regardless of the possibility of their passing equally well through the membrane by 12 Hober and Hoffmann, Pfliiger's Arch., 220, 558, 1928. STRUCTURE OF THE PLASMA MEMBRANE 11 selective solution. Of course it has been pointed out by'Michaelis that the collodion membrane, if it is dried enough to establish selective ion- permeability and therefore to give the maximum electromotive effect, allows those molecules to pass whose diameter is about the same as that KCI 14-1-M-l water NaBr FIG. 1. of glucose.13 A similar behavior is met with in the plasma membranes. It will be noted that among the organic non-electrolytes entering into the cell, there are some which permeate more quickly than might be expected in relation to their relative lipoid-solubility, or, more cor- rectly, in respect to their relative solubility in ether, supposing that the relative solubility in ether is to be acknowledged as a likely measure of the physiological phenomenon. Some of these substances are characterized by a relatively small molecular volume, for instance, ethylene glycol and glycerol. Therefore Collander may be quite right in considering their comparatively rapid permeation into plant cells as due to the porosity of the plasma membrane.14 In other cases where a disagreement occurs between velocity of penetration and solubility in ether, for example with urea and its derivatives, even the view of a sieve-like property fails to overcome the difficulties. But here we can see, as I have found with Watzadse, that the difficulties will be removed if, instead of the solubility in ether, the solubility in the previously mentioned oil mixture of Nirenstein will be correlated with the physiological phenomenon.15 13 Loc. cit. 14 Loc. cit. 15 Watzadse, Pfliiger's Arch., 222, 640, 1929. 12 RUDOLPH HOBER The assumption of the porosity of the plasma membrane in this manner being justified in several ways, it will be necessary to study as intimately as possible the properties of the artificial porous mem- branes and especially, because of their great stability, those of dried collodion membranes. Therefore perhaps it is not too audacious to consider the possibility, in relation to physiological conditions, that certain molecules with a diameter not too great and not too small might be stopped in the pores and obstruct them in the same manner that ultramicroscopic particles are not only kept back by an ultrafilter, but finally also obstruct its pores. FIG. 2. Diffusion of thiocyanate in 15', retarded by urethanes. From this point of view Anselmino has made experiments in my laboratory. He favored the obstruction of the pores by using narcotics, because they can be adsorbed by the collodion.16 The result was that the collodion membrane was obstructed to such a degree that the osmotic movement of water as well as the diffusion of molecules of small size was strongly retarded. Figure 2 reproduces a striking experiment. You see that the diffusion of thiocyanate is reversibly slowed by several urethanes, and that the homologous urethanes exert their influence characteristically so that the longer their carbon chains are, the smaller their limiting concentration will be, in the same way that we usually observe in narcosis. 16 Anselmino, Pfliiger's Arch., 220, 524, 1928. STRUCTURE OF THE PLASMA MEMBRANE 13 It will be necessary to find out still more exactly which substances are suitable for the obstruction of the pores and which are not. Michaelis has recently found that the speed of diffusion of glucose through a collodion membrane of suitable pore size decreases with time more and more, and he regards this too as effectuated by an obstruction of the pores.17 But in this case we do not have to deal with an adsorb- able substance. It still remains an open question for the future, how far the decrease of cell-permeability during narcosis, so often already observed, is to be attributed to the porosity of the plasma mem- brane. If this really happens, our ideas as to the nature of narcosis would be greatly supplemented. Now we will consider an especially difficult matter. As we have seen, the entrance of numerous organic non-electrolytes into the living cell may be considered as a matter of lipoid-solubility; the entrance of other organic non-electrolytes, of some ions and of water may be considered as a matter of diffusion through the pores of the plasma membrane. But there is a group of substances of a very remarkable physiological significance, which can neither directly enter by dissolving in the oily phase nor by migrating through the porous phase, but which, nevertheless, do obviously enter. To this group belong substances which constitute a considerable proportion of the nutritive material, such as many sugars and amino-acids. There can be no doubt that this passage is not merely a simple form of permeation, in the sense that it depends on a certain permanent and invariable physicochemical behavior of a membrane. Either the plasma mem- brane must change under definite conditions in such a way that a temporary removal of the barrier to diffusion is brought about, or, — as has often been supposed, — reversible chemical reactions of the food- stuffs occur even in the surface of the cell, so that either the products of reaction are enabled to pass through or a more or less complicated series of single reactions is terminated by the appearance of the food- stuffs inside the cell wall. Adhering to the physicochemical character of this lecture, we will discuss, basing our remarks on experimental data, only one of these forms of the ingestion of the nutritive substances, namely the alteration of the plasma membrane in such a wray that for a short time it becomes permeable to substances for which it is otherwise not permeable. As a matter of fact, there is a well-known form of intake of nutritive material which can be considered as an opening of the plasma mem- brane, that is, phagocytosis. For as the protoplasm is flowing around the particles of food in order to incorporate them, the superficial layer 17 Michaelis and Weech, Jour, of Gen. Physiol, 12, 55, 1928. 14 RUDOLPH HOBER necessarily must be partly destroyed. On the other side there exist further conditions for naturally opening the plasma membrane in a reversible manner, particularly as a so-called functional increase of permeability, that is, as an increase of permeability accomplished by function, or better, by excitation as preparation for function. I am not able to give an extended review of our knowledge of functional increase of permeability within the limits of this lecture, but I shall relate one striking demonstration of the bringing about of a reversible increase of permeability. If one brings Spirogyra cells into a solution of cyanol, a well-diffusing blue sulfonic acid dye, the protoplasts will remain unstained for several weeks. Some years ago in my laboratory Banus observed that while sending an alternating current of appropri- ate strength through the threads of algae, the blue dye would pass out of the solution into the interior of the cell, namely into the sap of the vacuoles.18 After this, the current being stopped, the algae were left for some time in the blue solution; then they were taken out and washed with pure water. It resulted that the vacuole retained the blue dye in spite of its diffusibility, the dye which entered being imprisoned as long as the cell was alive. Apparently the electric current had opened the plasma membrane, a substance to which the interior of the cell is closed under natural conditions had penetrated, and behind it the plasma membrane had shut up. In this way an event was produced, owing to experimental conditions, that is never realized in nature; but a natural phenomenon, the reversible increase of permeability, had been reproduced, possibly in a somewhat crude manner. Perhaps there occurred only a regeneration after an injury generated by the current. But, examining the conditions more closely, we may recognize that nature may sometimes duplicate them. For, in regard to the well-known studies of Bethe and ToropofT on gelatine diaphragms, it is highly probable that the flow of an electric current is accompanied by changes of hydrogen and hydrox}^ ion concentration on the cell boundary so that these active ions, either by hydration and liquefaction or by aggregation of the surface colloids, can amplify or narrow the paths to be taken by diffusing substances and can in this way produce reversible changes in permeability. Thus we learn more and more to regard the plasma membrane as a formation with varying properties so that its permeability exhibits different degrees succeeding one another in time. But the plasma membrane does not only vary in one and the same object temporarily, but, — and this shall be the last point to be discussed in this lecture, - it varies also in one and the same kind of cell from species to species. 18 Banus, Pfliiger's Arch., 202, 184, 1924. STRUCTURE OF THE PLASMA MEMBRANE 15 I shall only demonstrate this with one especially simple object, namely, the red blood corpuscles again, and with this object I wish to demon- strate further in what direction the research into the nature of cell permeability is to be extended. Finally I return in this way once more to the phenomena of porous permeability and of solution per- meability of the cells. As we have seen before, the limiting membrane of the blood corpuscles, according to the electropositive charge in the wall of its pores, allows only the anions to exchange by diffusion from one side of the membrane to the other. Further, it has been pointed out by different authors that each anion passes through the corpuscle mem- brane with a specific velocity. Now Mond in my laboratory has raised the question of the existence of differences in the relative velocities from species to species as evolving from the different sizes of the holes in the sieve-like membrane, and in order to decide this question, he examined the exchange of chlorine ions against sulfate ions, which are known to wander especially slowly.19 Mond actually found considerable differences in the different animals. The inter- change is quickest in the blood corpuscles of man, then there follow pig, horse, cattle. The conclusion that we have to come to in the experiment just described with differences in pore size has been supported by Mond by comparing the sulfate ion with the tetrahydric alcohol erythritol as a non-electrolyte which is insoluble in lipoids and which is known to penetrate into the blood corpuscles and other cells as slowly as the sulfate ion. The same result occurred, namely, the speed of permeation was greatest with the corpuscles of man and the least with those of cattle. But there exist not only differences from animal to animal in the porous permeability of the cells; the same state of affairs holds for the solution permeability. It is well-known that almost every basic dye enters the living cell, but there are rather few acid dye-stuffs that are suitable to it. As to the sulfonic acid dyes, evidently only those enter which dissolve in the oil mixture worked out by Nirenstein and, as has been illustrated by a table in the beginning of my lecture, the dyes enter the cells the more as their relative solubility in the oil mixture is greater. Now I have discovered that the partition coefficient of blood corpuscles to surrounding solution differs under the same conditions from one species to the other; for example, the coefficient is greater with the blood corpuscles of the pig than with those of cattle and sheep, and with these greater than with those of the horse.20 This is demonstrated for two dyes in Table Y. 19 Mond and Gertz, Pfl tiger's Arch., 221, 623, 1929. 20 Unpublished experiments. 16 RUDOLPH HOBER It appears at once that two explanations may be attempted: either we have to assume that the blood corpuscles of all four animals contain lipoids of the same quality, on which the dyes are distributed, but the quantity is greatest in the corpuscles of the pig and is smallest in the corpuscles of the horse; or we have to do with nearly the same quantity of the lipoids in every kind of corpuscle, but the lipoids differ qualitatively as to their power to dissolve dyes, the power being greatest with the pig and smallest with the horse. It is my opinion that we must prefer the second explanation; for whenever the passage of the dyes is dependent here upon the lipoid solubility,— and unquestionably this is the case, — then we must expect that a dye- stuff penetrating into the blood corpuscles of the pig will also get through the corpuscles of the horse, even if their lipoid phase is very small; but I have found that, on the contrary, the corpuscles of the horse are nearly impermeable to several of the staining sub- stances examined. Thus we conclude that not only the properties of the porous phase of the cell boundary, but also its dissolving properties, vary from animal to animal. TABLE V Partition of Dyestuffs to Blood Corpuscles Kind of Corpuscle Dye Initial Con- centration Final Con- centration Partition Coefficient Horse Tropsolin 1 0.0025 0.0017 1.9 Cattle Tropsolin 1 0.0025 0.0015 2.7 Pie Tropseolin 1 0.0025 0.0013 3.7 Horse Bromophenol blue 0.0025 0.0022 0.55 Cattle Bromophenol blue 0.0025 0.0020 1.0 Pig Bromophenol blue 0.0025 0.0015 2.7 Ladies and gentlemen, I have come to the end and I shall repeat now the previously quoted words of Professor Jacobs: "It may be emphasized that what is most needed in the field of cell permeability at the present day is facts." And in relation to that I wish to add to what I have already said to you a quotation from Professor Ralph Lillie's lecture concerning the scientific view of life. He said: 'What is required is the imagination or construction of some model that will reproduce in intelligible form the essential features of the phenomenon under consideration. Intelligibility is the essential criterion of the scientific view; it aims at making phenomena intellectually compre- hensible." We perceive better than anywhere else the striking ad- vantage of using models in the development of our knowledge of cell permeability, beginning with Traube and Pfeffer and passing from STRUCTURE OF THE PLASMA MEMBRANE 17 Overton to Collander and Michaelis. It is very peculiar that in this direction the physical chemists have realized almost nothing from these weary, but very fascinating and instructive studies of what is required as a model of the cell membrane. This is to their own dis- advantage, I believe, because they here overlooked fundamental problems worthy of pursuit by the methods of exact science which are applicable to the membranes, the qualities of which have been discussed in this lecture. Under these circumstances the physiologist is constrained, and will be constrained still more in the future, to leave his proper work, as he must for a shorter or even for a longer time put away physiology, and become pure physicist or pure physical chemist in order to answer preliminary questions of great importance to physiology. Otherwise he will be open to the great danger of fabri- cating hypotheses. But whoever among the physiologists resolves to leave his physiological studies, he may encourage himself by re- membering that it was Jacques Loeb who, feeling obliged to do so in regard to his science and to himself, created in the last years of his life a monumental work on the physical chemistry of the protein bodies. In this way he manifested anew his perseverance and his enthusiasm, both properties which distinguished also Reynold Spaeth, whose memory is with us today. THE COPPER CONTENT AND THE MINIMAL MOLECULAR WEIGHT OF THE HEMOCYANINS OF BUSYCON CANALICULATUM AND OF LOLIGO PEALEI HUGH MONTGOMERY (From the Department of Physiology, Harvard Medical School, Boston, and the Marine Biological Laboratory, Woods Hole) The view, originally put forward by Fredericq (1878), that copper is a normal constituent of hemocyanin and that it has a significance in the respiratory function of this protein similar to that of iron in hemoglobin has been substantiated by later investigations, particularly those of Begemann (1924) and Redfield, Coolidge and Montgomery (1928), which show that the combining ratio of copper to oxygen is the same in the blood of a large number of invertebrates. A knowledge of the quantity of copper in hemocyanin consequently provides significant information with regard to its respiratory function. Inas- much as the amount of copper in the various hemocyanins does not appear to be the same, such data gives unequivocal evidence of the specific character of the respiratory pigments in the different groups of invertebrates. Furthermore, because of the very small number of copper atoms in the hemocyanin molecule, the copper content is a most valuable basis from which to estimate the minimal molecular weights of these proteins. In this paper an investigation of the hemocyanin of the whelk, Busy con canaliculatum, and of the squid, Loligo pealei, is described. Mendel and Bradley (1906) studied the respiratory protein of the blood of the whelk, which they called hemosycotypin, — a name derived from the then current generic name of this form, Sycotypus. They report that it contained zinc as well as copper.1 They concluded that copper composed only 0.043 per cent of the weight of the molecule, a value very much smaller than that obtained in the case of other hemocyanins and one which leads to very high estimates of the protein content of the blood when the oxygen capacities demonstrated by 1 It seems preferable to include "hemosycotypin" among the hemocyanins because it has been demonstrated that the combining ratios of copper and oxygen are the same in this case as in that of other hemocyanins and because recent obser- vations in this laboratory appear to make it doubtful whether the zinc is a true constituent of the protein molecule. Inasmuch as specific differences appear to exist between the hemocyanins of different groups of animals, confusion will be apt to result if each hemocyanin is given a different specific name. 18 COPPER CONTENT OF HEMOCYANIN 19 Redfield, Coolidge and Montgomery (1928) are taken into account. The copper content of the hemocyanin of the squid does not appear to have been previously examined. The copper content of these hemocyanins has been determined on material purified according to several standard procedures applicable to protein substances. Analyses for copper were made by the method described by Redfield, Coolidge and Shotts (1928). Between 10 and 20 c.c. of the hemocyanin solutions were used in each sample. The samples were dried in an oven at 100-110° C. for 48 hours, cooled in a dessicator and weighed. This procedure was repeated daily until successive weights did not vary more than i mgm. The samples of dried hemocyanin weighed between 100 and 300 mgm. Digestion, the electrolytic separation of copper, and its estimation were carried out exactly as described, except that in the titration 15 drops of potassium iodide were used instead of 10, as this modification was found to sharpen the end point. We have not succeeded in producing definitely crystalline prepara- tions of the hemocyanin of Busycon canaliculatum by methods which have been found applicable in other cases. Dhere, Baumeler and Schneider (1929) have also been unsuccessful in crystallizing this hemocyanin. However, on prolonged dialysis against distilled water a precipitate is formed which appears to be composed of short rods and which gives a silky sheen on shaking similar to that characteristic of crystalline protein preparations.2 Busycon hemocyanin appears to be a globulin, as it is insoluble in the region of its isoelectric point in salt solutions of sufficient dilution. This property has been used in purifying our material as well as the usual procedure of salting out with ammonium sulphate, employed by Redfield, Coolidge and Shotts (1928) in the preparation of Limulus hemocyanin. 2 In an attempt to produce crystals, a number of preparations of hemocyanin, all of which showed a silky sheen on shaking, have been made by different methods from several species. The precipitated particles were too small, however, to be recognized under the microscope as definite crystals, though a very fine rod shape was observed in many cases. By the addition of 2 drops of serum to 1-2.5 c.c. of 0.05M acetate buffer solution of pH 4 to pH 5, the hemocyanins of Busycon canalicu- latum and of Busycon carica were precipitated and showed a sheen on shaking. In the case of the bloods of the eight different species; Limulus polyphemus (horse- shoe crab), Busycon canaliculatum, Busycon carica, Libinia emarginata (spider crab), Loligo pealei, Homarus americanus (lobster), Callinectes sapidus (blue crab), and Ovalipes ocellatus (lady crab), the hemocyanin was precipitated by diluting the serum 20 to 200 times and adding a few drops of 0.006 per cent acetic acid to 5 c.c. of the diluted serum. The acid must be added slowly or a precipitate will be formed which will show no sheen. Too much acid redissolves the precipitate. In several cases these hemocyanin precipitates were concentrated by centrifuging and redissolved, whereupon the solutions appeared distinctly blue. This color dis- appeared when the solution was reduced with sodium hydrosulfite so that evidently the hemocyanin was not denatured by the process. 20 HUGH MONTGOMERY TABLE I Copper Content of Hemocyanin of Busycon canaliculatum Specimen No. Method of Preparation Dry Weight Copper Copper grams me,m. per cent IVa Three washings at isoelectric point 0.2071 0.496 0.240 0.2053 0.496 0.242 0.2067 0.492 0.238 0.2070 0.482 0.234 IVb Four additional washings at isoelectric 0.1541 0.366 0.237 point 0.1538 0.378 0.245 0.1544 0.371 0.240 0.1554 0.378 0.243 VI Three washings at isoelectric point 0.0914 0.225 0.246 0.0919 0.204 (0.227) 0.0916 0.217 0.238 0.0917 0.230 0.238 0.2694 0.642 0.238 0.2699 0.634 0.235 0.2698 0.633 0.235 0.1816 0.437 0.241 0.1821 0.433 0.239 VII Salting out and dialysis 0.1694 0.440 0.260 0.1680 0.436 0.260 0.1700 0.441 0.260 0.1699 0.443 0.260 0.1693 0.436 0.258 0.1693 0.438 0.258 0.1693 0.434 0.256 0.1693 0.440 0.260 0.1693 0.441 0.260 0.1693 0.434 0.256 VIII Salting out and dialysis under con- 0.1075 0.263 0.242 ditions leading to precipitation 0.2130 0.517 0.242 0.2120 0.530 0.250 X Salting out and dialysis 0.3967 0.948 0.239 0.3948 0.944 0.238 0.3965 0.950 0.238 0.3978 0.948 0.237 0.6492 1.535 0.236 0.6485 1.554 0.240 0.6487 1.534 0.236 XI Salting out and dialysis 0.5494 1.318 0.240 0.5481 1.308 0.238 0.5483 1.309 0.239 0.5478 1.311 0.239 0.5472 1.315 0.240 COPPER CONTENT OF HEMOCYANIN 21 Specimen IVa was made from blood which had been preserved with toluene in the cold room for two weeks. It was diluted with ten times its volume with distilled water and brought into the region of its isoelectric point by the careful addition of 0.01N HC1. The precipitate resulting was separated by centrifuging and put into solution in the original volume of water by the addition of an amount of sodium hydroxide equivalent to the hydrochloric acid previously added. This process was twice repeated. The precipitate finally obtained was washed with distilled water. The final product con- tained only a trace of chloride. Whenever acid or alkali was added, it was run in through a glass tube which had been drawn to a fine point while the hemocyanin was being vigorously stirred. In order to determine whether further purification of this product could be obtained, the entire process of purification was repeated four more times on a portion of Specimen IVa, the resulting preparation being designated Specimen IVb. Specimen VI was made in a manner similar to Specimen IVa. Specimen VII was made from blood which had been preserved half-saturated with ammonium sulphate for a month. The precipitated hemocyanin was separated by centrifuging and dissolved in a large volume of 5 per cent saturated solution of ammonium sulphate. The solution was centrifuged in order that a small amount of insoluble material might be discarded, and the solution was reprecipitated by the addition of saturated ammonium sulphate. This process was repeated twice. The solution was then dialyzed against 0.001N sodium hydroxide under 20 cm. Hg reduced pressure for two weeks, at the end of which time it was free of sulphate. The preparation of Specimen VIII included the same steps as Specimen VII, except that it was dialyzed against 0.001N sodium hydroxide for five weeks at atmospheric pressure. At the end of the fifth week a precipitate appeared in the solution which gave on shaking a silky sheen similar in appearance to that produced by protein crystals. The precipitate consisted of rod-shaped particles about 2 IJL in length. The solution still contained traces of sulphate and was consequently centrifuged and the precipitate washed three times with a large volume of distilled water. The sulphate test was then negative. Specimens X and XI were prepared from material which had been kept over two years precipitated in half saturated ammonium sulphate. They were purified by reprecipitation with ammonium sulphate (pH 8.0), repeated three times, followed by dialysis against 0.0001 sodium hydroxide for 18 days. The preparation and analysis of Specimens X and XI were made by Miss Elizabeth Ingalls. The results of the analyses of these preparations are given in HUGH MONTGOMERY Table I. The copper content obtained in the case of preparations made in the various ways is very nearly the same. This fact may be taken as evidence that fairly pure preparations of the protein have been obtained. The fact that the copper content of Specimen IVb was not materially increased over that in Specimen IVa by additional washing is further evidence for the adequacy of the method of purifi- cation employed. The best representative value of the copper content of Busycon canaliculatum hemocyanin appears to be 0.24 per cent. Specimen VII yields consistent values 0.02 per cent higher than this. Inasmuch as Specimens VIII, X and XI, prepared by the same general method, agree with the general series, it is probable that the high value obtained in the case of Specimen VII should be attributed to some systematic analytical error rather than to superiority in the method of prepa- ration. Two specimens, which were obtained by the dialysis of fresh blood without other attempt at purification, yielded a product which con- tained about 0.22 per cent copper. This material was free of chloride and had the same nitrogen content per unit weight as the others. The result would appear to indicate that another protein may be present in the blood, but that if so, it exists only in small amounts. In the case of Limnlus, the hemocyanin appears to account for about 95 per cent of the protein of the serum. In order to investigate this possibility further an attempt has been made to determine how far the nitrogen content of the blood of Busycon canaliculatum may be accounted for by the hemocyanin contained in it as estimated from the quantity of copper present. The nitrogen content of Specimen X was determined by the Kjeldahl method. Successive analyses yielded 15.6; 15.5; 15.7; 15.5; 15.4; 15.7; mean 15.5 grams nitrogen per 100 grams dry weight. The copper content of Specimen X was 0.238 grams per 100 grams dry weight. One part of copper consequently corresponds to 65.2 parts of nitrogen. Two specimens of blood were analyzed for copper and nitrogen. The first contained 0.074 mgm. copper per c.c. and 4.92 mgm. nitrogen per c.c. From the copper content it may be estimated that it contained 4.84 mgm. nitrogen as hemocyanin. The second specimen of blood contained 0.066 mgm. copper per c.c. and 4.14 mgm. nitrogen per c.c. The hemocyanin concentration as estimated from the copper content would account for 4.3 mgm. nitrogen. It is evident from these measurements that hemocyanin will account approximately for all of the protein nitrogen in Busycon blood. One preparation of the hemocyanin of the allied species, Busycon COPPER CONTENT OF HEMOCYANIN carica, was made. The blood had been preserved in a precipitated condition in half-saturated ammonium sulphate for one year in the cold room. The hemocyanin was separated, purified by the procedure employed in the case of Bnsycon canaliculatum Specimen X. Analysis of the copper content of the purified material yielded the following values: 0.217, 0.235, 0.238 per cent. The copper content of the hemocyanin of this species appears to be approximately the same as that of Busy con canaliculatum. The hemocyanin of the squid, Loligo pealei, may be readily crystallized by methods similar to those first employed by Henze (1901) in preparing crystalline Octopus hemocyanin, and consequently lends itself well to purification. Squid hemocyanin is insoluble in solutions containing high concentrations of ammonium sulphate. It was found that if enough saturated ammonium sulphate solution is added to the blood to form a very slight cloud of precipitated hemocyanin, a fuller precipitation in the form of crystals can then be produced by several procedures designed to decrease the solubility of the hemocyanin in the solution. These were: (1) the careful addition of increasing quantities of ammonium sulphate, (2) increasing the hydrogen ion concentration as in the Hopkins-Pinkus (1898) method of crystallizing albumen, or (3) raising the temperature. These methods can be used with success in combination. Crystallization by raising the temperature, which is presumably due to increasing the "salting out" effect of the ammonium sulphate at the higher tempera- ture is particularly efficacious and has the advantage that it involves the addition of no reagents and may consequently be accomplished slowly so as to favor the formation of crystals. It was found that by raising the temperature from 0° C. to 30° C., a heavier crystalline precipitate is produced than by raising it to room temperature only. A temperature change within a range which will not denature the protein did not crystallize all the hemocyanin that was in the solution. Consequently, the yield may be increased by combining the tempera- ture method with the addition of ammonium sulphate or of acid. When crystallization is produced in this manner, there is formed first a fine precipitate, visible under the microscope but apparently amorphous. This changes in a few minutes to fine rods and then to bundles of needles and finally to large needles. The process is much like that described in the case of Eledone moschata hemocyanin by Robert (1903). The appearance of the crystalline rods is similar to that figured by Dhere (1919, figure 4), in the case of the oxyhemocyanin of Helix pomatia formed in the presence of sodium sulphate. If large excess of reagents are added suddenly, the precipitate produced is 24 HUGH MONTGOMERY amorphous. Crystallization of squid hemocyanin was obtained more readily from fresh blood than from preparations which had been preserved in a precipitated condition in concentrated ammonium sulphate or from previously crystallized hemocyanin. Crystals which had been kept for a year in the cold room in their mother liquor (half saturated ammonium sulphate), were found to have become insoluble in distilled water. This phenomenon was observed by Craifaleanu (1919) i.n the case of crystals of the hemocyanin of Octopus vulgaris. Craifaleanu called this form "para-hemocyanin." TABLE II Copper Content of Hemocyanin of Loligo pealei Specimen No. Method of Preparation Dry Weight Copper Copper grams mgm. per cent I Salting out and dialysis 0.1485 0.384 0.258 0.1486 0.371 0.250 0.1502 0.388 0.257 0.1490 0.376 0.252 II Crystallization and dialysis 0.0785 0.194 0.244 0.1620 0.386 0.238 0.1624 0.390 0.242 V Salting out and dialysis 0.4579 1.155 0.252 0.4594 1.161 0.254 0.4593 1.178 0.256 0.4601 1.159 0.252 0.4592 1.154 0.252 Analyses of the copper content of the hemocyanin of Loligo pealei have been made upon three preparations. Specimens I and V were prepared from blood which had been precipitated by the addition of ammonium sulphate to half saturation and kept in the cold room at about 5° C. for two years. The material had a fishy odor, which dis- appeared when it was shaken with air and from which the final prepa- rations were entirely free. The precipitate was separated from the supernatant fluid with the centrifuge and was dissolved with a small volume of 5 per cent ammonium sulphate. The solution was again centrifuged to throw down any insoluble material, and the fluid was drawn off and reprecipitated by the addition of saturated ammonium sulphate. This process was repeated twice. The solution was finally dialyzed until it was found to be free of sulphate. Specimen II was prepared by crystallization from fresh blood. The blood was chilled to 0°, and then sufficient saturated ammonium sulphate was added to COPPER CONTENT OF HEMOCYANIN 25 produce a very slight precipitation of hemocyanin. The temperature was then raised from 0° to 20°, when full precipitation was obtained. The precipitate was in the form of needle-shaped crystals about ten ju in length. The crystals were separated from the mother liquor by centrifuging and dissolved with 5 per cent saturated ammonium sulphate. Insoluble material was removed by centrifuging, and the hemocyanin was then reprecipitated as before. This second pre- cipitate was not crystalline, however. The preparation was then dialyzed against water until free of ammonium sulphate. All threp preparations had a clear blue-green color and became colorless in the characteristic way upon reduction with sodium hydro-sulphite. Table II contains the data obtained from analyses of these prepa- rations of squid hemocyanin, which all yield values for the quantity of copper in the molecule close to 0.25 per cent. It is interesting to compare the values obtained for the copper content of the hemocyanin of Busycon and Loligo with those previously reported for other species, particularly with regard to their systematic relationships. In Table III are collected the various determinations TABLE III Copper Author Cancer per cent 0.32 Griffiths (1892). Homarus. . . 0.34 1 t Sepia 0.34 U Octopus vulgaris 0.38 Henze (1901). Loligo pealei 0.25 Helix pomatia 0.25 Burdel (1922). 11 U 0.29 Begemann (1924). Busycon canaliculatum . . . Limulus polyphemus 0.24 0.173 Redfield, Coolidge and Shotts (1928). of the copper content of hemocyanin which occur in the literature. It is noteworthy that the value obtained in the case of Busycon canaliculatum and Busycon carica does not differ greatly from those attributed to the other gastropod, Helix pomatia. The value obtained for Helix pomatia by Begemann, whose method of copper analysis we have employed, exceeds the value obtained with Busycon by an amount well in excess of the apparent experimental errors. These hemocyanins appear also to differ in certain other respects. Busycon hemocyanin cannot be crystallized by methods which succeed in the case of Helix (Dhere, Baumeler and Schneider, 1929). Busycon hemocyanin is insoluble in the region of its isoelectric point in the presence of quite ^V08 "^ /<£, * ^»^ »\^ (uui LIBRARY 1C — ; 26 HUGH MONTGOMERY appreciable amounts of salt. Helix hemocyanin, on the other hand, appears to be readily dissolved by very small concentrations of salt under these circumstances (Svedberg and Heyroth, 1929). It is surprising that such a great difference exists between the copper content of the hemocyanin of the squid and that of the octopus. Inasmuch as the properties of the respiratory pigments in these two cephalopods appear to be very similar, we believe it to be desirable to redetermine these values by methods of preparation and analysis which are strictly comparable. The weight of hemocyanin containing one atom of copper is given by dividing the atomic weight of copper, 63.57, by the fraction of the weight of hemocyanin due to this element. In the case of Busycon canaliculatum this fraction is 0.25 X 10~2. The minimal molecular weight of Busycon hemocyanin thus appears to be approximately 26,500, when estimated upon the basis of its copper content. It has been shown, however, by Redfield, Coolidge and Montgomery (1928), that when hemocyanin becomes associated with oxygen to form oxy hemocyanin, one molecule of oxygen is combined with a quantity of hemocyanin containing two atoms of copper. Inasmuch as it appears highly unlikely that the oxygen molecule is dissociated into its constituent atoms in its reaction with the respiratory protein, it seems safe to assume that each molecule of oxyhemocyanin is com- bined with not less than one molecule of oxygen. The hemocyanin molecule must consequently contain at least two atoms of copper. Estimated on this basis, the minimal molecular weight of Busycon hemocyanin is approximately 53,000. In a similar way it may be calculated that the minimal molecular weight of the hemocyanin of Loligo pealei, estimated on the basis of its copper content, is 25,400, and when the oxygen-combining relations are taken into account, the combining weight appears to be approximately 51,000. SUMMARY The hemocyanin of Busycon canaliculatum contains 0.24 per cent of copper and 15.8 per cent of nitrogen. Its minimal molecular weight is approximately 53,000. The copper content of the hemocyanin of Busycon carica appears to be the same. The hemocyanin of Loligo pealei contains 0.25 per cent of copper and has a minimal molecular weight of approximately 51,000. COPPER CONTENT OF HEMOCYANIN REFERENCES BEGEMANN, H., 1924. Over de ademhalingsfunctie van haemocyanine, thesis, Utrecht; for abstract see Jordan, H., 1925. Zeitschr.f. vergl. Physiol., 2: 381. BURDEL, A., 1922. Contribution a 1'etude des hemocyanines, thesis, Fribourg. CRAIFALEANU, A., 1919. Boll. Soc. Natur. Napoli, Anno 32: 88. DHERE, C., 1919. Jour, physiol. et path, gen., 18: 503. DHERE, C., BAUMELER, C., AND SCHNEIDER, A., 1929. Compt. rend. Soc. de biol., 101: 759. FREDERICQ, L., 1878. Arch, de Zool. esp. et gen., 7: 535. GRIFFITHS, A. B., 1892. Compt. rend. Acad., 114: 496. HENZE, M., 1901. Zeitschr. physiol. Chem., 33: 370. HOPKINS, F. G., AND PINKUS, S. N., 1898. Jour. Physiol., 23: 130. ROBERT, R., 1903. Arch. f. ges. Physiol., 98: 411. MENDEL, L. B., AND BRADLEY, H. C., 1906. Am. Jour. Physiol., 17: 167. REDFIELD, A. C., COOLIDGE, T., AND MONTGOMERY, H., 1928. Jour. Biol. Chem., 76: 197. REDFIELD, A. C., COOLIDGE, T., AND SHOTTS, M., 1928. Jour. Biol. Chem., 76: 185. SVEDBERG, T., AND HEYROTH, F. F., 1929. Jour. Am. Chem. Soc., 51: 539. GROWTH AND DIFFERENTIATION OF THE COLONIES OF ZOOTHAMNIUM ALTERNANS (CLAP. AND LACHM.) E. FAURE-FREMIET COLLEGE DE FRANCE, PARIS INTRODUCTION In a preceding publication (1922) I have insisted on the fact that colonial Vorticellidse constitute an intermediary step between a popula- tion of like cells (cultures of free Infusoria) and a multicellular organ- ism; unlike free cells with unlimited power of division, whose population growth theoretically follows a geometrical progression. The col- onies of Epistylis, of Carches'mm, or of Zoothamnmm generally have a limited growth, following a special cycle, independent of a possible sexual cycle. In these colonies the lineage of each cell is perfectly de- nned by dichotomous ramifications of a common peduncle, and it is pos- sible to show in a large number of cases the existence of somewhat dif- ferential divisions giving two sister cells whose power of multiplication is different. In certain species (Eplstylis arenicolce, Epistylis Perrieri} the first divisions can be dichotomous and equal, so that the mass growth of a number of individuals follows a geometrical progression ; but soon the sister cells resulting from each division multiply unequally, and the growth approaches more or less an arithmetical progression. On the other hand, the study of the growth of the common peduncle, which is considered as a product of the protoplasmic activity, shows that the latter may decrease in course of time. But the Vorticellida* colonies form, from time to time, migrating individuals which may be of the same size as the other individuals (Carchesium, Epistylis} or more voluminous (some Epistylis, some Zoothamniuni, called heteromorphic). In these individuals, and in these only, appear secretory granules already observed by Engelmann and more recently (1926) by Wesenberg-Lund, which seem to be connected with the formation of the peduncle, and one can consider the hypothesis of an active substance, or of a transformable substance, produced in a definite quantity and periodically, by certain individuals, which is divided among the descendants of the latter and at the same time is diminished little by little. It appears then that the growth of a group of cells may be limited by 28 ZOOTHAMNIUM ALTERNANS 29 factors somewhat internal but altogether independent of the hypothetical notion of a " factor of senescence " 1 The Zoothamnium called heteromorphic, about which I have given some detail in my paper of 1922, seems to give the most typical ex- amples as to the role of these internal, or properly cellular factors, in the general form of growth of a colony and its limitation. Claparede and Lachmann described in 1858 a marine species, Zoothamnium altcrnans (described later by Mobius under the name of Z. Cienkowskii) ; the aspect of the colonies, they say, is that of " un arbre a branches courtes et tres regulierement alternantes. La forme de ces families a sa cause dans un arret de division spontanee qui frappe en general 1'un des deux individus issus de chaque division. Lorsqu'un individu A se divise en deux individus B et B1 ', 1'un des deux, B par exemple, ne se forme qu'un pedoncule fort court et son developpement reste stationnaire a partir de ce moment, tandis que 1'autre, B', secrete un pedoncule plus long, puis se divise en deux nouveaux individus, C et C', dont le premier, qui est tou jours du cote de la branche opposee a celui ou se trouvait 1'individu B, ne forme qu'un pedoncule tres court et ne se divise pas davantage tandis que C' forme un pedoncule plus long et se divise en deux individus D et D' et ainsi de suite." That is not all ; in Z. altcrnans and in Z. arbuscula Ehrb. or Z. genlculatwn Ayrton (see Wesenberg-Lund, 1925, and Furssenko, 1925) the migrating individuals which' will be the origin of new colonies and will thus begin a new cycle, are distinguished not only by a few mor- phological characters, but also by their voluminous size and the well- determined place where they originate in the colony, generally at the junction of the main branches. These large migrating individuals are the " ciliospores " of Wesenberg-Lund or " macrozoides " of Furssenko, much larger than the " trophozoides " or " microzoides " which consti- tute the most numerous individuals of the colony. Ehrenberg had observed these individuals in Z. arbuscula, and had noticed that they result from the growth of an individual not unlike the others, but always situated at the junction of a branch. This author admits that one of the two individuals issued from a bipartition on the branch while the other grows without dividing, thus being, he says " the aunt " of the individuals of the branch. Claparede and Lachmann find this same condition in Z. altcrnans, but sometimes this growing indi- 1 In other publications (1925-26) I tried to show that in several very different cases the idea of a factor of senescence could be replaced either by the hypothesis of differing speeds in a group of transformations necessary to cellular activity, or by the assumption of a "probability" of transformation which would be too long to develop here. (See Faure-Fremiet and Laura Kaufman, 1928, and Faure- Fremiet and H. Garrault, 1928.) 30 E. FAURfi-FREMIET vidual may undergo a division. Zootliainnimn alternans (Claparede and Lachmann) is found frequently on the coasts of Brittany ; I have found it in abundance in Woods Hole and was able to follow the dif- ferent stages of the colony cycle and of the formation of the " cilio- spores." I observed a few phenomena of conjugation, quite sporadic, but I have not observed a sexual cycle analogous to the one discovered by Wesenberg-Lund in Z. gcniculatum or described by Furssenko in Z. arbuscula. TECHNIC In order to follow the complete evolution in a large number of colonies, I have used numbered slides, ruled in squares with a diamond point. These slides were first placed in a crystallization dish containing numerous colonies of Z. altcrnans. After several hours, they were re- moved and placed in a Petri dish containing sea water and examined under a binocular microscope. All individuals recently attached were carefully located and designated in numeral order ; those whose peduncle had already developed or had already given the first division were removed with a needle. After this operation, the slides were placed vertically on frames floating in an aquarium through which ran a strong current of sea water ; this was done to avoid the deposit of particles and of microorgan- isms. The slides were then examined periodically and the different stages of the development of each colony were carefully recorded in function of time. When the cytological examination of a colony is necessary, it is always easy to detach this colony with a fine pipette, in order to study it under the high power, in vivo, or after fixation. The best technic for the study of the nuclear apparatus is the fixation by OsO4 for a short time followed by boracic carmine stain. The presence (generally in the Vorticellidae) of a cuticle and the con- tractability of a peduncle constitute two technical difficulties which are not easy to overcome ; it may be necessary to cut the colony with a fine scalpel in order to isolate certain individuals which it is necessary to fix and stain. STRUCTURE OF THE COLONIES The appearance of colonies of Z. altcrnans is very nearly that of a palm (Fig. 1) ; they have a main trunk and oblique branches placed alternately in the same plane, on right and left of the axis; the main trunk always bears at the top a terminal individual of rather large size ; the lateral oblique branches bear a variable number of small individuals ; ZOOTHAMNIUM ALTERNANS 31 finally along the trunk, at the juncture of the lateral branches, are found the voluminous migrating- individuals either macrozooids or macrospores. Trt FIG. 1. A young colony of Zootliaiiniinin altcrnans (Clap, and Lachm.), showing the main trunk and the alternate lateral branches. TM, terminal macro- zooid ; Ci, ciliospores at different stages of growth, located on the anterior side of the colony at the first division of each branch D, E, G, H. The branch F , in this case, bears, at the same place, only two microzooids apparently identical with the others. The lateral branches of the colony observed in extension are almost always slightly curved in, and most of the individuals borne by these branches are inclined toward the outside of the curvature. The two sides of the palm are thus different, and one can define at the same time a base and a summit, an anterior and a posterior side. The elements of symmetry of such a colony are a main axis repre- sented by the trunk, and a median plane, antero-posterior, separating the two halve^ right and left. As for all the other species of the genus Zoothainnhtm, the colonial peduncle bears an elastic tube whose role is passive, and a continuous " cordon central," dichotomically ramified, which represents the pro- 32 E. FAURfi-FREMIET longation of the lower extremity of each individual ; this central cordon has itself a protoplasmic tube (/) limited by a fine film and surrounding a muscular fiber which terminates at the basal part of each individual by a conical group of myonemes.2 The migrating individuals, or " ciliospores," when liberated swim rapidly with their posterior ciliary crown. They are large individuals, flattened in the antero-posterior direction, and look like a top. They attach themselves by means of the scopula (/) and begin to secrete the peduncle. At the same time they lose their posterior ciliary crown and progressively take on again the ordinary subconical form. . 10 [» 100 p Time : 15 m. ZH. 10 H. FIG. 2. Fixation of the ciliospore and construction of the peduncle. At first the top-like ciliospore turns quickly on the slide, then the building of the peduncle begins ; the same individual is shown fifteen minutes after fixation. The ciliary crown slows down and disappears while the peduncle grows (Epistylis stage) dur- ing a short time (two hours) ; finally, one can see the differentiation of the " cor- don central" and the muscular fiber (ten hours). The peduncle is at first a solid cylindrical body of a fibrillar structure which grows rapidly ("Epistylis stage") ; after two hours it reaches 2 For the structure of the Vorticellidse in general, and of the peduncle in par- ticular, see Faure-Fremiet (1906). ZOOTHAMNIUM ALTERNANS a length of about 250 //.. The secretion then begins to slow down and a section of the peduncle is ring-like; there is a central canal, at the bot- tom of which remains attached a part of the body of the infusorian, which from now on will lengthen itself along with the tube of the peduncle and become differentiated in a central cordon with the muscular fiber or " spasmoneme " (Fig. 2). Six or seven hours (at the temperature of 21° C.) after the start of the secretion of the peduncle, the original individual undergoes a first unequal division which gives a macrozooid and a microzooid ; the plane passing through these two zooids and the common peduncle is the median plane of symmetry of the future colony. The large cell remains clearly axial after this first division and continues to form actively the prin- cipal peduncle of the colony. After four to seven hours it undergoes a second unequal division; the interval between the following divisions is longer, from ten to sixteen hours; but always during the growth of the colony the terminal individual is a macrozooid. each division of which separates a microzooid in the median plane of the colony. The succes- sive series of terminal microzooids constitutes a main strain perfectly schematized by the axial trunk of the colony. \\c. shall designate each cell of this series by a Roman numeral representing the division which started it ; we shall have then the origi- nal individual, or ciliospore, then the series of macrozooids, I, II, III, ... X, etc. ^Ye shall designate with capital letters the corresponding series of median microzooids detached from the main strain (microzooids of first order), A, B, C, . . . J, etc. Each branch of the colony is started by the division, alternately at the right and at the left of the median plane of each microzooid of the first order. But, in accordance with the dia- gram of Claparede and Lachmann, only one of the two cells resulting from such a division is the origin of a lateral limb ; we shall designate it by a small letter preceded by the coefficient 1 ; the other cell remains median and will be designated by its capital letter preceded by the same coefficient 1. At the beginning of the formation of the fifth branch, for example, we shall have first the division of the terminal macrozooid IV, which will give a new terminal macrozooid V and a median microzooid E. The latter will divide in a perpendicular plane to that of the division of IV, and will give two individuals, one of which, IE, remains in the median plane while the other, \c, situated for example at the right of this plane, will be the origin of the branch (Fig. 3). Each branch has also a main axis and lateral branches but does not have a well-defined median plane nor median individuals. The division 3 34 E. FAURfi-FREMIET of \c, for instance, gives rise to two cells apparently similar, 2e* and 2c'-. The individual 2el remains in the axis and gives at the new divi- sion 3cl (axial) and 3r2 (lateral) ; 3cl will give 4cl (axial) and 4c2 (lateral), etc. IV FIG. 3. Scheme of the branch E and the basis of the branch F, showing the lineage of the median microzooids IE and \F and the different microzooids. Likewise the individuals 2c2, 3c2, and 4c2 will give successively two or three generations, the elements of which we shall designate by the sym- bols 2c21, 2e22, 3c21, 3e22, etc; according to the rule of Claparede and Lachmann 3c22 does not divide, but 3r21 gives 3c211 and 3e212 ; the num- ber of generations formed by the lateral branches seems to be always rather limited. The median individuals of the second generation: \A, IB, 1C . . . IE, etc., can divide once and give IA1 and 1A2 for example. But while \A, IB, 1C, and their two immediate descendants remain microzooids identical to these designated by the small letters, ID, IE and the follow- ing ones, or the two cells of the second generation, ID1, ID2; IE1, IE2.. etc., undergo a considerable growth and are transformed into ciliospores, or migrating macrozooids, which soon detach themselves from the com- mon trunk to swim freely and to attach themselves later on. It appears clearly then that during the growth of a colony of Zoothanmiuni altcrnans the two cells resulting from the division of one initial cell are never equivalent as to their " potentialities." But in con- firming the observations of Ehrenberg and of Claparede and Lachmann, we may now make them more precise by showing that the progressive segregation of the power of multiplication and of the power of growth is very rigorously tied up with the respective position of the individual separated by the successive divisions. It seems then that a certain ZOOTHAMNIUM ALTERNANS number of divisions at least must be considered as differential divisions. The cytological examination confirms this interpretation. FIRST DIVISION OF THE INITIAL MACRGZOOID The first division is characterized, in a rigorously constant manner, by the unequal division of the macronucleus and of the protoplasm of the initial individual of the colony (Fig. 4) between the first two cells, the macrozooid / and the microzooid A (Fig. 5). A short time before this division, the macronucleus, which takes the shape of a long twisted rod, enlarges at one of its extremities in a compact mass. The other extremity is thin and often flattens slightly, and becomes elongated in the median plane of the individual. The two edges of this flat portion are often slightly thickened, so that a side view gives the impression of a structure in a horseshoe shape. The micronucleus remains near the thick extremity and soon lengthens into a spindle. Meanwhile the peristome and the scopula divide as well as the central cordon of the peduncle and soon an upper and a lower furrow, growing in depth toward each other, begin to separate two cells of very unequal size. The micronucleus completes its own division, then the macronucleus is divided unequally at the time when the two furrows join; the macro- zooid (which remains the terminal individual on the axis of the colony) retains the thickened part of the macronucleus and a micronucleus ; the microzooid (which becomes the first median individual A} retains the thin part of the macronucleus and a micronucleus (Fig. 6). Considering the irregular shapes of the body and of the macro- nucleus in Z. alternans, it is impossible to calculate the corresponding volume and to establish the values of the nucleoplasmic relation. Never- theless, it is clearly evident that the ratio N/P is greater in the micro- zooids than in the macrozooids, i.e., the macronucleus is divided into two daughter cells even more unequally than the cytoplasm. It is difficult to establish whether there exists a difference in com- position between the two unequal extremities of the macronucleus divided between / and A. The " nuclear reaction '' of Feulgen does not show any difference between these two parts, and their structure differs very little. Most frequently one can observe a linear orientation, in a continuous and parallel line of the chromatin granules (microsomes) in the thin part of the macronucleus which will be distributed by the division. On the other hand, the voluminous mass which remains in the macrozooid / shows an irregular distribution of its microsomes. This mass behaves as a chromatin reserve which would not be affected at all by the phenomena of division. Supposing that the terminal condensation of the macronucleus repre- 36 E. FAURE-FREMIET seats a kind of segregation of the chromatin material, we shall describe this first unequal division as a differential quantitative and qualitative division. FIG. 4. Ciliospores at the beginning of the peduncle's formation, showing the appearance of the macronucleus before the first division. FIG. 5. First cleavage of the ciliospore, giving the terminal macrozooid / and the median microzooid A. The figure shows the differential division of the macro- nucleus (figured by dotting) and the apparently equal division of the micronucleus (black — spindle stage). FIG. 6. Later stage of the first cleavage, showing the terminal macrozooid / and the median microzooid A ; macronucleus figured by dotting ; resting micro- nucleus black. FIG. 7. Fourth cleavage on the main strain giving the terminal macrozooid V and the median microzooid E. The qualitative equal division of the macro- nucleus (figured by dotting) is shown. LATER DIVISIONS OF THE INDIVIDUALS OF THE MAIN STRAIN The division of the individuals / and // presents exactly the same differential character as that of the initial individual. It is different at the time of division of the individual ///. In the latter, the macro - nucleus shows at the outset of the repartition a symmetrical thickening at each of its granular extremities which appear entirely homologous. ZOOTHAMNIUM ALTERNANS 37 The median part, finely striated, is divided, however, into two unequal parts by the division of the protoplasmic body, which isolates here again an axial and terminal macrozooid, V. and a median microzooid E (Fi2, and If2 (Fig. 8). On the other hand, the median microzooids, D , E, F, and the follow- ing undergo an unequal division, quantitatively and qualitatively dif- ferential, like that of the first three individuals : the ciliospores / and //. A short time before the division, when the median individual begins to lengthen in the transverse plane, its macronucleus takes the shape of an elliptic blade, presenting in a marginal point a large subspherical thickening. This thick part of the macronucleus, on the other hand, lengthens at the time of division and is divided between the two in- dividuals \D, Id, IE and \c, etc. (Fig. 9). These facts indicate that the differential division takes place at two different times from the fourth generation of the axial cells. For instance, when the division of /// divides into IV and D, the microzooid D has a little less than a half macronucleus ; but this half macronucleus is qualitatively similar to that of the macrozooid IV, having a granular terminal thickening. However, the microzooid D shows a nucleoplasmic relation, a ratio N/P superior to that of macrozooid IV, for the proto- plasm has divided much more unequally than the macronucleus. It is a small individual with a large macronucleus. When the microzooid D divides, the cytoplasmic division is almost equal, but the division of the macronucleus is qualitatively differential, because the thickened and granular part does not divide but goes whole to the median individual ID. The outcome is that the ratio N/P is still increased in this individual. The axial microzooids ID, IE, etc., can undergo a division and give for instance ID1 and ID2; but these two individuals, which remain median, soon begin to enlarge without dividing any further. 38 E. FAURfi-FREMIET The microzooids Id, Ic, etc., as said above, go through a series of divisions which always give individuals with long and slender macro- nuclei. m 8 FIG. 8. Cleavage of the median microzooid B, giving, with equal division of the macronucleus, the microzooids 1&1 and lb-. Comparison between the terminal macrozooid /// and the median microzooid C (resting stage). FIG. 9. Cleavage of the median microzooid D, giving the future median macrozooid ID (ciliospore) and the microzooid Id, with a differential division of the macronucleus. FIG. 10. One median macrozooid (\G for example) at the beginning of its growth, and one microzooid of the corresponding branch. The large difference in size of the macronucleus is to be noted. FIG. 11. Two median macrozooids during the time of growth. In the macro- nucleus, numerous large nucleoli are to be seen (figured as vesicles on the drawing). GROWTH OF THE MEDIAN MICROZOOIDS AND FORMATION OF THE « CILIOSPORES The median microzooids of the fourth generation (D or ID1 and ID2} and of the following generations (E, F, G, etc.) increase rapidly until they reach a length of about 55 p to 70 /A, in one day, two days, or two and a half days. ZOOTHAMNIUM ALTERNANS 39 The macronucleus, already voluminous, begins to grow and forms a very large horseshoe-shaped body. The micronucleus situated at the lower part in a slight depression lengthens into a spindle as in preparation for the division. While the macronucleus increases, rather refringent nucleoli appear in the midst of the chromatic granulations, not giving the reaction of Feulgen (Fig. 10). Soon, while the protoplasmic growth goes on, it seems that the nuclear growth stops. The very numerous nucleoli alone still increase in volume (Fig. 11). Then the outline of the macronucleus disappears, the nucleoli project on the surface of the chromatic mass, and one can observe very numerous stages of disintegration and of degeneration of the macronucleus and of its fragments (Fig. 12). 12 14 15 FIG. 12. Later stage of the median macrozooid's growth. Disintegration and disappearance of the macronucleus. FIG. 13. One median macrozooid almost ready to leave the colony : p, posterior ciliary crown; Ma and Mi, macronucleus and micronucleus of the new nuclear apparatus ; r, residual mass of chromatin. FIG. 14. Top view of a median macrozooid (same stage as that shown in Fig. 12). FIG. IS. Terminal macrozooid making the posterior ciliary crown and soon ready to leave the colony. 40 E. FAUR&-FREMIET Finally, one sees in the center of the cytoplasmic mass containing a rather larger number of residual masses, a short macronucleus, arched, staining very intensely, containing only very small nucleoli, and accom- panied by a resting spherical micronucleus (Figs. 13 and 14). This aspect, frequently observed, is that of a nuclear apparatus of new for- mation, and it is probable that the changes just described represent a phenomenon of endomixis. I was, however, unable to follow in the individuals stained in toto the fate of the spindle-shaped micronucleus observed in the preceding stages. It probably divides and makes up the new nuclear apparatus ; but this stage was not observed in my set of preparations. At the end of the protoplasmic growth and when the nuclear changes are completed, a furrow appears around the median individual, at about the posterior third. It is the future ciliary crown, whose vibratile elements appear soon afterward. At the same time the organism flattens in the antero-posterior direction, and takes the shape of a top. The cytoplasm is filled with diverse inclusions, a great number of which are probably nuclear residue. In the posterior region, above the " scopula," appear very numerous inclusions which are not very refringent. Neutral red in I'tvo colors them a brownish red. These inclusions correspond to the secretion granules whose existence I have already mentioned in the migrating individuals of different Vorticellidse. There are still a few lipoid granules, and, toward the middle of the body, numerous small inclusions fixing neutral red in an intense red color. Iodine fixation gives a mahogany color, but the latter is not any stronger than for the microzooids. The " ciliospore " which has thus been formed becomes almost lens- shaped. The peristome remains closed and the posterior ciliary fringe is animated with active movements which soon determine the liberation of the migrating individual (Fig. 15). GROWTH OF THE COLONIES OF ZOOTHAMNIUM ALTERNANS At a temperature of 21° C., in an aquarium with running water, the growth of the colonies of Z. alter nans goes on very regularly for a period of eight to ten days. Hence it is easy, by periodic examinations of a specific colony, to follow the increase in number of the individuals as a function of time. We have then a measure of the colony's growth. This measure is not very exact, because certain individuals grow without dividing and their mass is clearly larger than that of the others. However, the group of large cells given by the terminal macrozooid and the ciliospores is always rather restricted, and one can admit that the ZOOTHAMNIUM ALTERNANS 41 appearance of the development is rather well represented by the varia- tion in number of the individuals. A more important error may arise from the fact that some parasitic Infusoria (Acineta) very often get into the microzooids (especially the microzooid of the first branch) and multiply in this individual, which does not divide and soon falls off. Because of this, it is necessary at every investigation to trace a total scheme of the colony studied, indicating the place of each individual, which with some practice, may be quickly made by examining the colo- nies in extension in a thin water layer with a low power objective. By this means it is possible to keep an account of the accidental influences ; but when the number of individuals increases too much, beyond the eighth day, for example, this method of pointing becomes very difficult and soon impossible to use with precision. TABLE I Numbers of Colonies Examined Dafp T ' 1 2 3 4 5 6 7 8 9 10 11 12 July 14 11A.M. 0 0 0 0 0 0 0 0 0 0 0 0 15 11A.M. III III III III II III III III II III III 16 12 M. V V V IV IV III IV IV III IV IV 17 4:30 P.M. VII VIII VII V VI V VII VI V VI 18 9A.M. VIII IX IX VII VII VII IX VIII VI 19 10A.M. IX XII VIII VIII VIII XI X 20 11:30 A.M. XI XIV XII X IX XIII 21 9 P.M. XII XVI XIV XV 22 9 P.M. XVII XII Numbers of Colonies Examined T ' 13 14 15 16 17 18 19 20 21 22 23 11- July 14 11 A.M. 0 0 0 0 0 0 0 0 0 0 0 15 11 A.M. 11 III III II IV I III II III 16 12 M. IV V V III V V IV IV 17 4:30 P.M. VI VII V V VII VI VI 18 9A.M. VIII VIII VII IX VII 19 10A.M. 20 11:30 A.M. 21 9 P.M. 22 9 P.M. The simultaneous study of the growth of the various colonies placed in apparently identical conditions, on the same slide or on adjacent slides, shows at first that the speed of growth is not the same for all the colonies. We have already seen that the interval between two divisions varies in rather large proportion, in the same stage, in two different colonies (i.e. four hours to seven hours between the division of / and that of //; ten hours to sixteen hours between the division of // and that of///). 42 E. FAURfi-FREMIET Table I shows the records of twenty-three colonies (experiment commenced on the 14th of July) ; the figure 0 indicates the initial macrozooid at the beginning of the peduncle formation, and the Roman numbers indicate the number of the terminal individual on the main strain ; we see, thus, that on the fourth day, there may be a difference of two generations between different colonies and that on the eighth day the difference may be four generations. The whole number of indi- viduals borne by each colony differs, of course, proportionally. N 130 no no wo 90 80 70 60 SO 40 30 20 10 ll'f 96 168 193. H. FIG. and 7) ; 16. Curves of growth from four colonies of Z. altcnians (Nos. 1, 2, 4, number of the individuals in ordinates ; time (in hours) in abscissae. The data relative to colonies Nos. 1, 2, 4, and 7 are plotted in the curve of Fig. 16. These are only gross numbers, there being no cor- rection for some microzooids parasitised or dropped out. Besides, these various curves show that for each colony the rate of growth varies itself in the course of the growth; but it is difficult to determine the part of the accidental factors already mentioned and capable of introducing some disturbance. Fig. 17 represents in function of time the genealogical and complete view of a colony having given sixteen generations on the main strain. ZOOTHAMNIUM ALTERNANS 43 The essential data are given by the successive records of colony No. 2, completed, as regards the incomplete branches sprung from A, B, and C, by the data furnished by other colonies studied in the same experi- ment (3, 5, 20, etc.). Furthermore, the periods of some divisions have been settled according to the survey of the successive and periodical ex- aminations of colony No. 2 with interpolations ; I have kept account, in this case, of the interval settled with more precision than in other ex- periments in which either the first stages of the colony or the growth of a branch were connected at intervals of time most closely approached from hour to hour. The curve represented in Fig. 18 is drawn according to this scheme. The daily increase of the number of individuals shows the following numbers : Time (in hours) Number of individuals Increase of the unity of mass in 12 hours Number of zooids made in 24 hours 0 1 1 12 2 2 24 4 2 3 36 9 2.25 48 15 1.66' 11 60 23 1.58 72 31 1.34 16 84 41 1.32 96 55 1.34 24 108 66 1.20 120 84 1.27 29 132 104 1.23 144 122 1.17 38 156 137 1.12 168 147 1.06 25 The first part of this tabulation shows a rather regular increase and such that the number of the individuals, i.e., approximately the whole protoplasmic mass, doubles at regular intervals, from twelve hours to twelve hours. Of course, we find again here, at first the geometrical progression of the ratio 2 which characterized the multiplication by bipartition of a mass of cells which keep always the same speed of growth. If we choose for unity of time this period of twelve hours, we see, however, that after the second day the rate of growth of the unity of mass, which averaged about 2, slows down progressively from 1.66 to 1.58 and 1.34, then persists for some time at a median and constant level: 1.32, 1.34, 1.20, 1.27. Then, in a last period, this rate of growth again slows down with the * Y 44 E. FAURfi-FREMIET values 1.23, 1.17, 1.12, 1.06; but the difficulty in obtaining an exact enumeration does not permit a determination of its values when the colony approaches its greatest size. Then it appears that the growth progressively slackens in the whole of the colony ; the time necessary to double the protoplasmic mass grows as the protoplasmic mass increases; it is a limiting factor of the growth. But it is evident that this factor (or limiting factor), in the case of Z. alternans, is not a factor of senescence which affects equally all the individuals, and involves a sort of progressive segregation, whose nuclear phenomena give a parallel objective picture. JULY 1 > I 01 *•— M VI JJ| XvJ XIV 1 JN • | xu J V T -168 -Iff -14-*, \ \ 1 .131 XII 1 -ItO L v . K XI 1 ' ' -/at i X J •* > i IX -9s '- L vm H — ' -J . gt vn ?! - fz vi F — 1 - to ' W D 1 _ 3t> c m n -24 B A i _ It Ci . 0 15 H, FIG. 17. Genealogical view of a colony at the sixteenth generation; time in abscissae; lineage of each individual in ordinates. This leads us to examine the case of the main strains. After the second division, which takes place rapidly, four to seven hours after the first one and at a temperature of 21° C., the rhythm of the bipartitions of the axial macrozooid slows down, (sixteen to seventeen hours between second and third divisions), then remains sensibly constant. During the entire growth of the colonies, more than twenty bipartitions of the axial macrozooid succeeded each other at intervals of ten to six- teen hours. The growth of the axial peduncle was fairly constant. It seems then that during eight to ten days at least, the functional activity and the power of growth of the axial macrozooid remain con- stant, and, in the colonies already developed, one can observe the for- ZOOTHAMNIUM ALTERNANS 45 mation of a posterior ciliary crown around the terminal individual. Thus the axial macrozooid can become a migrating individual equivalent to a ciliospore, but one never observes in this case the endomictic trans- formation of the nuclear apparatus. We have seen how the nuclear segregation which is established during the differential divisions seems to determine the characteristic features of the median individuals and of the microzooids. However, we must admit that the later divisions of the microzooids are still different, although they are not accompanied by a visible nuclear segregation. According to the rule of Claparede and Lachmann, we can still distinguish in one branch one main strain and lateral strains. The fourth branch, for instance, after the differential divisions which separate ID and Id may be represented as follows: Id gives 2d- and 2dl. Let us give the exponent 1 to the main strain of this branch; 2d'2 gives 2rf21 and 2d" which do not divide any further ; 2rf1 on the contrary gives 3d- and 3d1. The smaller branch issued from 3d2 has r. principal axis, but the number of generations is reduced. The first division separates 3d", which does not divide any further, and 3d"\ which gives 3d'21- and 3d211 without descendants. The individual 3d-- gives 4dl and 4r/2 ; 4d~ gives 4d'22 without descendants and 4d21, which still gives 4d'21- and 4(/211 without descendants. The individual 4o!1 tf. ISO no ISO /to I/O 100 go so 1° 60 SO 1,0 30 SO w FIG. 18. Curve of growth of Z. altcrnans colony drawn from Fig. 17. 46 E. FAURfi-FREMIET finally gives 5d'2 without descendants, and Srf1 which divides into 6dl and 6d- without descendants. The interval which separates the microzooid divisions is at first of the same order (or even more rapid) than the interval which separates the divisions of the axial individual ; but it increases progressively and in such a colony, for example, the individuals of the sixth branch will represent six successive generations from the cell F, while its sister cell VI will have given during the same length of time ten successive genera- tions. We can see from Fig. 17, for instance, that the microzooids 2d~- and 2d21 live more than three days and a half without bipartition ; such a fact is more typical with some microzooids of the earlier branches, A and B, which maintain themselves for more than five days without division. But after this time (corresponding to ten generations on the main strain of the branch) these individuals do not appear larger than the others ; yet they feed and their protoplasm contains many digestive vacuoles. The decrease of the power of growth which characterizes these individuals is not dependent upon their age — and for this reason we cannot admit the notion of the factor of senescence — but of their position in the colony, as if the differential divisions assured the pro- gressive segregation of a factor of growth. But we can still notice that this segregation, as it may be seen by the form of the growth of the branch D, for instance, is yet continued during the divisions of the microzooids which show no longer a differential appearance. In short, if we bear in mind the main axis of the colony, its branches and its boughs, we see that the power of growth, and of multiplication, decreases according to a kind of gradient, in proportion with its removal from the main strain. The differential character of the cellular divisions seems to be the essential condition which slows down and restrains the growth of the colonies of Z. altcmans. But, theoretically at least, this restricted growth should go on indefinitely. It is not the case here. Secondary factors play here an important role; the development of different para- sites (Protozoa, Protophytes) make it impossible to obtain a normal growth of the colonies beyond ten days, under ordinary laboratory con- ditions or in a natural marine environment ; soon the last surviving in- dividuals leave the common peduncle. The microzooids often form in this case a posterior ciliary wreath ; their fate has not been determined.3 3 A few cases of conjugation have been observed between a terminal macro- zooid and a migrating microzooid. These cases were rare ; the later phenomena were not followed. s ZOOTHAMNIUM ALTERNANS 47 CONCLUSIONS The sexual cycle described by Furssenko and by Wesenberg-Lund in the voluminous species of Zoothanin'mm (Z. arbuscula Ehrb., Z. gcniculatuui Ayrton) is rather special and with Z. alternant (Clap, and Lach.) I have never observed anything- similar, either on the Britanny coast or in my cultures at Woods Hole; I will not, then, attempt to compare the evolutionary cycle of these different species. The objective that has led me into the minute study of these colonies of Vorticellidae is the cyclical evolution — generally considered — of an initial cell's lineage, which is here the foundation macrozooid or the " ciliospore." The growth of colonies of Z. alternans is limited, in a great meas- ure, by external agents such as parasitic infections, or the growth of animal and vegetable microorganisms which change the surrounding conditions of a specific colony. In the cultures watched as described above, these various circum- stances, somewhat accidental, are much reduced ; yet the growth of each colony appears to be limited in itself ; I have taken the common in- dividual— the microzooid — as unity of mass, and I have observed that the rate of growth decreases in function of time for the whole of each colony studied ; at the same time, some particular migrating individuals are formed and become the source of new colonies ; it is precisely this " cyclical " appearance of growth in the colonies of Vorticellidre that 1 have described in an earlier paper ( Faure-Fremiet, 1922) ; I have con- sidered two different hypotheses : ( 1 ) the formation during the evolu- tion of the migrating individuals of a limited stock of an hypothetical " active substance " which divides and becomes increasingly smaller with each generation of daughter-cells, or (2) a progressive modification of the intimate composition of the cells, variations which would lie " corrected " only during the evolution of their own migrating cells.' In any case, this cyclical and limited evolution gives to the colonies of Vorticellidce (Epistylis, Carchcsiuni, ZootJiaiiniiiun) somewhat of an individualized character. In this regard, the case of Z. alternans is very striking. At first, the successive divisions of the cells derived from the first individual and the regularity with which they follow one another in exactly determinate planes which fix the general features of the colony, closely recall the process of a strictly predetermined cleavage, but one which would be complicated with a continuous growth. Secondly, the existence in these colonies of a main strain and of secondary strains characterized by different nuclear qualities and dif- ferent evolutionary properties recall in a certain measure the separation 4 These suppositions have been examined and criticized in a very interesting work of G. Teissier (1928). 48 E. FAUR£-FREMIET of the germinative and somatic strains during the cleavage of an Ascaris egg- Thirdly and finally, we can characterize the individuality of the colony by the repartition of the power of growth and the power of multiplication of its cells according to a certain gradient. In connection with another species of ZootJictiniiiiini Wesenberg- Lund also considers the notion of the individuality of the colony, for the various individuals are tied by the continuous protoplasmic, thread of the ramified peduncle and this brings about in their mass rather a physiological unity. But the above-indicated characteristics are again met, more or less accentuated, in other colonial Vorticellidse in the species Epistylis and Carchcsium, for example, which do not show any protoplasmic connection between the zooids. The case of these colonies is then nearer that of a " population " of cells, and their cyclical evolution appears very similar to populations of free Infusoria, studied by so many authors. The case of Z. alternans is still, from this point of view, particularly interesting. In these species, the Claparede and Lachmann rule shows that two daughter-cells have not necessarily the same power of growth and of proliferation. I found the same rule (1922) in some species of the genera Carchesium and Epistylis, and more especially with Epistylis arcnicola (n. sp). Here there seemed to exist in the course of the successive biparti- tions a kind of progressive segregation of the power of growth, but we find in Z. alternans, as an objective support of this hypothesis, the differential divisions, which are produced at the origin of each lateral branch and which indicate a kind of nuclear segregation. In this species the main strain's cells which keep a constant nuclear appearance, keep also a constant rate of growth and, apparently, an indefinite multiplicative power. We witness, then, a cytological mechanism, probably independent of the external factors which rule the functional differentiation of the cells belonging to the same family, in a process of growth. This cytological factor, or those which are superimposed upon it, rules at the sa.me time the family's general mode of growth ; it intervenes as a limiting factor, independent of the colony's age, and quite distinct, by this fact, from a factor of senescence in the true meaning of this word. However, the colony's initial individuals, the " ciliospores," ap- pear to be characterized by a kind of " physiological potential " greater than that of the main strain's common individuals. As in all the colonial Vorticellidas that I have previously studied, they are characterized by large size and by the presence of definite ZOOTHAMNIUM ALTERNANS 49 granulations connected with the secretion of the basic peduncle's inert substance. During their particular growth, accompanied by a complete changing of the nuclear apparatus, the cells acquire these properties and we can thus show that near the end of the colony's cycle of growth an endomic- tic cycle exists, closely comparable to that observed in a population of free Infusoria. But we must remark that, here again, the particular evolution of these " ciliospores " and the endomictic phenomena of which they are the seat, are determined, not by their age, but by their place in the colony's plan, just as if this evolution were still connected with the same mechanism of differential division and of nuclear segregation.5 I am very glad to be able here to express my thanks to the Inter- national Education Board, to my American colleagues who made my residence at Woods Hole so profitable for me, and, very particularly, to Dr. Calkins and Mrs. Harnley, who have helped me in translating this paper. SUMMARY 1. The first division of the initial macrozooid (or ciliospore) deter- mines the median antero-posterior plane of the colony; the subsequent cleavages of the daughter individuals are brought about according to equally determined schemes, which give the main strain (or axial trunk) and the lateral branches, alternately at right and at left. 2. The individuals constituting the main strain are of a rather large size (axial macrozooids) ; their cleavage is always accompanied by a differential division giving rise to a new axial macrozooid and a median microzooid. 3. The differential divisions are characterized by an unequal division of the protoplasmic mass, accompanied either by a sensibly equal di- vision of the macronucleus (division supposed to be quantitatively differential), or by the unequal division of the macronucleus in which the larger mass (delicately granular) remains in the larger individual, while the thinner part (often of fibrillar structure) goes to the micro- zooid (division supposed to be qualitatively differential). 4. The cleavages of the ciliospores and those of the axial macro- zooids, I, II, and III are always differential as regards the protoplasm and the nucleus. The cleavages of the macrozooids IV and after give a cytoplasmic differential division and an equal nuclear division; the dif- 5 Long ago I mentioned an apparently differential division in Lagcnophrys, in which one of the individuals remained sedentary, while the other migrated and secreted a new shell (1904). 4 50 E. FAURfi-FREMIET ferential division of the macronucleus is carried back to the cleavage of the corresponding- median microzooids. 5. The common microzooids have a limited power of growth and of multiplication. 6. The median individuals having a large macronucleus after the dif- ferential division of the median microzooids D and progeny begin an active period of growth accompanied or unaccompanied by only one ulterior division : these forms constitute the median macrozooids or " ciliospores." 7. The growth of the ciliospores is accompanied by an important hypertrophy of the macronucleus followed at first by a disintegration, then by a reconstitution through an endomictic process. 8. During the growth of the median macrozooids, some grains of se- cretion accumulate at the individual's posterior end, then the ciliary crown grows, the ciliospore breaks away, swims freely, then settles down on a substratum and becomes the source of a new colony. 9. The character of the differential divisions on the main strain seems to determine the individual's differentiation of the colony; this differentiation depends not only on the individual's size, but also on its physiological potencies. 10. Independently of the obviously differential divisions, it is shown that the power of growth is divided among the microzooids according to a gradient, so to speak. 11. The unequal power of growth of the various individuals of a colony gives to its whole growth a behavior which approaches the be- havior of an organism. This unequal share constitutes for the growth of the whole a limiting factor very unlike a factor of senescence. LITERATURE CITED CLAPAREDE AND LACHMANN, 1858-1861. Etudes sur les Infusoires et les Rhizo- podes. Memoires dc I'lnstitnt Gcnez'ois, 5, 6, 7 and 8. FAURE-FREMIET, E., 1904. Epuration et rajeunissement chez les Vorticellidse. Compt. rend. Soc. Biol, 57: 428. FAURE-FREMIET, E., 1905. Structure de 1'appareil fixateur chez les Vorticellides. Arch. f. Protistcn, 6: 207. FAURE-FREMIET, E., 1910. La fixation chez les Infusoires cilies. Bull. Sclent. France et Bclglque, 44: 27. FAURE-FREMIET, E., 1922. Le cycle de croissance des colonies de Vorticellides. Bull. Sclent. France et Belgiquc, 56: 427. FAURE-FREMIET, E., AND GARRAULT, H., 1928. La courbe de decroissance de ponte chez " Margaropus Australis." Ann. Ph\siol. ct Ph\slcochlmle Biol., 6. FAURE-FREMIET, E., AND GARRAULT, H., 1925. La Cinetique de Developpement. Coll. Les Problemes Biologiques. Press. Univ. France. FURSSENKO, A., 1924. Zur Konjugation von Zoothamnium arbuscnla Ehrbg. Trav. Soc. Naturalistes de Leningrad, 54: fasc. I. ZOOTHAMNIUM ALTERNANS 51 FURSSENKO, A., 1924. Zur Biologic von Zoothamnium arbuscula Ehrenbcrg. Arch. Russes de Protistol., 3: 75. FURSSENKO, A., 1929. Lebenscyclus und Morphologic von Zoothamnium arbus- cula Ehrb. Arch. f. Protistenk., 67: 376. TEISSIER, G., 1928. Croissance de population et croissance des organismes. Examen historique et critique de quelques theories. Ann. Physiol. et Physicocliimie Biol., 4: 343. WESENBERG-LUND, C., 1925. Contribution to the biology of Zoothamnium geni- culatum Aryton. D. Kgl. Danske Vidensk, Selsk. Skriftcr, naturvidensk. og mathcm. Afd., 8 Raekke X, 1. THE INFLUENCE OF HUMIDITY ON THE BODY TEM- PERATURE OF CERTAIN POIKILOTHERMS F. G. HALL AND R. W. ROOT (From the Zoological Laboratory, Duke University) Poikilothermic animals are commonly accredited with possession of a body temperature closely approximating that of their environment. In general this appears to be true. However, there are some cases where the body temperature of certain " cold blooded " animals may be very unlike that of their surroundings. Such examples are given by Rogers and Lewis (1916) in a table which they have compiled from the investigations of numerous workers. It shows that not all investigators agree even as to the temperature of the same species. It is probable that much of the discrepancy is due to different types of method. On the other hand, a more careful examination of the conflicting results of various authors as to the correspondence between body and environ- mental temperature shows that the greatest variations occur when ani- mals are subjected to atmospheric conditions. The factors which influence the temperature of animals may 'be classified as follows : intrinsic — those that lie within the organism and act to produce a temperature different from that of the environment ; extrinsic — those imposed on the animal from without. The extrinsic factors are (1) conduction and convection, (2) radiation, (3) evapora- tion of water. A discussion of the role played by each factor is given by Pearse and Hall ( 1928) . It is the purpose of this paper to study the influence of the third factor, namely, the evaporation of water, on the body temperature of various poikilo therms. EXPERIMENTAL METHODS Apparatus. — The apparatus employed consisted of an air pump, several gas washing bottles — some containing concentrated sulfuric acid, others water — a chamber in which animals under experimentation were placed, temperature-measuring instruments, which included a potenti- ometer, a high sensitivity suspension galvanometer, and a copper-con- stantin thermocouple. The air pump was adjusted to supply air at a constant rate of 22.6 liters per minute through two possible air leads. One lead was through 52 INFLUENCE OF HUMIDITY ON BODY TEMPERATURE 53 four wash bottles containing concentrated sulfuric acid and the other through four similar wash bottles containing pure water. The amount of air passing proportionally through each lead was controlled by screw pinch cocks. Thus air of any desired humidity from 7 per cent to 100 per cent could be obtained. The relative humidity of the air was measured by a calibrated hair hygrometer suspended in an enclosed jar through which all the air passed before entering the experimental animal chamber. B 30 T T FIG. 1. Apparatus used to determine the influence of relative humidity on the body temperature of animals. The experimental chamber in which animals were placed is shown in Fig. 1. A cylindrical percolator (A} was immersed in a constant temperature bath (B). Animals were tied to a sliding rack (R) which was so arranged that only a small portion of the animals' bodies was in contact with it, thus allowing a maximum surface to be exposed to the moving air. The end of this rack closed the mouth of the percola- tor. Two precision thermometers (T) were inserted through the rack, one in the upper, the other in the lower portion of the percolator. The thermocouple lead wires (C~) also passed through the end of the rack. The mouth of the percolator was packed with cotton to lower the rate of conduction. The direction of air flow is shown by arrows. The temperature in all experiments was maintained at 20° C. Experimental Animals. — The species chosen for this investigation were: Amphibians — the frog, Rana pipicns Schreber; the salamander, PletJicdon glntinosiis Green ; the toad, Bufo fowleri Carman. Reptiles— the lizard, Sccloporus nndulatns Latreille ; the " horned toad," Pliry- uosoina corniiium Harlan; the turtles, Terrapene Carolina Carolina Linn., Cistudo major Agassiz, Chrysemys marginata Agassiz ; the alii- 54 F. G. HALL AND R. W. ROOT gator, Alligator mississippiensis Daudin. All animals were kept under good laboratory conditions, and were alive and active at the end of each experiment. Individuals were weighed at the beginning and end of each experiment. From four to ten individuals of each species were used and several determinations were made on each individual. Temperature Records. — Environmental temperatures were recorded by use of precision thermometers placed in the experimental chamber. The body temperature was determined with a thermocouple inserted through the anus well up into the animal's body. Each thermocouple used was calibrated against a precision thermometer (previously cali- brated by the U. S. Bureau of Standards). The temperature readings are believed to be accurate to ± 0.01° C. Records of the temperature of each animal and its environment were made at the following relative humidity points: 7 per cent, 25 per cent, 50 per cent, 75 per cent and 95-100 per cent. TABLE I Showing Variations in Body Temperature of Several Species of Poikilothcrms from Environmental Temperatures in At- mospheres of Different Relative Humidities Species Relative Humidity 7 25 50 75 95-100 Salamander -9.21 -8.60 -7.33 -0.70 -0.37 -0.72 -0.34 -0.39 -6.34 -6.75 -5.31 -0.70 +0.02 -0.57 -0.23 -0.26 -4.62 -4.68 -3.98 -0.15 +0.11 -0.52 -0.11 -0.15 -2.54 -3.01 -2.48 +0.30 +0.19 -0.41 -0.03 -0.08 -0.29 -0.13 -0.74 +0.64 +0.38 -0.12 +0.15 +0.18 Frog . Toad Lizard Horned "Toad" Turtle water Turtle, land . Alligator Plus signs signify a higher body temperature than that of the environment; minus signs indicate a depression in body temperature below that of the environment. RESULTS Amphibians. — The body temperature of the salamander, frog, and toad very closely approximated that of their environment when the surrounding atmosphere was saturated, or nearly so, with water vapor. In atmospheres of low humidity, however, a considerable depression in the body temperature below that of the environment was obtained. Salamanders showed the most marked depression, toads the least marked. The average results obtained are shown in Table I. Con- INFLUENCE OF HUMIDITY ON BODY TEMPERATURE 55 siderable weight loss was suffered by these animals. At low humidities their skins appeared dry and their bodies emaciated. Reptiles. — The response of reptiles to atmospheres of varying hu- midity was quite unlike that of amphibians. Whereas amphibians showed great depression in body temperature when exposed to a dry environment, reptiles showed only slight depression. In fact, if the relative humidity be maintained between 90 and 100 per cent, many rep- tiles will show a body temperature slightly higher than that of their sur- °c 10 8 '-£ 6 o> «j o> o> Ol o: -2 \ Amphib i a f \ Reptiha 0 20 40 60 80 100% Relative Humidity FIG. 2. Graph showing relation of body temperature to environmental tem- perature of amphibians and reptiles when subjected to different relative humidities. roundings. Lizards and water turtles (Chryscinys marginata) were influenced the most by low humidity. Apparently the water turtle is slightly more susceptible to the influence of humidity than the land form. Weight loss in the reptiles was practically nil. Subjection to low hu- 56 F. G. HALL AND R. W. ROOT midity for long periods of time showed no apparent injurious effect. Table I contains the average results obtained on all forms summarized to show the difference in response to surroundings of varying relative humidity. Fig. 2 shows the comparison of the response of amphibians as a group with that of reptiles, and shows the variation in the change of body temperature from that of the environment at similar relative humidities. DISCUSSION It is apparent from the results obtained that in atmospheres of low relative humidity, amphibians will have a much lower body temperature than that of their environment. Such a condition results from the evaporation of water from the surface of the body. The body tem- perature of reptiles is but slightly changed by similar conditions. Thus it is clearly indicated that the difference in response of these two classes lies in the type of integument. The amphibians with moist skin will readily lose water by evaporation. They have little means of retaining water as has been shown by Gray (1928). The moisture of their integument is in dynamic equilibrium with the water content of their environment. The inner tissues supply water when that at the surface has been evaporated (Hall, 1922). Thus, for example, a salamander behaves physically very much like a wet bulb thermometer. The de- pression in temperature is not as great, probably because water is not transported to the surface as rapidly as in the wick of a wet bulb ther- mometer. Amphibians are limited in their habitat to moist places. They possess a "reaction pattern" (Pearse, 1922), which permits them to live only under damp logs and stones or in marshes or other watery places. Thus they become more conspicuous on rainy days when the atmosphere offers a more favorable and less restricted environment for their activities. It is perhaps interesting to speculate that a frog may have a lower body temperature on a dry, sunny day than on a somewhat colder, rainy day. The possession of a scaled integument, characteristic of the reptiles, greatly increases the power of water retention. Reptiles give up water very slowly and will resist desiccation for long periods of time (Hall, 1922). Not only by possession of an integument, but by certain in- ternal physiological processes, such as the elimination of nitrogenous wastes as uric acid instead of urea, they conserve water. In conse- quence many reptiles live in very dry surroundings. Perhaps the principal explanation of the discrepancies in reports by many investigators of the correspondence between body and en- INFLUENCE OF HUMIDITY ON BODY TEMPERATURE vironmental temperatures is that they are due to a lack of control or record of humidity. In the light of these experiments any results ob- tained without knowledge of the relative humidity of the surroundings in which an animal's temperature is taken would seem meaningless. A further observation seems to indicate that the influence of changes o in humidity on the body temperature of these animals decreases as animals higher in the phylogenetic series are used. It appears that t; 10- o> xxo CD 8- 6- ^ 4- o> o> 0> 0 «0 00 Spec ies FIG. 3. Showing a comparison of the change in temperature of the body of each species studied when the humidity was lowered from 100 per cent to 7 per cent saturation. amphibians as they progress in evolution show a decrease in their sus- ceptibility to humidity variations. The same fact apparently holds for the reptiles. Fig. 3 represents the results arranged to show the maxi- mum change in body temperature relative to environmental temperature in each of the species used, the salamander showing the greatest change, the alligator the least. The reptiles seem to have a more stable body temperature than amphibians because they are less influenced by en- vironmental factors. Possibly the increased ability of water retention evolved in the reptiles is a " milestone " on the road to homoiothermism. 58 F. G. HALL AND R. W. ROOT SUMMARY 1. Amphibians show marked response in body temperature to en- vironmental variations in relative humidity. When subjected to an atmosphere of 7 per cent relative humidity at 20° C., a depression of several degrees centigrade may occur in their body temperature. 2. Reptiles show very little response to variations in relative hu- midity. The integument apparently prevents the evaporation of mois- ture from the surface of the body. 3. It is suggested that the evolution of the scaly integument of reptiles from the slimy and moist skin of amphibians, with the con- comitant power of water retention, is perhaps an important step in the evolution of homoiothermism. BIBLIOGRAPHY GRAY, J., 1928. The Role of Water in the Evolution of Terrestrial Vertebrates. Brit. Jour. E.rpcr. Biol, 2: 26. HALL, F. G., 1922. The Vital Limit of Exsiccation of Certain Animals Biol. Bull., 61 : 31. PEARSE, A. S., 1922. The Effects of Environments on Animals. Am. Nat. 56: 144. PEARSE, A. S., AND HALL, F. G., 1928. Homoiothermism. New York. ROGERS, C. G., AND LEWIS, E. M., 1916. The Relation of the Body Temperature of Certain Cold Blooded Animals to that of their Environment. Biol. Bull, 21: 1. THE POINT OF ENTRANCE OF THE SPERMATOZOON IN RELATION TO THE ORIENTATION OF THE EM- BRYO IN EGGS WITH SPIRAL CLEAVAGE T. H. MORGAN AND ALBERT TYLER (From the Marine Biological Laboratory, Woods Hole, and the William G. Kcrckhoff Laboratories of the Biological Sciences, California, Institute of Technology) If the entrance of the spermatozoon into the egg is instrumental in determining the planes of cleavage, and the cleavage planes bear a definite relation to the embryonic axes, it would still remain important to find out whether the side of the egg on which the sperm enters is a factor in locating the dorsal (or ventral) side of the embryo. In some eggs having an equal first cleavage, such as the frog, the ascidian and the sea-urchin, observations of this kind have been reported, and a distinct relation has been found between the side of the egg on which the sperm enters and the future dorso-ventral axis of the em- bryo. Curiously enough, despite the large number of careful observa- tions on the cell-lineage of eggs with a spiral type of cleavage, there is only one set of observations on the relation of the entering point to the first cleavage plane, and even here we do not know whether the side on which the sperm enters becomes the dorsal or the ventral side. In the course of our work another relation was found that is both novel and has a bearing on the interpretation of the so-called law of alternate right- and left-cleavage in spiral types. In Cumingia it was discovered that two types of second cleavage occur in equal numbers, one of which in ordinary parlance would be called a right-handed, the other a left-handed spiral, yet in both cases the third cleavage was found to be always dexiotropic. As a consequence of this relation it follows that in one case the first plane of cleavage corresponds to the median plane of the embryo, and in the other case the second plane of cleavage corresponds to the median plane, provided the later sequence of events is the same for both types. A third relation has not, so far as we know, been carefully studied, namely, whether in eggs with an unequal first cleavage, the plane of cleavage passes through the pole or consistently to the side. Without exception our observations show that the plane passes to the side on 59 60 T. H. MORGAN AND ALBERT TYLER which the smaller cell comes to lie, but the relations here are not the same in the three types examined, nor are the succeeding events always the same. However, these relations will be shown to have a significant bearing on the location of the median plane of the body. The Cleavage of Cumingia The early cleavage of the egg of the bivalve mollusk Cumingia tell'moides has been described by Morgan (1910) and Browne (1910). The following observations were made in the summer of 1929 at Woods Hole, Mass. The eggs and sperm were obtained by the usual method of isolating individuals in small dishes of sea water. The eggs were washed and samples removed for fertilization at once or soon after deposition. A square of vaseline was laid down on a slide and two fragments of No. 2 cover slips placed on the vaseline for additional support. A drop of eggs was placed in the square and a small drop of very dilute sperm-suspension was added. A cover slip was placed on the preparation and the slide was examined at once under the micro- scope. The eggs were brought under observation in less than thirty seconds after insemination. To some of the eggs one or more spermat- ozoa were already attached ; to others they soon became attached. Only those cases in which one or a few spermatozoa were attached were fol- lowed— if the insemination had been too heavy the slide was rejected. The egg of Cumingia is about 66 micra in diameter, and with the jelly about 107 micra. The glass supports were about 140 micra in thick- ness, which with the further help of the vaseline sufficed to prevent compression of the eggs. The pole of the egg of Cumingia can readily be identified by a clear area free from pigment. The outer pole of the first maturation spindle lies in the center of this area. The identification of the pole is later checked by the point of extrusion of the polar bodies. The sperm enters at any point of the periphery of the egg. On attaching itself to the egg the spermatozoon becomes immotile, its tail extending radially from the surface. About 30 seconds after attachment the egg rather suddenly becomes distinctly ovoid in shape, with the more pointed end at the point of attachment. This change in shape lasts 30 seconds or less. As the egg rounds out again the sperm enters. This phenom- enon enables one to identify the particular sperm that will enter, even before the sperm-head has penetrated. Other sperms in the jelly, ap- parently even touching the surface of the egg, do not call forth this striking reaction. The change in shape is something more than the formation of a fertilization cone, since it involves a change in form of the whole egg. Unless the entering sperm is exactly on the horizon, ENTRANCE POINT OF SPERM AND CLEAVAGE 61 the change in form of the egg may not he observed; also there seems to be some difference in different sets of eggs as to its appearance. The first polar body appears five or six minutes after fertilization. In making observations, all the sperms at or near the periphery of the egg were located on a drawing, and their relative position in three dimensions noted. Those that did not enter served as markers. Spermatozoa that are too far above or below the optical section of the egg cannot always be seen. When the polar body appears, the pole can be more accurately located in relation to the position of the entrance point. As a rule only one egg in each preparation was followed. The observations were made under magnifications of 284 and 440 diameters. The first cleavage appeared about 50 minutes after fertilization. The location of the plane was noted in the drawing with respect to the point of entrance. This was checked as far as possible by the posi- tion of the markers, since, if any shifting of the egg occurred, their positions would change. The first division; Fig. 1, a, b, is unequal. a' b FIG. 1. The first and second cleavages of Cumingia showing the two pos- sible types of 4-cell stage. In a the C-cell comes off " counterclockwise " ; in b' " clockwise." The smaller blastomere, following the convention for this type of egg, will be called AB, and the larger blastomere CD. The second cleavage, Fig. 1, a', b', divides AB equally (A and B), and CD into unequal parts (C and D) ; the C-blastomere being smaller and approximately 62 T. H. MORGAN AND ALBERT TYLER the size of A or B. Theoretically the C-cell might form from either side of CD (Fig. 1, a', b'). It is obvious, then, that there would be two possible configurations or arrangements of the blastomeres after the division that are mirror figures of each other (Fig. 1, a', and Fig. I, b'). As will be shown, it is important at this stage not to identify these two types as dextral or sinistral cleavages, although this would be the usual interpretation. The clockwise sequence ABCD may seem to imply that the second cleavage has been leiotropic and the third will be dexiotropic, or con- versely for the counter-clockwise sequence DCS A ; but by utilizing the usual lettering we do not wish here to commit ourselves to such an implication. The reasons for this will appear later. Entrance Point of Spermatozoon in Relation to the First Cleavage in Cumin gia Ninety-eight cases were recorded in which the relation of the en- trance point to the first cleavage plane was definitely ascertained. In 77 cases there was strict coincidence between the plane of the first di- vision and the entrance point. In 13 the entrance point was less than 45° from the cleavage plane. In 8 cases the divergence was greater than 45° and less than 90°. Whether the expectation of close coin- cidence should be 100 per cent and the departures be considered as due to abnormalities, or as due to errors of observation may be briefly considered. Polyspermy might introduce a complication, but it can be detected either by the presence of extra pronuclei, or by irregularities in the cleavage. Compression of the egg might be one of the factors de- termining the position of the cleavage plane. To avoid this, the sup- ports were made so thick that the space between the slide and the cover slip was greater than the diameter of the egg plus the jelly. If the sea water evaporates, the retreating edge of water may cause the egg to move, and the hypertonicity might cause irregularities in cleavage. This was avoided to a large extent by the wall of vaseline ; also eggs were selected that lay in the centre of the drop. Any movement of the eggs can be detected by their position with respect to neighboring eggs. The change in shape that the egg undergoes before cleavage is not a serious source of error, especially if checked by the presence of " markers " on the egg, but during division the change in shape of the egg may cause slight changes in position. Therefore, whenever pos- sible, the egg was constantly watched throughout this period. In some cases when the cleavage is horizontal the egg may roll over. This is prevented to some extent by avoiding jarring of the table etc. When ENTRANCE POINT OF SPERM AND CLEAVAGE 63 one or more of these factors was observed to come into play, the egg under observation was rejected. The first cleavage plane does not pass through the pole (as deter- mined by the position of the attached polar bodies), but slightly to one side. When considered from the entrance point of the sperm, the pole of the egg being up, this plane may be said to pass to the right or to the left of the pole. Whenever the first plane passes to the right of the pole, the AB-ce\\ comes to lie to the right of the entrance point (Fig. 2, a) ; whenever it passes to the left of the pole, the AB FIG. 2. The cleavage planes of Cumingia with respect to the entrance point of the spermatozoon, a, 2-cell stage with AB to right of sperm-entrance point ; b and c, the two possible types of second cleavage. comes to lie to the left (Fig. 3, a). This simple relation, which is constant in all the eggs examined, has apparently been overlooked by earlier observers in eggs of this type. The polar bodies adhere to the surface of the CD blastomere, and are carried into the furrow during the first division. Of the 77 cases of coincidence between the entrance point and the first plane, the AB was to the right in 40 cases and to the left in 37 cases. It appears that the chances are equal that the smaller cell lies to the right or to the left of the entrance point. The bearing of these two possibilities on the location of the plane of bilateral symmetry will be considered presently. It is obvious that when the small cell (AB) lies to the right of the entrance point there are two possible types of second cleavage (Fig. T. H. MORGAN AND ALBERT TYLER 64 2, b, and Fig. 2, c) ; similarly when the small cell (^/£) lies to the left (Figs. 3, b, and 3, c). As a matter of fact it was found in these 77 cases of coincidence that when the AB was to the right, only one of the two theoretical types appeared, namely, that shown in Fig. 2, b. When the AB was to the left, again only one of the two theoretical types appeared, namely, that shown in Fig. 3, b. Ordinarily the cleav- FIG. 3. The same as Fig. 2, except that the AB-ce\l lies to the left of the sperm-entrance point. age giving the first type (Fig. 2, b) would be called a leiotropic second cleavage, implying that the third would be dexiotropic. The second type, Fig. 3, b, would be called a dexiotropic second cleavage, implying a leiotropic third. However, a study of the third cleavage of Cumingia has shown that the division is always dexiotropic. This information was obtained from eggs preserved at the time of the oncoming third cleavage. The orientation of the spindles with respect to the poles was determined in 84 eggs, and in every case they showed the cleavage to be dexiotropic (Fig. 5). The observation shows in the first place that it would have been erroneous to conclude that because the third cleavage is dexiotropic, the second must have been leiotropic. It would have been equally erroneous to have concluded from the two types of four-cell stages that the direction of the spiral would be dif- ferent in the two types. By parity of reasoning it would seem unjusti- fiable to infer that because a given egg shows a leiotropic second cleav- age, the first cleavage must have been dexiotropic, and thus to designate the egg as a dexiotropic egg. . ENTRANCE POINT OF SPERM AND CLEAVAGE 65 Such reasoning might have led one to infer that a dexiotropic third cleavage in Cumingia means that the second cleavage must have been leiotropic. A study of preserved eggs in the anaphase of the second division gave no indication of a spiral arrangement of the spindle. The spindles in the CD- and AB -cells appear to lie in the same hor- izontal plane (Fig. 4, a, b, c, d~), instead of being tilted in opposite directions, as has been described for other eggs at this division (Mead, Conklin). Of course it is possible that the tilting of the spindles in the Cuniingia egg is too slight to be visible, but nevertheless it is in- FIG. 4. Two-cell stages of Cumingia showing the positions of the spindle for the second cleavage, a and b, polar views ; in a the C-cell will come off clockwise, in b the C-cell will come off counterclockwise, c and d, antipolar views ; in c the C-cell will come off counterclockwise, in d, clockwise. The two poles of the spindles appear to lie at the same level in all cases. teresting to note that in this egg in which two different types of four- cell stages occur the spindles do not show a visible tilting. The spindles, in the AB- and CD-cells, are horizontal as shown in the figures (Fig. 4, a, b, c, d). However, they are not parallel, but, especially in the CD-cell, the spindle makes an angle with the plane of division. In order to answer the question, if it should arise, as to whether both types of cleavage in Cumingia produce normal embryos, a few eggs of each type were isolated. Normal embryos developed from each. The normal trochophore swims in a dexiotropic spiral. This also occurred in the embryos from these two types. Moreover, all the em- 5 66 T. H. MORGAN AND ALBERT TYLER bryos from a culture swim in the same kind of spiral. In adult Cumin- gia the two valves of the shell are different in the articulation joint on the median dorsal side. All shells examined were alike, i.e., not right or left, but all the same. Location of the D-Ccll in Relation to the Entrance Point It has been found that when the first plane passes to the right of the pole (Fig. 2, a) the next division is always of such a sort that the D-cell is later away from the point of entrance of the sperm (Fig. 2. M. Similarly when the first plane passes to the left of the pole (Fig. 3. «) the next division is always of such a sort that the future Z)-cell is again away from the point of entrance (Fig. 3, b). The records from living eggs show that in 32 cases in which the cleavage plane passed to the right of the pole, the L>-cell lay on the side opposite the entrance point, giving the arrangement of the blastomeres shown in Fig. 2, b. In 30 FIG. 5. Four-cell stages of Cuniiiigia showing the position of the spindles for the next division. In all cases the spindles show that the next division will be dexiotropic. cases in which the plane passed to the left, the .D-cell also lay on the side opposite the entrance point, as in Fig. 3, b. No exceptions to this rule are found. So far the description has been restricted to those cases where the first cleavage plane coincided very nearly with the entrance point. In addition there were a few other cases, as reported above, where the coincidence was not so close and where there were no reasons to sup- ENTRANCE POINT OF SPERM AND CLEAVAGE 67 pose that errors of observation were made. There were 13 such cases recorded in which the cleavage plane was less than 45° from the en- trance point. If the entrance point is arbitrarily brought to the nearest point of the actual cleavage plane, then there are 3 cases in which AB is to the right of the entrance point, and 10 cases in which AB is to the left. The same relations of Z)-cell to entrance point obtain for both of these sets of cases as for those in which there was strict coincidence. There were also 8 cases in which the first cleavage plane was more than 45° from the entrance point. This divergence is too great to make a comparison profitable. Relation of the Entrance Point of the Sperm to the Plane of Bilateral Symmetry The evidence reported above has an important bearing on the re- lation of the point of entrance of the sperm to the plane of bilateral symmetry of the body. It has been shown that in 78 per cent of the cases close coincidence was observed between entrance point and first cleavage plane. In about half of these the first cleavage passed to the right of the pole (Fig. 2 a), giving the type of 4-cell stage shown in Fig. 2, b. At the next cleavage, the third, the 1-d micromere forms dexiotropically (Fig. 5). If from this point onwards the cleavages FIG. 6. Diagrams indicating the location of the 4-d cells in the two types of cleavage shown in Fig. 2, b, and in Fig. 3, b. alternate, left and right, the 4-d cell will come off leiotropically and will lie next to the second plane of cleavage as shown in Fig. 6, a. It has been shown (Lillie, 1895) for at least one pelecypod (Unio) that the 4-d blastomere gives rise to the larval mesoblast. and establishes the plane of bilateral symmetry. This means that the second plane of cleavage coincides approximately with the median plane of the body. In the other half of the recorded cases the first cleavage passed to the left of the pole (Fig. 3. a) giving the type of 4-cell stage sho\v; 68 T. H. MORGAN AND ALBERT TYLER in Fig. 3, b. The l-d again forms dexiotropically, Fig. 5. It follows from the same reasoning that the 4-d micromere comes off leiotropically, and will here lie next to the first plane of cleavage as shown in Fig. 6, b, and this plane of cleavage will now approximate the median plane of the body. It may seem, then, that either the first or the second plane of cleav- age may become the median plane of the body. This follows only on the assumption made above, which, although known to be true for other eggs, has not been entirely shown in this case. It is possible, for example, that the second somatoblast which determines the median plane may be formed at different divisions in the two cases. If, for example, in the type shown in Figs. 3, a, and 3, b, the second soma- toblast appeared one division earlier or one division later, the median plane would be the same as in the other case (Fig. 2, a, and 2, b). As shown by the evidence, when the first cleavage plane passes to the right of the pole, the plane of bilateral symmetry coincides with the second cleavage plane, and when it passes to the left, with the first cleav- age plane. What determines the passage of the first cleavage plane to the right of the pole in some cases and to the left in others is un- known. The fact that about 50 per cent of each type occurs suggests that it is merely a matter of chance. If we assume that the unfertilized egg has its materials radially arranged around the polar axis, and that the entering sperm determines through movements of the contents of the egg (or otherwise) that materials correlated with the determination of the D-cell come to lie opposite the entrance point of the sperm ; and furthermore, that the cleavage plane does not pass through this ma- terial, then a possible interpretation suggests itself. It is obviously not necessary to make this assumption in quite the same crude form as suggested above in order to express these relations, for, at the time of the first division, all of the egg appears to be involved in the process. The risk of making such a generalization will be apparent when another egg, Chatoptcrus, is examined. The Cleavage of C licet opterus The eggs were washed in sea water, and allowed to stand about 20 to 30 minutes during which time the first polar spindle forms. A drop of eggs was put onto a slide prepared in the same way as for Cumingia. The egg measures 106 micra in diameter, without the jelly, and 111 micra with the jelly. The same thickness of cover slip sup- port etc. was used as for Ciuningia. A very small drop of very dilute sperm-suspension was added to the eggs which were examined immedi- ately. In most cases the spermatozoa were already attached as though ENTRANCE POINT OF SPERM AND CLEAVAGE 69 the combination had been made almost instantaneously. The sper- matozoon enters 15 to 30 seconds after insemination and may be missed unless the preparation is examined very quickly. The pole of the egg can be identified by the clear area in which the spindle for the first maturation division lies. The sperm enters at any point, and a slight fertilization cone appears at the point of entrance. The extra sperm which do not enter remain attached, and serve as markers. The exact position of the pole is given by the location of the polar bodies. The cleavage of the egg of Chatopterus has been described by Mead, Wilson, and Lillie, and the relation of the median plane of the body to the first cleavage plane determined, but so far no one has examined the relation of the entrance point of the sperm to the first cleavage. The third cleavage of the egg is dexiotropic, and the fourth leiotropic, so that 2-d (the first somatoblast) comes off near the second cleavage plane, and 4-d (the second somatoblast) is similarly placed. This de- termines that the median plane of the body lies near the second cleavage plane. The Relation of the Entrance Point of the Sperm to the First Cleavage Plane As in Cniningia the location of the sperm that had entered was recorded on the drawing, and the individual eggs watched until the cleavage furrow appeared. In 48 eggs there was a fairly strict co- incidence; in 35 eggs the entrance point was less than 45° from the plane of the first division, and both to the right and left of the plane. In 33 eggs it was more than 45° and less than 90° to the right and left. Thus in only 41 per cent of the cases was there a close agree- ment between entrance point and cleavage plane ; but if the entrance point is not in some way correlated with the direction of the first cleavage plane, even this percentage of coincidence would not be ex- pected. Taking first the cases where coincidence occurs, it was found that in 23 cases the first plane passed to the right of the pole, which means that the AB-cel\ lay to the right of the entrance point as in Fig. 7, a. In 25 cases it passed to the left of the pole, thus placing the AB-cell to the left of the entrance point as in Fig. 7, b. In both cases, however, the second cleavage gave the same arrangement of cells, namely, that shown in Figs. 7, a', b'. According to the usual convention these four-cell stages would be obtained from leiotropic second cleavages (which is actually true for the Ch&topterus egg), but in one type, Fig. 7, a', the Z7-cell would lie away from the entrance point of the sperm, and in the other type near the entrance point (Fig. 7,6')- 70 T. H. MORGAN AND ALBERT TYLER The third cleavage in all cases observed, both in the living and in the preserved eggs, was dexiotropic. If the subsequent cleavages alternate left and right, the 4-d cell in both types will come to lie near to the second plane of cleavage (Fig. 7, a', b'). This means that the second plane coincides with the median plane of the body, although in one type the entrance point of the sperm would be to the right of the median plane, and in the other it would be to the left. 4d FIG. 7. Diagrams indicating the position of the first cleavage with respect to the polar body, and the entrance point of the spermatozoon ; also the location of the 4-d cell. In a' the position of the 4-rf resulting from the type of first division in a is shown, in b' that in b. The Cleavage of Nereis The egg of Nereis is particularly well suited for a study of relation of entrance point to cleavage, not only because the slow entrance of the sperm makes for accuracy of observation, but also because after the sperm-head has entered, a portion is left sticking to the fertilization membrane, and, if exactly on the horizon, may be still seen at the time when the cleavage begins. The technique was the same as for the Cumingia eggs, but since the egg is larger, thicker supports made from ENTRANCE POINT OF SPERM AND CLEAVAGE 71 glass tubing were used. Owing to the great thickness of the jelly a relatively large space between the cover and slide is essential. The location of the first cleavage with respect to the entrance point of the sperm has been studied by Just. The observations reported here were made to determine not only the constancy of the relation, but also to determine whether the AB-ce\\ always forms to one side of the en- trance point — a relation not previously reported. It was found that whereas the AB lay to the right in a very large number of cases, there were a few cases where it lay to the left. Nevertheless, at the four- cell stage only one arrangement of blastomeres was found (even in those with AB to the left), namely, that shown in Figs. 7, a', or 7, b'. The first plane of cleavage coincided with the entrance point in 33 cases. In 17 cases it was less than 45°. In 14 cases it was more than 45° and less than 90°. It is apparent from these observations that the agreement (51 per cent) is far from perfect. Of the 33 cases of close coincidence, the first plane passed to the right of the pole in 28 cases, and in five cases to the left. Of the 17 cases less than 45° away, it passed to the right in 11 cases, and to the left in 6 cases. This conclusion was reached by arbitrarily shifting the entrance point to the nearest surface point in the cleavage plane. Here again there were more cases where AB lay to the right than to the left. The configuration of the cells after the second cleavage is always of the same type (Fig. 7, a', or /, //), whether the first cleavage passes to the left or to the right of the pole. In the 28 cases in which the first plane passed to the right, the Z)-cell formed away from the en- trance point and in the five cases in which it passed to the left the .D-cell formed near the entrance point. The third cleavage of Nereis, as is well known, is always dexiotropic. The succeeding divisions of the egg alternate left and right. Hence, in both sets of cases the 4-d cell conies to lie near the second cleavage plane, which Wilson has shown to be near the median plane of the body. In 1912 Just reported results of experiments on Nereis eggs, in which the entrance point was marked by the path of India ink in the ielly. He found coincidence varying from 50 per cent in one set to 60, to 80, to 95 per cent in other sets. He placed emphasis on those sets in which the greatest amount of agreement occurred. The excep- tions he supposed were due to errors of technique, since by a change in technique he found in one set of 60 eggs, 100 per cent coincidence. Our own results gave only 51 per cent exact coincidence. That the vaseline we used was not injurious was shown by removing the eggs from the slide after the 4-cell stage and finding that they produced nor- 72 T. H. MORGAN AND ALBERT TYLER mal trochophores. We tried the India ink method in the hope of ob- taining a large number of observations from a single preparation, but abandoned it because of the uncertainty in many cases of following the marker exactly to the surface of the membrane, and unless this can be done with absolute certainty there remains too great a chance of making a wrong inference, especially when the coincidence is not quite exact. In our opinion continuous observations on single eggs, while much more tedious, are safer. DISCUSSION The main interest in these observations concerns the two types of the four-cell stages in Cumingia. As pointed out, one type arises in eggs in which the ^5-cell forms to the right of the entrance point, and the other where it forms to the left. Since these two types give rise to two different planes of bilateral symmetry, on the assumption made, the problem of the determination of these planes seems to resolve itself into the problem of what determines that the cleavage plane lies to one or to the other side of the pole. Since these two types appear with equal frequency in Cumingia, it may seem that it is only a matter of " chance " to which side of the pole it passes. In Nereis there is only one type of four-cell stage and the AB-cc\] in the majority of cases (85 per cent) forms to the right of the entrance point. To this extent it conforms to the rule found for Cumingia. Since the AB-ce.ll of Nereis lies to the right of the entrance point in 85 per cent of cases, its location does not here seem to be a matter of chance. In Chatopterus there is again only one type of four-cell stage, but here the AB-ce\\ lies equally often to the right or to the left of the entrance point. Since there are here three different types of behavior leading to the forma- tion of normal embryos, it may be inadvisable at present to try to re- duce them all to one mechanism. The spiral type of cleavage common to all these eggs might incline one to attempt to find an explanation of the fact that the first cleavage plane passes to the right (with respect to entrance point) or to the left consistently in the different types. In Cumingia the egg regulates according to whether the AB-cd\ lies to the right or to the left of the entrance point. In Chatopterus, although the ^!5-cell again may lie either to the right or left of the pole there is no regulation, because the second cleavage plane coincides with the median plane. In Nereis no regulation is necessary, in this sense, in the majority of cases because these all conform to the same rule, but in the few exceptional cases the result is the same as in Chcctopterus. As already stated, an examination of the second cleavage spindle of Cumingia has not shown a spiral arrangement of the spindles. It ENTRANCE POINT OF SPERM AND CLEAVAGE 73 is equally obvious, however, that, just prior to the division, the spindle in the CD-cell lies well to one side, indicating the future position of the C-cell. After the division, the A and C blastomeres approach each other, more nearly in the polar than in the antipolar hemisphere in both types, while the B and D cells meet in a straight line at or near the antipole. If this be taken as evidence for a spiral second cleavage, then there are both leiotropic and dexiotropic second cleavages in Cnuiingia. Since the third cleavage is always dexiotropic this would contradict the " law " of alternating spiral cleavages. It has been pointed out in the text that the two types of cleavage of Ciiniingia give rise to two different planes of bilateral symmetry. In one type the median plane coincides with this first cleavage plane, and in the other type with the second. This conclusion, however, is based on the assumptions that the law of alternating cleavage holds from the third cleavage on, and that the 4-d blastomere gives rise to the germ bands. BIBLIOGRAPHY BROWNE, E. N., 1910. Effects of Pressure on Cumingia Eggs. Arch. f. Ent- wickelungsincclianick d. Organ., 29. CONKLIN, E. G., 1902. Karyokinesis and Cytokinesis in the Maturation, Fertiliza- tion and Cleavage of Crepidula and other Gasteropoda. Jour. Acad. Nat. Sci. of Phila., 12. JUST, E. E., 1912. The Relation of the First Cleavage Plane to the Entrance Point of the Sperm. Biol. Bull, 22. LILLIE, F. R., 1895. The Embryology of the Unionidae. A Study in Cell Lineage. Jour. Morph., 10. LILLIE, F. R., 1906. Observations and Experiments concerning the Elementary Phenomena of Embryonic Development in Chastopterus. Jour. Ex per. ZooL, 3. , MEAD, A. D., 1897. The Early Development of Marine Annelids. Jo-ur. Morph., 13. MORGAN, T. H., 1910. Cytological Studies of Centrifuged Eggs. Jour E.vper. ZooL, 9. WILSON, E. B., 1883. Observations on the Early Developmental Stages of Some Polychaetous Annelids. Stud. Biol. Lab., Johns Hopkins Unir., 2. WILSON, E. B., 1892. The Cell Lineage of Nereis. Jour. Morph., 6. THE EFFECT OF LOW OXYGEN TENSION ON THE PULSA- TIONS OF THE ISOLATED HOLOTHURIAN CLOACA BRENTON R. LUTZ (From the Bermuda Biological Station for Research,'1 the Mount Desert Island Biological Laboratory, and the Physiological Laborator\ of Boston University School of Medicine) The sequence of events in the respiration of Sticliopns niocbii Semper has been adequately set forth by Crozier (1916). In laboratory aquaria the rhythmic activity of the cloaca is distinctly periodic. A series of several pulsations is followed by a pause during which water is expelled from the respiratory tree. Then another series of inspira- tions begins. The number of inspirations in a series was found by Crozier (1916) to range from five to eleven, the greatest number being found in the largest animal. Pearse (1908) pointed out that, if the respiratory pulsations of Thyonc briareus are prevented for some time by repeated mechanical stimulations, the contractions which ensue when stimulation ceases are greatly augmented in amplitude. Oxygen deficiency has often been associated with periodicity and augmentation of response in various tissues. Douglas and Haldane (1909) have described periodic breathing in man under low oxygen tensions, and Douglas (1910) found the same type of breathing at high altitudes. Magnus (1904) and Frey (1923) reported that a stop- page of the oxygen supply to beating smooth muscle results immediately in an increase in amplitude. The present paper deals with the phe- nomena which have been observed on decreasing the oxygen available to a rhythmically beating isolated strip of circular muscle from the cloaca of Stichopus niocbii Semper. This holothurian is found in great num- bers in the shore waters at the Bermuda Biological Station. During the summer of 1927 the author repeated some of the experiments on a ring preparation from the cloaca of Cucumaria frondosa, very abun- dant at the Mount Desert Island Biological Laboratory, Maine. METHOD Crozier (1916) has shown that the cloaca in situ in the isolated posterior end of Stichopus will maintain its pulsations for many hours. No reference to the use of an isolated strip of this organ could be 1 Contribution number 158. 74 HOLOTHURIAN MUSCLE AND OXYGEN LACK 75 found in the literature. The present work was carried out with an opened ring of the circular muscle of the cloaca. A cloacal-end prep- aration was first made similar to that described by Crozier (1916). The cloaca was then excised by cutting the radial muscles with a scalpel and freeing the organ from the anal rim by a transverse cut. From the muscular tube thus obtained a strip was made, one to two centi- meters broad, and from four to six centimeters long. This strip was suspended vertically in a vessel of sea water by means of an L-shaped glass rod and a counterbalanced aluminum lever. A 250 cc. graduated cylinder cut off to hold about 125 cc. was found convenient as a vessel to hold measured amounts of sea water, or through which sea water could be made to flow continuously. The temperature of the water was recorded and found to vary little during an experiment, or from day to day. Therefore no special precautions for maintaining constant temperature were necessary. RESULTS Records were taken from strips of Siiclwpits cloaca beating under the following conditions: (1) in a continuous flow of sea water, (2) in a limited amount of sea water, (3) in boiled sea water with added carbon dioxide, (4) in boiled sea water of various degrees of aeration, and (5) in normal sea water with potassium cyanide added. Continuous Floiv of Sea }\\itcr. — When sea water was made to flow continuously through the vessel at the rate of about 100 cc. a minute, the strip beating therein gave a tracing which was exceedingly uniform over a period of several hours, as may be seen in Fig. 1. Both am- plitude and tone increased during the first hour. This condition was maintained for an hour or more. Then the tone began to fall very gradually while the amplitude remained about the same. After five to seven hours from the beginning, the amplitude began to decrease slightly. The rhythm was exceedingly regular and no indications of periodicity appeared. The rate of beat decreased slowly from the start, in one case almost 50 per cent after seven hours and forty-one minutes ; but the preparation was still vigorous and regular. Limited Amount of Sea ITater. — When a strip was allowed to beat in a limited amount of sea water, that is in 100 cc. without change, the amplitude began to increase in about three hours and distinct periodicity developed as seen in Fig. 2. The increase in amplitude continued for an hour or more, becoming 230 per cent in one case. The tone was maintained until the increase in amplitude occurred, when it gradually fell ; but the increase in amplitude was not entirely due to a decrease in tone since the contractions of the strip raised the lever a greater 76 BRENTON R. LUTZ N 3 HOLOTHURIAN MUSCLE AND OXYGEN LACK 77 distance above the base line than in the beginning. Finally both am- plitude and tone fell markedly. The rate of beat decreased constantly from five or six at the beginning to two or three per minute during the periods of beating. The length of the periods of inhibition of beat gradually increased to three or four minutes. Boiled Sea Water. — Sea water which had been boiled in a narrow- necked flask and cooled to laboratory sea water temperature (28° C. or 29° C. ) was used. When the muscle strip was immersed in 100 cc. of this water, the first two or three beats usually increased in amplitude, but both tone and amplitude almost immediately fell and the strip ceased beating in from three to five minutes as shown in Fig. 3, A. If the FIG. 3. A. Cloaca! strip in 100 cc. of boiled sea water. pH 8.4. B. Cloacal strip in 100 cc. of boiled sea water treated with c.arbon dioxide, pH 5.8. Aeration at X. C. Cloacal strip in 100 cc. of boiled sea water, pH 8.3. Aeration at A'. pH 8.2 immediately after aeration. pH 8.2 after 102 minutes. water was aerated within three minutes by sucking it into a hypodermic syringe and squirting it back forcibly, a partial recovery occurred, which showed periodicity at first but later an uninterrupted rhythm (see Fig. 3, C). Several attempts to bring about recovery after waiting a longer period failed. The pH of the boiled sea water (indicator method) was sometimes as high as 8.8 as compared with 8.1 to 8.3, the pH for unboiled sea water in this region. 78 BRENTON R. LUTZ Boiled Sea Water with Added Carbon Dioxide. — Inasmuch as boil- ing removed the carbon dioxide as well as the oxygen, the former was replaced by means of a carbon dioxide generator. This resulted in boiled sea water ranging from pH 5.8 to 7.7. At the latter value the strip ceased to beat in three and one-half minutes and at the former value cessation occurred in three minutes. Aeration of the water after a three minute period of cessation failed to induce recovery (see Fig. 3, B). It seems therefore that neither the lack of carbon dioxide in the boiled sea water nor the increased alkalinity was the cause of the cessation of the pulsations. A moderate excess of carbon dioxide was produced by treating 125 cc. of unboiled sea water with carbon dioxide until the pH was 7.0. This procedure was brief and probably did not remove much oxygen. In experiment 75 (Fig. 4) the amplitude began to decrease slowly after an hour, the rate decreasing gradually from the beginning. Neither augmentation of amplitude nor periodicity had appeared when the ex- periment was stopped after two hours and fifty-six minutes. When, however, an excess of carbon dioxide was produced by adding a few drops of N/10 HC1 to a preparation beating in 100 cc. of unboiled sea water, there was an immediate rise in tone and increase in am- plitude which soon gave way to a fall of tone and amplitude and finally to cessation of beat. It is therefore not probable that an accumulation of carbon dioxide in the immersion fluid as a result of tissue activity in a limited volume of water is the cause of the appearance of perio- dicity although it might be called upon to account for the increase in, amplitude. Boiled Sea Water of Various Degrees of Aeration. — When a de- creased oxygen content of the sea water was produced, either by mix- ing boiled sea water with unboiled sea water or by partial aeration of boiled sea water, the augmentation and periodicity appeared much sooner than when a limited volume of unaltered sea water was used, the onset varying from a few minutes to two hours, according to the degree of oxygen lack. In one case the boiled sea water had been stored for several hours in a narrow-necked flask with only a few square centimeters of water surface exposed to the air. A strip beating in 100 cc. of this water became periodic at once and each successive period showed an increase in amplitude which finally amounted to about 200 per cent. The tone and the rate of beat, however, fell rapidly. In another experiment in which 100 cc. of boiled sea water had been partially aerated, wave-like variations in amplitude appeared 13 minutes after immersion, and gradually developed into periodicity 53 minutes after the start. The amplitude increased from 10 mm. to 24 HOLOTHURIAN MUSCLE AND OXYGEN LACK 79 80 BRENTON R. LUTZ mm. and was still high when the experiment was stopped at the end of 93 minutes. The tone fell rapidly during the first five minutes and then more slowly during the next ten minutes after which it was un- changed. The rate decreased about fifty per cent during the first half hour and then remained constant. In experiment 62 (Fig. 5) the boiled sea water (100 cc.) was par- tially aerated. Before boiling the pH was 8.3, but after boiling and partial aeration it was 8.4. The amplitude of a strip beating in this water decreased at first with a tendency to form waves. Then for a period of 80 minutes the amplitude remained constant, but at the end of this period the amplitude began to increase, becoming 65 per cent greater than that during the previous period of uniform amplitude. Periodicity appeared in about two hours from the beginning of the experiment, the number of beats in each period ranging from ten to sixteen, while the period of interruption varied from one minute and a quarter to three minutes. The pH was still 8.4 about forty-five minutes after periodicity and augmentation were well developed. Ap- parently these phenomena were not due to increased acidity of the sur- rounding medium, nor was the cessation of beat in the cases of extreme oxygen lack due to an increase in the concentration of the salts resulting from boiling. When 75 cc. of boiled sea water were mixed with 25 cc. of unboiled water, the pH of the mixture was 8.8. The first few contractions increased in amplitude about 20 per cent, lasting for about three minutes. Then a fall occurred, and the amplitude remained uniform in height until waves in amplitude appeared in one hour indicating the onset of periodicity, which became well marked about half an hour later. At this time the pH was still 8.8. Effect of Potassium Cyanide. — When ten drops (about 0.5 cc.) of .M/10 potassium cyanide were added to 100 cc. of sea water in which a strip had been beating for a few minutes, the results were similar to those obtained with partial aeration. An increase in amplitude oc- curred within two minutes which varied in different cases from 18 to 400 per cent. The tone increased at about the same time. Periodicity occurred within fifteen minutes. In one case it began in three minutes, and the rate of beat was increased about one beat per minute for a brief period after the addition of potassium cyanide. Finally the tone and amplitude fell and the strip ceased to beat (see Fig. 6). An examination of the results presented above suggests that lack of oxygen is responsible for the appearance of the two chief phenomena noted. Since augmentation and periodicity did not occur with a con- tinuous flow of water but did occur in three hours when the amount HOLOTHURIAN MUSCLE AND OXYGEN LACK 81 of water was limited to 100 cc., one might expect that one or more of several factors were responsible, such as, an increase in carbon di- oxide, an increase in unoxidized acids, a depletion of essential ions, or a depletion of oxygen. However, when the carbon dioxide content of sea water was increased at the beginning, the phenomena did not FIG. 6. Experiment 82. Cloacal strip in 100 cc. of sea water. Time of im- mersion 4:28. At A", 4:35, 20 drops of A//10 potassium cyanide added. appear, although a temporary increase in amplitude and tone could be produced upon addition of hydrochloric acid. Moreover, in the experiments in which the phenomena did appear, the pH of the sea water was either unchanged or decreased very slightly. Since both augmentation and periodicity were made to appear much sooner when the water was partly depleted of oxygen at the beginning, or when potassium cyanide was added, the inference is that oxygen lack was either directly or indirectly responsible. When a ring preparation made from the cloaca of Cucninarid frondosa was allowed to beat in a limited volume of sea water, namely, 25 or 30 cc., periodicity appeared in 50 minutes on the average in eight out of ten preparations. Two showed no periods. Augmentation of amplitude occurred in five 'cases. When boiled sea water was used the periodicity appeared in 25 minutes on the average in 14 out of 15 preparations. One showed no periods. Augmentation of amplitude occurred in 13 preparations. These results, especially when considered in the light of the results on Stichopus, indicate that lack of oxygen is a factor tending toward an early development of periodicity and augmentation. DISCUSSION Periodicity is a part of the normal respiratory sequence of a holo- thurian, the rhythmical contractions of the cloaca being inhibited while the body muscles squeeze out the sea water from the respiratory tree 6 BRENTON R. LUTZ through the relaxed anal valve. Crozier (1916), however, found no evidence of periodicity in the cloacal-end preparation of Sticlwpus and came to the conclusion that the stimulus for spouting has its origin outside the cloaca. It should be noted that he used larger volumes of water than were used in the work reported in this paper. Appar- ently no oxygen deficiency existed in his preparations, in which the cloacal pumping probably produced a sufficient movement of water to keep it aerated beyond the needs of the preparation. Since the isolated cloacal strip will exhibit regular periods of inhibition, the inference is that a part of the normal mechanism for spouting lies within the cloacal muscle. Since periodicity is lacking with sufficient aeration and appears quickly under conditions of oxygen deficiency, one is inclined to be- lieve that low oxygen tension is a factor in determining the normal respiratory sequence in the holothurian. Periodicity is commonly observed in the respiratory activity of vertebrates, as for example in the breathing of hibernating animals, in Cheynes- Stokes respiration, and in respiration at high altitudes. The causes of this phenomenon are usually associated with the chemical conditions in the respiratory center. Most authors have offered ex- planations which concern directly or indirectly the hydrogen ion con- centration of the blood or fluid surrounding the cells. Gesell (1925), however, has called attention to the hydrogen ion concentration within the cells of the former, pointing out that when oxygen is present carbon dioxide is formed, but if oxygen is lacking lactic acid results. In either case the activity of the center increases as the acidity rises. As- suming a critical level, one needs further to call upon a mechanism for altering either the level or the acidity to account for periodic in- hibition. The augmentation of amplitude observed with a decrease in the available oxygen is in accord with the work of Magnus (1904) and of Frey (1923), who worked on vertebrate smooth muscle. Gross and Clark (1923), in an investigation on the influence of the oxygen supply on the response of the isolated intestine to drugs, stated that cutting off the oxygen resulted in a decrease in amplitude and tone. They did not comment on the immediate brief increase in amplitude and tone shown in their published graphs. The literature offers many additional observations which indicate that a certain degree of oxygen lack results in increased activity of tissue. Kaya and Starling (1,909) found that lowering the oxygen tension resulted in excitation in the whole nervous system. Sherrington (1910) found that a certain degree of asphyxia favored the elicitation of the scratch reflex, and sug- gested that the hyperexcitability of the reflex was due to oxygen HOLOTHURIAN MUSCLE AND OXYGEN LACK lack. Mathison (1911) showed by the use of hydrogen, nitrogen, and carbon monoxide that the initial effect of oxygen lack on the medul- lary centers is clearly stimulating. Gasser and Lovenhart (1914) found by the use of carbon monoxide and sodium cyanide that decreased oxi- dation stimulated the medullary centers at first but later depressed them. Kellaway (1919) demonstrated that lack of oxygen may lead to stim- ulation of the adrenal glands, and Lutz and Schneider (1919) have observed a dilatation of the pupil in men during a period of breathing nitrogen. They also presented evidence to show that the cardiac and the respiratory medullary centers in man respond very quickly to changes in the partial pressure of oxygen. A decrease in oxygen ten- sion increased the activity of these centers, while an increase in oxygen tension decreased their activity. Glazer (1929) found that intravenous injection of sodium cyanide in a dog increases the reflex response of the anterior tibial muscle, and Winkler (1929) obtained a similar effect with low alveolar oxygen tension. In the muscle-and-nerve-net preparation reported in the present paper, it appears that the carbon dioxide content and the acidity of the surrounding fluid are not primary factors in controlling its activity. This conclusion is supported by the work of Hogben (1925) who found that, on adding acid to the perfused heart of Mala and of Homants, the pH could be lowered from 7.0 to 5.6 without producing a change in the mechanical phenomena. Reduction beyond this point produced an immediate effect on the character of the rhythm. Nor could any alteration be noticed in the beating of the smooth muscle of Helix and of Aplysla on changing the pH from 7.0 to 6.0. In fact it is possible that the pH outside of the cell may vary markedly without greatly al- tering that inside of the cell. The oxygen tension appears to have some influence on tissue acidity. Frey (1923) presented evidence which shows that without oxygen the tissue rather than the surrounding fluid first changes its hydrogen ion concentration, and if this approaches the optimal value, an increased ability to respond ensues. The anaerobic production of acid in cellular activity and the role of oxygen in the re- covery process suggest that oxygen lack is acting indirectly when cellular activity is first increased and is subsequently depressed. SUMMARY 1. An isolated muscle strip from the cloaca of Stlchopns mocbll Semper and a ring of muscle from the cloaca of Cncuinariu- frondosa were used in sea water as rhythmically beating preparations. 2. In a continuous flow of sea water the contractions (Stichopus) were nearly uniform in rate and amplitude over a period of several hours, but a gradual decrease in both finally occurred. 84 BRENTON R. LUTZ 3. In a limited volume of sea water (100 cc.) the amplitude (Stich- opus} began to increase after three hours and a distinct periodicity of the regular rhythm developed. In the case of the cloacal ring of Cucumaria beating in 25 or 30 cc. of sea water, periodicity appeared in 50 minutes on the average. 4. In boiled sea water the strip (Stichopus) ceased beating in from three to five minutes, but partial recovery took place if the water was aerated within three minutes. If the carbon dioxide was replaced in the boiled sea water, cessation of beat occurred as before. In 25 or 30 cc. of boiled sea water the ring of cloacal muscle from Cncumaria developed periodicity in 25 minutes on the average. 5. A moderate excess of carbon dioxide in sea water (pH 7.0) did not bring on augmentation nor produce periodicity. 6. In boiled sea water of various degrees of aeration the augmenta- tion of amplitude and the periodicity appeared sooner than in unboiled sea water. When little oxygen was present both phenomena appeared almost immediately, while the pH of the surrounding fluid was un- changed. 7. When potassium cyanide was added to the sea water an increase in amplitude and tone occurred and periodicity appeared. 8. Evidence from the literature is cited supporting the view that decreased oxygen tension results at first in increased activity of mus- cular and nervous tissues. This view is further supported by the evi- dence presented in this paper.2 BIBLIOGRAPHY CROZIER, W. J., 1916. Jour. Ex per. Zool, 20: 297. DOUGLAS, C. G., 1910. Jour. Physiol, 40: 454. DOUGLAS, C. G., AND HALDANE, J. S., 1909. Jour. Physiol., 38: 401. FREY, W., 1923. Zeitschr. f. gcs. E.vpcr. Med., 31: 64. GASSER, H. S., AND LOEVENHART, A. S., 1914. Jour. Pharm. E.rpcr. Thcrap., 5: 239. GESELL, R., 1925. Physiol. Rev., 5: 551. GLAZER, W., 1929. Am. Jour. Physiol., 88: 562. GROSS, L., AND CLARK, A. J., 1923. Jour. Physiol., 57: 457. HOGBEN, L. T., 1925. Quart. Jour. E.rper. Physiol., 15: 263. KAYA, R., AND STARLING, E. H., 1909. Jour. Physiol., 39: 346. KELLAWAY, C. H., 1919. Proc. Physiol. Soc., Jow. Physiol, 52, Ixiii. LUTZ, B. R., AND SCHNEIDER, E. C., 1919. Am. Jour. Physiol., 50: 327. MAGNUS, R., 1904. Pfliiger's Arch., 102: 123. MATHISON, G. C, 1911. Jour. Physiol, 42: 283. PEARSE, A. S., 1908. Biol. Bull, ~15: 259. SHERRINGTON, C. S., 1910. Quart. Jour. E.rper. Physiol, 3: 213. WINKLER, A. W., 1929. Am. Jour. Physiol. 89: 243. 2 The writer wishes to express his thanks to Dr. Edward L. Mark, who generously accorded the privileges of the Bermuda Biological Station, and to Dr. H. V. Neal for the many courtesies extended at the Mount Desert Biological Laboratory. PHENOTYPICAL VARIATION IN BODY AND CELL SIZE OF DROSOPHILA MELANOGASTER W. W. ALPATOV (From the Institute for Biological Research, Johns Hopkins University) I. The purpose of this paper is to contribute to the solution of the question of the relationship of the cell size and body size, using well- known and standard material. The literature devoted to this question is very extensive, but most of the work done cannot be considered to fulfil the requirements of exact experimental investigation, in regard either to the control of conditions, or the homogeneity of the material, or the precision and accuracy of the treatment. Comparatively modern compilations of the data available have been made by Levi (1906) and Martini (1924). Concerning the more limited problem of the correlation of body size and cell size in Diptera there have been two recently published papers. Loewenthal (1923) attacks a problem which corresponds to one part of the present investigation, namely the influence of underfeeding on the body and cell size of the blow-fly. The first criticism which may be made of Loewenthal's work is that he does not give any indication of the ages of the normal and underfed maggots. It therefore is not clear whether the observed smaller size of the hypodermis cells is due to differences in the age of larvae or in the feeding. At the same time Loewenthal does not find any difference in the cell size of the gonad rudiments, in spite of their difference in size. The following conclu- sion is reached (p. 91) :" Danach ist die Korpergrosse der ausgebilcleten Imagostadiums unabhangig von der Zellgrosse und allein bedingt von der mehr oder minder grossen Zellanzahl." Further a totally incorrect statement is made concerning the absence of cell divisions during the larval life (p. 92) : " Mit Abschluss der Embryonalentwicklung stellen die larvalen Zellen ihre Vermehrungstatigkeit ein, das ganze Wachstum der Larve von wenigen mm Lange nach dem Schliipfen aus dem Ei bis zur Lange von 2 cm einer verpuppungsreifen Ruhelarve beruht allein — wenn man von den wahrend der Larvalperiode fiir^lie Gesamtgrosse nicht ins Gewicht fallenden Imaginalanlagen*absieht — auf dem Gros- senwachstum der Zellen." Przibram's and Megusar's (1912) investi- gations showed that this is not the case in the postembryonal develop- 85 86 W. W. ALPATOV ment of Sphodromantis (Orthoptera, Mantidae) and I (1929) have shown also that the metamorphosis of Drosophila is connected with six simultaneous divisions of the cells of the whole body. The same subject of the relationship of the size of an organ and the size of the cell has been touched upon by Bridges (1921, 1925). In both of his papers differences in the cell structure, namely, nuclear structure, are shown to be connected with the size of the whole body and its organs. It was discovered that these intersex-producing females (triploid) could be identified by their somatic characters, namely, large coarse bristles and large roughish eyes (1921, p. 253). In the second paper it is the size of the ommatidia which is shown to be different in flies having different chromosomal complexes. ' The cells of triploid individuals are readily seen to be larger than the cells of diploids, and correspondingly their facets are larger" (Bridges, 1925, p. 709). I became interested in the problem of body size and cell size years ago while working on the oceanographic expedition of the Floating .Marine Scientific Institute to the Russian arctic seas. The first ex- pedition in 1921 gave ver)^ impressive material on the geographical variation in the dimensions of the body of different marine animals. It could be particularly easily shown on such a group of animals as Isopoda, which have a postembryonal development ending with a definite imaginal stage analogous to that of insects. Extensive biometrical data on variation of Isopoda, taken from localities with different tempera- tures, showed perfectly that the colder regions (for instance, the Kara Sea) are populated by races which have a larger body size than regions with warmer water temperature (Barents Sea). On the second expe- dition I strove to collect some material on the histology of local races of some of the species of Isopoda. But the severe conditions of naviga- tion during this and following summers did not allow the accomplish- ment of this intention. During the winter of 1927-28, working at this Institute, I succeeded in working out a more or less accurate method of producing Drosophila imagoes of different sizes, using two factors, temperature and underfeeding. The method of counting the number of hairs on the wings of Drosophila as a method of estimating the number of cells on a certain surface of the wing was discovered by a friend, Dr. Th. Dobzhansky (1929), who was kind enough to explain it to me. I have the pleasure to express also my deepest gratitude to Dr. Raymond Pearl for criticism and valuable suggestions. II. Two factors have been used in producing flies of an abnormal size. It was shown in an earlier paper that the first of them was the low tern- DROSOPHILA MELANOGASTER 87 perature, which decreases the rate of development and produces flies of a larger size (see Alpatov and Pearl, 1929). The method of collecting new-born larvae has already been described (Alpatov, 1929) . Flies belonging to Wild Line 107 have been taken for parents of our ex- perimental animals, the collected larvae being 0-2 hours old at the moment of putting them on food. The bottles had been planted with yeast 2 hours before the putting on of larvae, and watered with a few drops of distilled water. Electric and low temperature Hearson incu- bators were used for keeping the bottles with flies. Five bottles with 50 larvae each were kept at 18° C, five others at 28° C. The develop- ment from the moment of the populating of the bottle till the moment of the pupation was more than twice as long in the cold series as in the warm. It is unnecessary to discuss here at length the question of temperature and development rate, this having been done in another paper (Alpatov and Pearl, 1929). The technique of breeding in the experiment with underfeeding was the same except for the fact that the yeast was put in the bottles with synthetic medium at the moment of populating the bottles with larvae. A method of getting undernourished larvae by taking larvae from the food before the normal end of feeding has been used by various workers, for instance, Ezhikov (1917, 1922), Smirnov (1926, 1927), Cousin (1926), Herms (1928) and others. Most of these authors did not attempt to determine with sufficient accuracy the moment of taking the larvae from the food, Herms being in that respect an exception. In the present investigation, larvae were taken from the food exactly 48 hours after the moment of populating the bottles with 0-2 hour-old larvae. Larvae which reached the desired age were taken from bottles and placed in half-pint bottles containing plain agar. The mouths of the bottles were covered with 40 mm. watch glasses and sealed with plastaline used in modelling. This was done in order to prevent the larvae, which become very active, from crawling out. The day after the larvae had turned into pupae the watch glasses were replaced by the usual cotton stoppers. Table I shows that the larvae with a subnormal period of feeding pupate earlier than normally fed ones. This can be compared with Kopec's (1924) statement that ". . . if we begin to apply starvation to older specimens during developmental stages . . . the transformation of these animals is accelerated." A little longer prepupal development of the normally fed larvae, those which served as controls to the underfed being compared with the 28° flies of the early October experiment, cannot be very easily interpreted. It might exist in a difference in conditions — perhaps a difference in yeast growth which lengthened the duration of development of larvae in the second set of experiments. 88 W. W. ALPATOV TABLE I Data on the Conditions of the Development of Flies Reared for the Study of the Problem of Cell Size Temperature limits of variation Average Time of the beginning of the ex- periment Time from egg until pupation, in hours Time of feeding Underfed flies Kept at 28° — October 24, 1928 80.39±.50 48 hours Normally fed flies Kept at 28° . October 24, 1928 93.16±.74 Until normal leaving of the food 28° flies 27.1-28.9° 28.2° October 8, 1928 87.40±.36 Until normal leaving of the food 18° flies 17.0-20.0° 18.2° October 8, 1928 200.86±.89 Until normal leaving of the food The flies have been collected in 70 per cent alcohol and measured in glycerine under a cover glass. The following characters on the wings of collected flies have been studied : the length and width of the wing, and the number of hairs on a surface equal to 0.1 square mm. on the lower surface of the wing. Fig. 1 represents the points of measure- ment and the place where the hairs have been counted. Region of 6r/stte counting FIG. 1. Measurements of the wing. AB, length of the wing, CD, width of the wing. The square shows the area of the bristle countings. For the measurements the following optical systems were chosen : Spencer 25.4 mm. objective and a micrometer ocular in a No. 2 ocular. The countings of the hairs of the lower surface of the wing were done in a way approved by Th. Dobzhansky. Pieces of paper with squares DROSOPHILA MELANOGASTER 89 representing 0.1 square mm. at a given magnification have been pre- pared by projecting through an Abbe camera lucida 0.1 mm. from an object micrometer placed on the microscope stage. A Spencer microscope was used with objective 4 mm. and ocular [ 10. The hairs have been projected by means of the camera lucida and drawn with a sharp pencil. Only hairs whose bases happened to fall inside the square have been counted (Fig. 2). A/or/na/Sy fed a £o/d fe/nperafare \ OOO_ T / W pq S •tt o > ^ s « 5 ^s •+<. s o *o o '— ' I- i = O O O e^ -H § -H _jj ^ -H **• 41 41 °^ "H 3 °- S3 S ^ 10 • ON ON o OO NO -NO »H O -• —' 1 C "° > r '> 03*0,,^^ 1 C K> oj T3 VtJ 4; *•" c* *"S ^ 03 C 0 41 rt C ° § T-J o c'-DU ^ -S .2 r ; C +-1 ^^ S T3 .2 r i C -M W 03 03 03 03 03 ea CO C/3 CO _ 1 o 1 HI r- IT +J 03 •S.2 **— ' CN IH CN bjO o3 C O. C +j c 'C 3 JJ 03 . 03 ^j 3 a. —<•_-. c/3 V 03 Oj -a -w O 3 O O ^ W^-*3 *~ V C - "5 ON fee S* ^ G, J ^ 5 CM !5; OH !?; c; a 1 ** o >o a'S JJ "^ ~4H M 01 a 1 ' II U 1-1 t~" CV O) as 3 O tN (V O ci t>^ 3 8 2 _H 1O §q f^ r^ JJ \O -U rs 00 o °° •o • O JQi f5 ^^ O^ ^ o T— I C »•* OJ c > rt T3 •— 4J 'l r- ^O ^^ o3 *— ' ^) r" O ^H o3 s r^ * r"! C^ j "c '-2 U 03 03 03 rt CO CO C 00 c OO o 00 *-w OO *•" ^11 03 1 " 00 uj IH -o 00 ^ _j 3 c ^ _ 2 03 03 ui IH •4-J O O o , o ON 10 C 7 o io ni t^» 1) IH fli £ JS C || tg ^ ON * * fe s OH ^^ DROSOPHILA MELANOGASTER 95 • CQ >o CM •fl **> ~H t~- o ON o • ^°. 5S o ^ o "-; 4^ § Jj 10 «N O ^ ** o ^ ^ " 1 '> C K"* •o 1 '> C t^ nj "O i^ 03 "P — rt 'H — ~ rt g ° o b e c ^— i 03 tj S "O .2 ' 1 -S § ° c -SU 03 "3 03 o3 03 03 C/5 C/5 ft ,_ 1 Q 1 aj r1 " ON )_' JS 0 2 i nj ON 4-> • — *-*-< CN u 1 bC "rt ON C ^ T3 C ON .S a S rt ' nj 4-> 3 a "M "O V- OJ 03 j^ •a 2 0 3 ^o o >o U 0) si -4-* «^ _ O *^H 1 ^> 2* vo ii ^y 11 03 * K^. 5 £ fe; £ ^ e OJ C5 O 10 8 e'.3 4 •H to r^ O S p 5 O OO ° 10 ^ ^O t^* 82 o oo S.. 1 8 -H ii -H S -H ^f O ^ ^H" p - > C ^ > n> 03 "O i_ 03 "O ^_ w o3 C O ^ « c o "c '£ U S -o .2 , ; c -3 U ^— ' C/5 o3 03 4*i SoO o oo O QQ *+J OO ^™ 1 03 1 ^44 u OO •o oo S "*~* C tx 03 • 03 "O u 03 0) 03 c/J 1 -M o 3 £ 0 1 CJ C CL> si ^ "S § " U g.0 || fe; S fe; 96 W. W. ALPATOV and the weight of the chrysalids in Lynmntria disbar (L.). This nega- tive correlation found in twelve experimental groups out of sixteen is particularly well expressed in males. s/./ 3ff W W #2 43 #7 ¥8 IfS SO S/ 52 S3 SV 5S 56 S7 SS S9 60 61 62 63 Length of Me using /n /n/cro/neter diws/ons FIG. 5. Correlation between the length of the wings and the number of bristles per 0.1 mm.2 on the lower surface of the wings of the male. s/./ J9 VO Vt V2 «3 «¥ US V6 V? YS V9 SO S/ 52 53 S¥ ff f6 S? 58 59 SO 61 62 63 6^> s •- 03 ^J s •.: (fi W ~ "c -ii ^"^ i^^ ^J ^H X. ^ « < ^ 5 S- ^ s? ^j O ^ ^ ~- G <^ •*-_ ^ C c «*^ '- S "o"" _ cc . 1 O q w £.£ £ IF £ tN q 00 ^H i CJ ' -U <£ oo o 00 "o £ £ i- ~ g -H o 8 lO •ti o d ~ o o rrj' T— < 25£ d 10 <_ c 0 o O'" ^^ 1— 1 _ CO . •z H P 60 j; j; 0 o " ^ oo oo CN q •H PC ) *~ rri •o 4-1 ^ PC -HI o 0 ". SO § || c D "o^ £ -H o q o 00 •H Tf d a^ 1 2"3 o, S ^ _g ."s o> TD c u "O > * CS C "3 cd S co U X.' O OO O oo •H ?< •H ^» q o «— i sq PC -5 rsi Cs q 4^ '0 ^_, "* >o o o j: 4^ d "1 -f >o ' so 4s* C OO q 0 -H fN 4i sO q IO 1 1 i 0 o Hj 4^ rC q Os q q 10 PC ^-< % o d PC PC JQ d IO 1 o l-H -U Os -^j PC O Os sO 00 >0 5 °°' OO sO $ 10 " d s$ 4i OO q 4^ 1 >o . *^ Os OO ^ c _o -4— > .2 o u c •c "^ s C a ^o ^ c/5 U >J 98 W. W. ALPATOV IV. Correlation tables shown in Figs. 5 and 6 contain the basic data on the number of hairs on 0.1 mm.2 and length of the wings. The horizontal axis gives the wing length in divisions of the ocular microm- eter, each division being equal to 28.333 microns. Table VI represents constants derived from Figs. 5 and 6 with the addition of wing length of 28° flies. The wing length is expressed in millimeters. TABLE VII Average Width of the Wings and Width Index, i.e., Width Expressed in Per Cent of the Length Underfed flies Normally fed flies 28° flies 18° flies Width of the Index Width of the Index Width of the Index Width of the Index wing wing wing wing Males . . . .7151 58.23 .8970 59.33 .8871 ±.0021 59.62±.ll 1.024 58.85 Females . .7253 57.29 1.015 58.30 1.004 ±.002 58.51±.12 1.102 57.31 Let us discuss the influence of the factor under consideration on the wing as a whole. Table VII gives us the constants for the width in millimeters as well as the width in percentage of the length. There is a pronounced sex difference in the size of the wing, the females in all groups being larger than the males. The relative width of the wing is larger in the males, as can be seen by comparing males and females in all groups, and particularly those of the 28° group. The difference is 6.9 times larger than its probable error. (The indices in this case have been calculated by the use of Pearson's formula.) Another point of interest concerning the relative width of the wing is that in the fe- males as well as in the males the underfed and 18° flies seem to have narrower wings than the " normal " 28° flies. The sex difference is also influenced by abnormal conditions. Table VIII shows that in " normal " 28° conditions the sex difference is the greatest, while un- derfeeding and low temperature reduce the difference. The lower line in Table VIII contains recalculated data from the experiment described in a former paper (Alpatov and Pearl, 1929). The effect of low tem- perature and consequently of the slow development can be seen in this case also. It is difficult to find an adequate explanation of this phe- nomenon, which very likely is connected with certain differences in male and female postembryonal development, that is, with different time of the manifestations of different characters during the larval or pupal life. DROSOPHILA MELANOGASTER 99 Turning our attention to the main problem of our investigation, one glance at the correlation tables shows that the larger the size of the wing of the corresponding group of flies, the smaller the number of cells on the area of 0.1 mm.2 In other words, the larger flies, con- sidering inter-group variation, have also larger cells. The coefficients of correlation for each of the six groups of flies have been calculated separately. They are given in Table VI. Only in the case of underfed males and females is the correlation significant and negative. The con- clusion is that in underfed flies the size of the body is negatively cor- related with the number of cells on a definite surface of the wing. A possible but very dubious explanation of the absence of such correlation in the case of normally fed and cold temperature flies might be that the variation in the wing length of Drosophila developed from normally fed larvae is so small that the correlation could not manifest itself. TABLE VIII Sex-Index of the ll'imj Lcntjth, i.e., Male Wing Length Expressed in Per Cent of the Female When studied Underfed flies Normally fed flies (28°) 28° flies 18° flies 1928 95.33±.54 86.76±.15 86.71±.15 90.61 ±.26 1927 — — 88.18±.16 91.93±.18 So far as the variation of the flies belonging to different groups is concerned, it can be seen that the coefficient of variation of the number of cells does not show any definite difference in different groups. At the same time the variation of underfed flies in the length of the wing is much greater than that of the flies which had a normal feeding, no matter at what temperature. Previous investigators who have worked on variation of flies under conditions of under-feeding have also de- scribed the increasing variation of experimental animals (see Smirnov and Zhelochovtsev, 1926). We have now to approach the problem of the actual surface-size of the cells and its relationship to the size of the whole organ. Table IX represents all the calculations relating to this question. The surface of a cell in square microns was determined by dividing 10,000 microns (0.1 mm.2) by the number of hairs on that surface. It can easily he seen that the larger the flies the greater the surface of the cell. An- other point of interest is the pronounced sex difference in the size of the cells, the females having much larger cells than the males. This has been pointed out by Dobzhansky (1929). The next step was to 100 W. W. ALPATOV en S oo X 5 H-H U ^ « 'I H '? •>, 1/3 a ^ — c "o a; flj " cd r- ca c^ 3 "&> i_ r_^ C 3 co co cd u OJ a; UH o o 0 "— ' o q NO o o ON ON 00 ON oo' NO NO 00 NO NO *-H 00 3 ni ni ^ LO uo Is* s CN CN T-H ON -H oo o oo 0 CN O 4} T— 1 HJ fN Hi o Hi r— CN T— 1 3 OO o 4} o o q 00 ON o "I CN O oo' OO CNJ T— ( "o 4— ' be C II OJ "a; cd o u rt 3 •-H- 10 O bit) J3 -c C ra -4—i ""o •M c « 0) J3 o -o 'o 8 OJ "o Surface a -s CO Calcula •4-J C o DROSOPHILA MELANOGASTER 101 come from the surface values to linear values which has been done by calculating the length of the cell, which was obtained by taking a square root of the surface of the cell. The data on wing length gave the possibility to calculate the percental decrease in the wing length taking the wing length of 18° flies as a basis. Multiplying by this percental decrease in wing length the number giving the " length " of cell in cold temperature (18°) flies, we obtained the figures represented in our table under the heading " Calculated length of cells." Comparing them with the dimensions obtained by taking the square root, we can easily see that the assumption that the wing length varies proportionally to the length of its constituents does not hold true. The three columns on the right of Table VII represent the changes in wing size and cell size expressed in per cent of 18° (cold) flies. The same relationship between these two characteristics is shown in a percental scale on Fig. 7, the diagonal line represents the relationship in case of a proportional change in wing length and cell length ; the dotted line shows the actual percental decrease in cell size in different groups of our flies. SO 6O 7O SO SO /OO /O 2O ~30 /O 20 JO SO 6O 7O 8O SO SOO Length of Me Length of the FIG. 7. The dotted lines represent the relationship between the percentage of decrease of the wing length and the percental decrease in the corresponding percental length of the cells calculated by taking the square root of the surface of the cells. The general conclusion of all these calculations is that the reduced size of cells alone cannot explain the reduction of the organ. The only possible way to explain it is the assumption that the decrease in the organ size — in our case in wing size — is not the result of a decreased size of its cells alone, but also of a reduced number of cells. This last conclusion has a certain bearing upon the problem of the cell constancy in the organism. 102 W. W. ALPATOV If our discussion is correct, the organism can evidently react to the factor decreasing in size not only by decreasing the size of the cells but also the number of cell divisions. The present limited material does not warrant further discussion, but it may be hoped that other investigations in the field of cell-biometry may create a similar basis for understanding the variation of the whole organism as Die Zellulars Pathologic of Virchow did for the interpretation of the pathology of the whole organism. SUMMARY 1. Dobzhansky's method to determine the number of cells under the surface of the wing membrane of Drosopliila melauogastcr by count- ing the number of hairs has been used in the present investigation of the relationship of the organ size to the size of its cells. 2. Underfeeding and development at low temperature have been the factors to produce flies under and above the normal size. 3. The functional relation between the time of feeding of larvae and the size of the wings of larvae being the expression of the upper part of the logistic larval growth of the third larval instar can be expressed by a cubic parabola. 4. There is a definite tendency for large larvae (i.e., fast-growing ones) to pupate earlier, which finds a certain analogy with Pearl's cor- relation that " between growth rate and duration of life (in this case, duration of larval life) to the beginning of death the correlation is negative and significant in degree." 5. As far as all three groups of flies (underfed, normal and cold flies) are concerned the size of the wings is negatively correlated with the number of hairs on a definite surface of the wing when the groups are considered as ^vholcs (inter-group correlation). The existence of such a negative correlation could be shown also within the group of underfed females and males, but not within the other groups. 6. Expressing in per cent the increase in size of the whole organ and the increase of the linear dimensions of the cells there is a dis- crepancy in the rate of changes. This leads to the conclusion that the changes in size of the wing cannot be accounted solely by the changes in the size of the cells. The number of cells must play also a certain iole in this process. LITERATURE ALPATOV, W. W., 1929. Growth and Variation of the Larvae of Drosopliila wclanogaster. Jour. E.rp. ZooL, 52: 407. ALPATOV, W. W., AND PEARL, RAYMOND, 1929. On the Influence of Temperature during the Larval Period and Adult Life on the Duration of the Life of the Imago of Drosopliila melanogaster. Am. Nat., 63: 37. DROSOPHILA MELANOGASTER 103 BRIDGES, CALVIN B., 1921. Triploid Intersexes in Drosophila mclatwgaslcr. Science, 54: 252. BRIDGES, CALVIN B., 1925. Haploidy in Drosophila mclanogastcr. Proc. Nat. A cad. Science, 11: 706. COUSIN, G., 1926. Influence du temps reserve a la nutrition sur les phases du cycle evolutif et les metamorphoses de Calliphora erythrocephala. Com[>t. rend. Soc. Biol, 95: 565. DOBZHAXSKY, TH., 1929. The Influence of the Quantity and Quality of the Chromosomal Material on the Size of the Cell in Drosophila melanog aster. Arch. f. Entit'icklngsnicch. d. Ore;., 115: 363. EZHIKOV, J., 1917. Influence de 1'inanition sur la metamorphose des mouches a ver. Rev. Zool. Rnssc, 3. EZHIKOV, J., 1922. Uber anatomische Variabilitiit iiber dirckt Wirkung ausserer Einfliisse. Rev. Zool. Russc, 3. HERMS, W. B., 1928. The Effect of Different Quantities of Food during Larval Period on the Sex Ratio and Size of Luc ilia scricata Meigen and Thco- baldia mad ens (Thorn). Jour. Econ. Entom., 21: 720. KOPEC, S., 1924. Experiments on the Influence of the Thyroid Gland on Meta- morphosis and Weight of Insects. Mcmoircs de I'lnstltut national poJonais d'economie rnrale a Pullaivy, 5: 356. LEVI, G., 1906. Studi sulla grandezza della cellule. Arch. Ital. siETArraRpt-taaia e« urn eon 1928 FIG. 1. A curve showing the relation between average length and weight of larval specimens of A. maculatum at various ages. Data of 1928. yolk mass. Increase in length of this axis during the early growth stages does not result in equivalent increase in the total length of the embryo. It is not until the embryonic axis becomes straightened that the total length shows marked increase. AMBYSTOMA MACULATUM 187 That the departure from a single sigmoid curve during the early development is not a distinct phase of growth may be demonstrated. When an " index of build " (Length VWeight) is computed (Table I), it is clearly indicated that length and weight are not directly associated during the early stages. This index varies constantly to a period shortly before hatching when the embryonic axis becomes linear. It is fairly constant, however, for the free living larval stages. A group of experiments carried under approximately constant tem- perature conditions gives indication that the degree of curvature may vary under environmental circumstances. Four groups of salamanders FIG. 2. Curve of growth showing weight and linear increment. Data of 1929. at the neurula stage were placed at approximately constant temperatures of 4° C. ± 1, 13° C. db 1. 19° C. ± 1 and 27° C. ± 1. When the four groups of length-age data acquired from these animals were plotted in such a way as to rule out the change in growth rate due to temperature, that is, when all the data, after allowance is made for appropriate thermal coefficients of growth, are plotted as though the animals were raised at 19° C., the shape of the curve is not the same for each group. The linear increase in the 4° sample was practically n 188 W. T. DEMPSTER 0-. GE ST bfl 1 - ti,ro O — < ON -H ro >O r_ •— »^I f»5 _LJ _U "5 -* -f -t< ^ _L_ ON 1 1 1 n Tl | | || 1 O IO NO NO -t* ON CN CN >O NO r^.-H.-HCNror«Of5r<5 -t -f -f -f a V ST. S3 a ••-* o <• CN — i xo *O ro O ON "0 -foo^f-tr— OOONOO-f t-~ T-H O O ON 01 O •— i 01 O] 01 O NO »— I NO IO »-H ON IO -f NO NO CN "5, *•-* I a T-H t^. CN *-H ON ON ON CN OO O "0 IO >O ON NO T-H T-H CN *O ON l^~ CN T-H r-~ ON IO OO O OO ON ON rr}rNOCNCNCNCNT-i ON ON O I-H •— i -H -f Tf t-~ w j PS •Ci s «i ^ §0 a: w H OOO\'*OOO*OVOON»— <*— i •»— i \O *O *O \O o o IO CN 00 CN "0 uo.^ON— i rt< OO --H f«. VO -*NONOONOON--HT-HCN T-H T-H T-H T-H a H ro ^t* OO t^» NO O f^O •— i CM rj< ON •— i O O 10 o CN NO 'O >O 0 ON CN ON -t1 — H 10 O CO >-H IO T— I T— I CN CN NO OO "^ ^i* ^O t^- O ON *-H ON CN t~** ^^ ON •^ NO NO NO NO NO UO K to H 0 H £ T-H T-H OO OOIOOOOOO •-H -^f OO ON OO •— i »— i O oo' IO r— CO ~£ »— ' ^O CN *^0 IO ^^ ^O NO 1 00 T-H r-f. ^H O NO -f >O CN O OO O fO O O NO IO NO IO *— I ON *-H *^ 10 ON CN oo *O ON OO CN *O Tf t*"~ T-H CN ^^ ^0 OO »~H r*o NO ^^ ON CN OO OO -~< T-H T-H CN CN O) ~^ IO *O t^^ t^* J"^ t~^ f^» t"» 8 ex. X O J z H IA OO'— i^-i lOOrO-tONrOCNl/OOOOONCN T— IT— iT-HT-HCNCNCNCNCNOlCN CN Tf T-H r^- oo NO O* ON t~*» T-H NO CO NO to 4i ^ 4i 41 4i 4i 41 '— i •— i IO »— i •— lOO-rfOOOCNOO OOOCNl^oV>-Ht^lOONlOt— O-) wo ~^ ON ON CN f~-> ON T-H 10 f^- OO OO f^ OO Cvicv]rs,CNCNrororO'*-tl'^. Zool.. Vol. 55, 1930). 190 W. T. DEMPSTER relatively large specimens (Fig. 1, a, a', b, b'}. It seems quite prob- able that these specimens had not yet entered the third period of growth, i.e. the terminal period of slow growth. During the second year, when the pond became dry, two records, one before the pond became dry and the other immediately afterward, are available on the length and weight of recently metamorphosed specimens. In the first of these both values are higher than in the second. The second group was un- doubtedly " forced " by the drying of the pond to metamorphose before reaching the stage at which the first group metamorphosed. Alice (1911), who has studied the seasonal succession of pond fauna, has indicated 'that there is a periodic change in numbers of species and individuals found in forest ponds. There is an increase in numbers of species which is slow during the spring months and rapid in early summer, less marked in July and in late August the number falls to the spring value. There seems to be a correlation between the period of highest productivity of the pond as reported by Allee and the period of rapid growth of the salamanders recorded here. 1929 FIG. 3. Graph showing the relative percentage of water and organic sub- stance in larval salamanders of different ages. Data of 1929. Relations of Water, Solids and Ash to Growth. — The eggs shortly after they were laid had a weight of 7.32 mgm. consisting of 4.98 mgm. of water and 2.34 mgm. of solid, of which .097 mgm. was ash. During the embryonic period the dry weight was fairly constant. Actual in- crease was associated with increase in inorganic matter and water (Table I). The ash percentage, however, was practically constant while the water increased in this period from 68 to 94 per cent. After the animals began to eat, the dry weight increased considerably so AMBYSTOMA MACULATUM 191 that the percentage of water decreased. To the period of metamor- phosis there was a gradual increase of inorganic matter from 1 to 2 per cent. Water per cent decreased from 94 to 85 per cent and the percentage dry weight increased from 6 per cent to 15 per cent. This relationship is brought out in Fig. 3. Until the animals began to feed, growth was purely a process of hydration ; afterwards it was due both to imbibition of water and to increase in organic and inorganic ma- terials. These findings are in accord with the work of Davenport and Schaper on the Anura. Recently metamorphosed specimens showed still further decrease in the percentage of water content. Data on the percentage of water in older metamorphosed specimens (Table II) show that this early decrease may be later compensated. The pro- portion of dry weight, ash and water, however, seems to be variable for the land forms. Two specimens from an indoor aquarium in December showed a decrease in water content to 80 per cent body weight and an increase in inorganic matter to 4 per cent. TABLE II Showing the relative content of water, solids and ash in terrestrial stages of A . maculatum. LENGTH WET WEIGHT LENGTH3 WATER DRY WEIGHT ASH WATER DRY WEIGHT ASH WEIGHT mm. grams grams grams grams per cent per cent per cent September 1929 50.5 .603 214 .491 .112 .011 81.51 18.49 1.77 71 1.659 218 1.412 .247 .031 85.12 14.88 1.86 76 2.504 177 2.206 .298 .032 88.10 11.90 1.27 82 4.616 120 4.063 .553 .084 88.02 11.98 1.82 88 4.709 146 4.012 .697 .101 85.10 14.80 2.14 93 5.602 144 4.955 .647 .102 88.45 11.55 1.82 104 5.875 193 4.939 .936 .124 84.07 15.93 2.11 December 1928 136 9.432 269 7.848 1.584 .427 83.21 16.79 4.53 138 8.792 301 7.044 1.748 .366 80.12 19.88 4.16 SUMMARY 1. Growth in weight of embryonic and larval Ambystoma maculatum from the time that eggs are deposited to the period of metamorphosis may be expressed as a single sigmoid curve. 2. The length curve, except for a short period before hatching when the embryonic axis is curved, is likewise sigmoid. 3. The Ambystoma population of a pond is quite homogeneous, the specimens metamorphosing at approximately the same time. 192 W. T. DEMPSTER 4. Under natural conditions the relation between weight and length from year to year seems to be constant during the stages before feeding. Later the relationships are variable because of feeding differences. 5. Growth to the time of food ingestion is associated with imbibition of water. Later growth to the time of emergence of the salamanders is correlated with a process in which the percentage of water content decreases. During this period the inorganic constituents gradually increase. LITERATURE CITED ALLEE, W. C., 1911. Seasonal Succession in Old Forest Ponds. Trans. III. Acad. Sci., 4: 126. BISHOP, S. C, 1926. Notes on the Habits and Development of the Mudpuppy, Necturus maculosus (Rafinesque). Bull. N. Y. State Mus.. No. 268: 5. DAVENPORT, C. B., 1897. The role of water in growth. Proc. Boston Soc. Nat. Hist., 28: 73. DAVENPORT, C. B., 1899. Experimental Morphology. Part 2. Effect of Chem- ical and Physical Agents on Growth. Macmillan and Company. EYCLESHYMER, A. C., AND WILSON, J. M., 1910. Normal Plates of the Develop- ment of Necturus maculosus. Normcn. zur. Entivick. dcr Wir. Heraus. v. F. Keibel. Gustav Fischer. PATCH, E. M., 1927. Biometric Studies upon Development and Growth in Amby- stoma punctatum and tigrinum. Proc. Soc. E.vp. Biol. and Mcd., 25: 218. ROBERTSON, T. B., 1923. The Chemical Basis of Growth and Senescence. J. B. Lippincott Co. SCHAPER, A., 1902. Beitrage zur Analyse des thierischen Wachstums. Arch, f. Entzv.-Mcch., 14: 307. SMITH, B. G., 1911. Notes on the Natural History of Ambystoma jeffcrsonwnum, A. punctatum and A. tigrinum. Bull. Wis. Nat. Hist. Soc., 9: 14. UHLENHUTH, E., 1919. Relation between Thyroid Gland, Metamorphosis, and Growth. Jour. Gen. Physiol.. 1: 473. WILDER, I. W., 1924. The Relation of Growth to Metamorphosis in Eurycea bislineata (Green). Jour. E.rp. Zoo/., 40: 1. THE EFFECTS OF TEMPERATURE CHANGES UPON THE CHROMATOPHORES OF CRUSTACEANS DIETRICH C. SMITH i (From the Harvard Biological Station, Soledad, Cienfucgos, Cuba and the Zoological Laboratory, Harvard University.} Temperature changes as they affected the chromatophores of crus- taceans were not neglected in the researches of early investigators; those of Jourclain ( 1878 ) being the first recorded in the literature to consider the possible influence of this factor. At 5°-6° C., according to his observations, the rapidity at which color changes occurred in Nica cdnlis was appreciably reduced, ceasing entirely as the temperature ap- proached nearer to zero. At this point the animals were almost trans- parent, except for areas partly covered with matted white spots. Jour- dain removed the eyes of those crustaceans and noted that the reddish color assumed at room temperatures, under such circumstances, dis- appeared entirely when the temperature of the water was lowered only to reappear again on the restoration of the temperature to its former level. Matzdorff (1883) observed no effect whatever of either high or low temperatures upon the chromatophores of Idotea tricuspidata. Somewhat later however, Gamble and Keeble (1900), after a few ob- servations upon Hippolyte varians, reported observable color response following exposure to heat and cold. Their specimens in common with most other crustaceans possessed several differently colored pigments, reds and yellows predominating, located with one exception in discrete bodies or chromatophores. During the day the reds and yellows were usually expanded, but at night these pigments were retracted into their chromatophore centers and if it were not for a blue pigment, diffused at this time throughout the tissues and free from any chromatophore. the animals would be colorless. Gamble and Keeble selected three of these transparent blue prawns, which they called " nocturnes," and placed one in water at 15.5° C. (60° F.), one in water at 8° C. and the last in water at 32° C. (93° F.). The first animal, in reality the control as it was kept under normal temperature conditions, turned greenish-brown as was to be expected. The second one at 8° C. main- tained the nocturnal blue color, showing after thirty-five minutes some traces of recovery, though one hour later this was still incomplete. 1 National Research Fellow in the Biological Sciences. 193 194 DIETRICH C SMITH The prawn placed in 32° C. was almost immediately killed by the heat, but remained nevertheless a brilliant nocturne for several hours, even though during the first five minutes of this experiment the temperature descended to 28° C. (83° F.). Menke (1911) experimenting with Idotca, produced a contraction of the chromatophores in about 15 minutes by raising the temperature of the water from 11.5° C. to 20.5° C. This contraction was sustained for about one hour when the pigment partially re-expanded. Five and one-half hours later on, lowering the temperature to 12° C., the chro- matophores again became completely expanded. But if at this time, instead of lowering the temperature of the water, it was raised to 30° C., the chromatophores also re-expanded completely. Complete ex- pansion was also produced by lowering the temperature from 14° C. to 4° C. Doflein (1910), working with Lcander xiphias placed several specimens in complete darkness at 5°-8° C. for two to three days. At the lapse of this time the chromatophores and the tissues of the animal were completely impregnated with blue pigment, all other pigments being completely retracted into their respective centers. But as Fuchs (1914) points out, these results might follow either from continued exposure to cold or to darkness. Megusar (1911) working with Gelasimus, Potamobius, Palceinonetes, and Palcstnon, observed an ex- pansion of the chromatophores on the sudden transfer of any of these animals from water at 16°-18° C. to water at 25°-30° C. Similarly a contraction of the chromatophores followed a sudden transfer from water at 25°-30° C. to cooler water at 16°-18° C. The results of these experiments are admittedly confusing, though as Fuchs (1914) observed, no reasonable doubt can be entertained as to the ability of temperature changes to produce an effect of some sort upon the pigmentary responses of the crustaceans. Further in- vestigation of the subject was thought desirable in the hope of ascer- taining, if possible, just how important a factor the action of heat and cold is in determining the distribution of the chromatophore pig- ment of this group. For this purpose a fresh water shrimp, kindly identified for me as Macrobrachiuni acanthurus Wiegmann by Dr. W. L. Schmitt of the United States National Museum, was selected as the subject of the experiments. These shrimps were obtainable in large numbers from the Arimao river and its immediate tributaries in the vicinity of the Harvard Biological Station, Cienfuegos. Cuba. I am happy to acknowledge my thanks and appreciation to Dr. Thomas Bar- bour for his assistance in putting the facilities of the Harvard Cuban Station at my disposal. When caught, the chromatophore pigments of these shrimps were EFFECTS OF TEMPERATURE ON CHROMATOPHORES 195 more or less extended, giving the animal a reddish-brown color. This color varied somewhat with the size of the animal, the smallest being the lightest. As collected, the shrimps ranged from 2 cm. to 10 cm. in length, measured from rostrum to telson. Males varying from 2 cm. to 3 cm. in length were selected for the experiments. Females were rejected, as at this time their abdomens were practically opaque owing to the fact that they were carrying their eggs. A word or two regarding the color changes of Macrobrachiwn will be an aid to the understanding of what is to follow. Taken to the laboratory and placed in white glazed porcelain bowls, the shrimps in daylight soon became transparent and colorless ; microscopical examina- tion of the abdomen and telson showing the chromatophores to be completely contracted. If such animals were placed upon a black back- ground, they assumed a dark reddish color with the chromatophores well expanded. A somewhat superficial examination disclosed the presence of two types of chromatophore pigments, both apparently located in the same chromatophore, one being reddish-brown in color and the other yellow. These facts were derived from a microscopical examination of the living pigmentary units. Detailed histological study of the chromatophores was not attempted. Occasionally under somewhat varying conditions, animals were seen with an unmistakable bluish color observed both in the light and in the dark. The blue pigment producing this color when examined under the microscope was clearly not confined to the chromatophores, but was free in the tissues, though its concentration did appear greater about the processes of the pigmentary centers. Gamble and Keeble (1900) reported that a blue color was the regular accompaniment of the noc- turnal phase of Hippolyte varians, a phase characterized by the retrac- tion of all other pigments into their respective centers. According to their statement, the blue pigment in Hippolyte responsible for the noc- turnal coloration arises as a discharge product of the chromatophores, leaving these organs on the contraction of the yellow and red pigments, and apparently being derived from them. Left free in the tissues, the blue pigment is permanently divorced from its point of origin and persists in coloring the body of the prawns until it eventually disappears. In these experiments upon Macrobrachium acanthurus determina- tions were first made of the action of heat and cold upon the color changes in normal shrimps. The method used was as follows : Two or three animals were placed in white porcelain bowls and covered with water at room temperatures. To this was added either warm or cold water, as desired, until the particular temperature demanded by the experiment was reached. Here it was either kept constant or 196 DIETRICH C SMITH altered as necessary. The responses of the shrimps to temperature changes when kept upon a black background were tested in the same manner. Numerous experiments with normal shrimps adapted to white back- grounds demonstrated conclusively that such animals darkened when exposed to temperatures either high enough or low enough to stimulate the chromatophores. Surprising as it may seem, once the response was complete, no criteria of any sort could be established separating the darkening produced by heat from that produced by cold. The color assumed in either circumstance was a deep red-brown, while micro- scopical examination showed the pigments of the chromatophores to be equally well extended at high and low temperatures. The protocols of the two following experiments may be taken as typical of many others : 2:13 — 28° C. Two colorless shrimps previously kept on a white background for a day were placed in a white porcelain bowl and small pieces of ice added to the water. 2 : 16—10° C. No change in color. 2:18 — 10° C. Shrimps appear slightly reddish. 2:22 — 15° C. Shrimps somewhat darker. 3:00 — 15° C. Shrimps a pronounced brown. 3:25—28° C. Shrimps still brown. 9 : 30—28° C. Shrimps colorless. 2 : 13 — 28° C. Two colorless shrimps previously kept on a white background for a day were placed in a white porcelain bowl and warm water gradually-added. 2 : 16—36° C. No change in color. 2:18—36° C. Shrimps faintly reddish. 2 : 21 — 36° C. Shrimps pronouncedly brown. 3:00—28° C. Shrimps colorless. In all of the experiments the appearance of the red-brown color was more rapid at high temperatures than at low. With heat only 10-15 minutes were necessary to make the animal completely dark, while with cold 30-45 minutes were required. But regardless of whether the shrimps were exposed to heat or to cold, once the point of maximal darkening was reached, the intensity of the color was equal in both cases. When the animals were subjected to warmth the lowest temperature capable of expanding the chromatophores was found to be 35° C., while temperatures as high as 40° C. could be withstood without subsequent death, though at this temperature and slightly below it, the shrimps re- mained motionless, and with the exception of gill movements showed no signs of life. Therefore, within the range of 35° C. to 40° C. the color of the shrimp is determined by the temperature of its en- EFFECTS OF TEMPERATURE ON CHROMATOPHORES 197 vironment rather than the type of background on which it happens to be. Similarly shrimps placed in water colder than 6° C. died immedi- ately, while any temperature above 15° C. and, of course, below 35° C., failed to produce an expansion of the chromatophore pigment. Therefore, between 6° C. and 15° C. the color of the shrimp is also determined by temperature rather than background. It might be well to mention here that the temperature of the water in which the shrimps normally lived ranged from 25° C. to 30° C. As a check upon these results and to determine whether there was any possibility of temperature changes producing a contraction of the chromatophore pigment, experiments similar to those just described were performed upon dark shrimps while they were upon a black back- ground. But such animals when exposed to various temperatures rang- ing between 6° C. and 40° C. showed no alteration whatever in the expanded condition of their chromatophores. Recovery of normal color and activity was the rule when shrimps subjected to effective temperatures were returned to water at about 28° C. But this recovery was more rapid in shrimps treated with warmth than those treated with cold. The former required but 30 to 40 minutes, and the latter 6 to 7 hours before normal temperatures and a white background again brought their chromatophore pigment to complete contraction. Tests were also made of the responses of blinded shrimps to tem- perature changes, blinding being accomplished by cutting off the eyes at the base of the eye stalk. Shortly after this operation the pigment of the chromatophores began to expand and within an hour or so, re- gardless of background, this expansion was complete and the animals were red-brown in color. Shrimps in this condition placed in warm and cold water and left so for an appreciable length of time — two to three hours — showed no color change of any sort. Similarly shrimps anaesthetized with 0.05 per cent chloretone. failed to show color re- sponses to either heat or cold. Neither high nor low temperatures are then capable of exerting any contracting effect upon the pigment of the chromatophores, even when these organs are removed from the in- fluence of any stimuli directly or indirectly produced by the retina. As Perkins (1928) has shown in Pahcnwnetcs, the withdrawal of pigment into the centers of the crustacean chromatophore is con- trolled by a hormone elaborated in the eye stalks, a fact which was later substantiated by Roller (1928) on Crangon and Lcandcr. Pos- sibly then, as temperature changes acted to expand the chromatophore pigment, there was an inhibition by heat or cold of the mechanism controlling the production of this contracting secretion. Before such 198 DIETRICH C. SMITH an hypothesis could be tested, it was necessary to ascertain definitely whether or not such a secretion played a part in governing the chro- matic responses of Macrobrachinin. Consequently Perkins' experi- ments were repeated upon this animal. Five or six shrimps were paled by placing them upon a white background for a day or more, after which their eyes were removed and thoroughly macerated in 2 cc. of 0.7 per cent NaCl. One tenth cc. of the resulting solution was then injected into the abdomens of several shrimps in the dark condition and with well expanded chromatophores. In all cases the following reactions were noted : Shortly after injection, 5-10 minutes, the shrimps began to assume a bluish color which gradually increased in intensity until within 30 minutes it had reached its maximum ; this was followed by a gradual retraction of the pigment into the chromatophore centers, a retraction which persisted until the shrimps had assumed a transparent blue color. These results closely parallel the effects reported by Perkins in Paltzmonetes, even to the formation of the blue color, and offer com- plete substantiation of his findings. As control experiments 0.1 cc. of the extract was injected into the abdomens of several shrimps in the light condition with no observable effect. Similarly injection of 0.1 cc. of 0.7 per cent NaCl into blinded shrimps produced no pig- mentary response. The existence of a hormone produced by the action of light upon the retina and released into the circulation to affect a contraction of the chromatophore pigment is then demonstrated in the shrimps used in these experiments. Is the formation of this hormone in any way in- hibited by either high or low temperatures? Apparently not, as the following experiments show. Two sets of extracts were prepared, one from the eyes of shrimps darkened on a white background by warm water (37° C.) and the other from the eyes of shrimps darkened on a white background by cold water (15° C.), both groups being subjected to their respective temperatures for the same length of time, namely 45 minutes. Two sets of blinded shrimps were then selected, one set being injected with 0.1 cc. of one extract and the other set with the same amount of the other extract. These animals were then replaced in water at room temperature and the results noted. In both cases these darkened shrimps paled within the specified length of time, but with this difference,- — the blue color previously described appeared in only one out of three of the shrimps injected with the extract prepared from the eyes of animals kept at low temperatures, while it appeared in all of the shrimps injected with the extract prepared from the eyes of animals kept at high temperatures. Neglecting for the present the significance of this variation, it is obvious that extreme temperatures EFFECTS OF TEMPERATURE ON CHROMATOPHORES 199 in no way inhibit the manufacture or even the potency of the chro- matophore-contracting hormone elaborated by the eye stalks. This gives us a clue as to the manner in which heat and cold affect the chromatophores of crustaceans. Unfortunately, these experiments cannot give us a conclusive solution to this problem, though the data at hand strongly indicate a direct effect. Positive information is not to be derived from experiments on limbs or bits of integument isolated from the bodies of these shrimps, as the chromatophores of such ex- cised pieces expand at once. Consequently subjecting such preparations to temperature variations accomplished no change in the distribution of their expanded chromatophore pigment. But since experiments on blinded and chloretonized shrimps give no evidence of any other type of response to temperature changes than those seen in normal light shrimps, and since neither heat nor cold affect the secretion of the chromatophore-contracting substances elaborated in the eye stalks, it seems reasonable to assume that the responses of the chromatophore pigment of crustaceans to high and low temperatures are direct. A word or two in regard to the blue color and its relation to tem- perature changes. Keeble and Gamble (1903) state that the blue color observed in nocturnal Hippolyte disappears completely at 60° C., while, as shown in an earlier paper (Gamble and Keeble, 1900), this color is maintained at 8° C. under conditions that in other prawns kept at a somewhat higher temperature (15° C.) produce its loss. This latter observation is in accord with the experiments of Doflein (1910) on the occurrence of a blue color in Lcandcr when the animals were kept for an extended period in darkness and cold. But aside from this, it is perhaps worthy of note that, as already mentioned, blinded animals injected with the extracts prepared from the eyes of shrimps subjected to cold showed only in one third of the cases a visible blue color, whereas blinded shrimps injected with an extract from the eyes of animals kept at high temperatures never failed to become pro- nouncedly blue. Furthermore, throughout the course of these experi- ments the blue color was repeatedly observed in connection with shrimps subjected to high temperatures, while the records disclose only one instance where it was seen in connection with shrimps exposed to low temperatures ; a case where an animal kept at 6° C. for about 30 minutes turned blue when returned to 28° C. Perhaps this indicates a relationship between changes in temperature and the appearance of the blue color worthy of further investigation. A survey of the work of previous investigators dealing with the action of heat and cold upon the crustacean chromatophore reveals a wide divergence of opinion. As we have already seen, Gamble and 200 DIETRICH C. SMITH Keeble (1900) claimed that both high and low temperatures produce or at least maintain a retraction of the pigment, a statement with which Jourdain (1878) and Doflein (1910) are in agreement as far as low temperatures are concerned. Menke (1911), on the other hand, reports that in Idotca extreme high and low temperatures tend to produce an expansion of the chromatophore pigment, though moderately high temperatures (20°-25° C.) lead to a contraction. Megusar (1911), however, observed an expansion of the chromatophore pigment with heat and a contraction with cold, though this author apparently did not subject his animal to temperatures lower than 15° C. In the most re- cent communication Roller (1927) maintains that temperature changes have no effect whatever upon the distribution of the chromatophore pigment in Cranyon. The results of the present writer's investigations are more in accord with those of Menke than with those of other workers, since Menke also observed an expansion of the pigment at both ends of the effective temperature scale. Macrobrachiitui acanthtints is a semi-tropical form, habituated to water normally remaining at 25°— 30° C. the year around. Therefore, the response to temperature changes of such forms might reasonably be expected to vary somewhat from those seen by Menke in Idotca, a form adapted to life in cooler waters. Consequently we need not be greatly concerned when Idotea responds to temperatures of 20°-25° C. and Cuban shrimps do not. For the latter such tem- peratures are obviously not warm. The important feature is that for both types an expansion of the chromatophores is produced on exposure to temperatures either extremely high or extremely low. Among the lizards and amphibians high temperatures as a rule produce a contraction of the chromatophores and low temperatures an expansion. Among the vertebrates in general, variations from this scheme are found in certain amphibians whose chromatophores are ap- parently insensitive to heat and among the fishes, where innervated melanophores react to warmth by expansion and to cold by contraction. The denervated melanophores of fishes respond, however, to tempera- ture changes as do the chromatophores of lizards and amphibians. Since in these last two groups such reactions are presumably direct, and since they are certainly direct in denervated fish melanophores, it is permissible to say that among the vertebrates the independent re- sponse of the chromatophore to heat is a contraction and to cold an expansion. In the crustacean chromatophore where there is a high probability that reactions to temperature variations are direct, though this is admittedly not certain, an expansion of the chromatophore pig- ment is produced both by heat and cold. On the basis of our present EFFECTS OF TEMPERATURE ON CHROMATOPHORES 201 knowledge then there seems to be little resemblance between the pig- mentary reactions to heat and cold in the vertebrates and the crustaceans. Among the vertebrates, especially in the lacertilians, the ability of the pigment cells to respond to temperature changes is sometimes given a thermo-regulatory significance. But the crustacean chromatophorc can certainly serve no such purpose, especially as the chromatic re- sponses of this group are controlled by factors other than heat and cold. It is inconceivable, for instance, that the form used in these experiments would ever encounter in its usual environment temperatures high enough or low enough to bring about changes in the distribution of its chromatophore pigment other than the distribution determined by background or light intensity. SUMMARY 1. Expansion of the chromatophores of Macrobrachium acaiit hunts. a Cuban shrimp, follows immersion of these animals in fresh water at any temperature between 6° and 15° C. or between 35° and 40° C. This reaction occurs regardless of the background upon which the shrimp is placed. Between 15° C. and 35° C. the chromatophores of this shrimp expand when the animal is placed upon a black background and contract when the animal is placed upon a white background. 2. In blinded and chloretonized shrimps, the chromatophores are expanded and this expansion is in no way altered by changes in back- ground or temperature. 3. Neither high nor low temperatures have any effect upon the potency or manufacture of the chromatophore-contracting substance elaborated by the eye stalks. BIBLIOGRAPHY DOFLEIN. F., 1910. Lebensgewohnheiten und Anpassungen bei dekapoden Krebsen. Festschrift. Hertwig, 3: 215. FUCHS, R. F., 1914. Der Farbenwechsel und die chromatische Hautfunktion der Tiere. Wintcrstein's Handbuch der vcryl. Physiol., Jena. 3: Halfte 1, Teil 2, 1344. GAMBLE, F. W., ANL KEEBLE, F. W., 1900. Hippolyte varians : a Study in Colour- change. Quart. Jour. Micr. Sci.. 43: 589. JOURDAIN, M. S., 1878. Sur le changements du couleur de Nika edulis. Cornet. rend. Acad. Sci.. Paris, 87: 302. KEEBLE, F. W., AND GAMBLE, F. W., 1903. The Color Physiology of Higher Crustacea. Phil. Trans. Roy. Soc.. London, 196B: 295. ROLLER, G., 1927. Uber Chromatophorensystem, Farhensinn und Farbwechsel bei Crangon vulgaris. Ztschr. vcrgl. Physiol.. 5: 191. ROLLER, G., 1928. Versuche uber die inkretorischen Vorgange beim Garneelenfarb- wechsel. Ztschr. rergl. Physiol., 8: 601. 202 DIETRICH C. SMITH MATZDORFF, C., 1883. Ueber die Farbung von Idotea tricuspidata Desm. Jena. Ztsclu: f. Mcdizin u, Naturw., 16: 1. MEGUSAR, F., 1911. Experimente iiber den Farbwechsel der Crustaceen. (I. Gelasimus. II. Potamobius. III. Paljemonetes. IV. Palsemon.) Arch. f. Entiv.-Mcch. d. Org., 33: 462. MENKE, H., 1911. Periodische Bewegungen und ihr Zusammenhang mit Licht und Stoffwechsel. Arch. gcs. Physio!., 140: 37. PERKINS, E. B., 1928. Color changes in Crustaceans, especially in Palaemonetes. Jour. Exp. Zool, 50: 71. Vol. LVIII, No. 3 JUNE, 1930 THE BIOLOGICAL BULLETIN PUBLISHED BY THE MARINE BIOLOGICAL LABORATORY UNUSUAL TYPES OF NEPHRIDIA IN NEMERTEANS WESLEY R. COE OSBORN ZOOLOGICAL LABORATORY, YALE UNIVERSITY The earlier investigators on the morphology of the nemerteans failed to find in any of the species of the family CephalotrichicUe the pair of longitudinal nephridial ducts which are so conspicuous in most nemerteans, and some of them came to the erroneous conclusion that in this family the nephridia are absent. Wijnhoff (1910) corrected this error, proving that a well-developed excretory system is actually- present in the females of several species, but of a different nature than had been found up to that time in any nemertean. Instead of having all the terminal organs connected with a single longitudinal canal of comparatively large size, each end bulb has its own efferent duct leading to the exterior of the body. Wijnhoff was unable to determine the exact configuration of the organs or the details of their histological structure, although she de- scribes and figures the terminal organs in their relation with the lateral blood vessels and shows the groups of granular cells adjacent to the end bulbs. METANEPHRIDIA IN CEPHALOTHRIX MAJOR On the coast of California occurs a species of Cephalothrix (C. major Coe), in which the worms reach a size many times larger than those of other known species of the genus and in which the histological structure of the extremely complex excretory organs is clearly shown. In this species, as in those studied by Wijnhoff, there is a series of isolated nephridia in close contact with the lateral blood lacuna- on each side of the body. Each nephridium consists of a multinucleate terminal organ, or end bulb, with slender flagella on its free border, and with a narrow canal leading to an enlarged glandular and con- voluted tubule and thence by an efferent duct to a minute pore on the dorsolateral aspect of the body (Figs. 1, 4, 5, 8, 9). 203 14 ebL tern VV f **/••, O t, '« *• ,. ri FIGS. 1-4. Metanephridia of C. major. FIG. 1. Entire nephridium with widely opened efferent duct. FIG. 2. Terminal organ (nephrostome) associated with a large area of gelatinous parenchyma. FIG. 3. Terminal organ close be- neath epithelial lining of blood lacuna. FIG. 4. Diagram of portion of body near anterior end of intestinal region, showing position of nephridium (nc) and efferent duct; bl, blood lacuna; bm, basement membrane of body wall; con, convoluted tubule; ebl, epithelium of blood lacuna; ictn, hn, ocm, inner circular, longitudinal and outer circular musculatures ; iep, intestinal epithelium ; in, integument ; In, lateral nerve; ncd, efferent duct; nc[>, nephridiopore ; par, parenchyma; ps, probos- cis sheath; re, rhynchocoel ; tc, terminal chamber; to, terminal organ. TYPES OF NEPHRIDIA IN NEMERTEANS 205 The number of such independent nephridia is very large, more than 300 being found on each side of the body in an adult worm measuring a meter or more in length. All of them are found in the anterior half of the body. The most anterior ones border the blood lacunae anterior to the mouth, the others being situated beside the lateral lacuna? in the region of the foregut and extending posteriorly beyond the anterior limits of the gonads. Although the nephridia are not paired on the two sides of the body, there is more or less regularity in their arrangement. Anteriorly they are more widely spaced than somewhat farther back, and they are most closely placed and of max- imum size in the region where the foregut opens into the intestine, that is, in the region somewhat posterior to the most anterior gonads. More posteriorly they are not only farther apart, but are appreciably smaller and with fewer nuclei. \ ebi • . -.. 7| J FIGS. 5-8. C. major. FIG. 5. Nephrostome (ne) imbedded in bulbous mass of gelatinous parenchyma (par). FIG. 6. Nephrostome close beneath epithelium of blood lacuna (cbl). FIG. 7. Transverse section of nephrostome, showing outer circle of nuclei (n) belonging to the flagella-bearing cells and the inner circle of smaller nuclei (n") lining the end canal. FIG. 8. Diagram of nephridium in longitudinal section. The actual distance between adjacent nephridia is commonly from 0.1 mm. to 0.2 mm. in the mounted sections, although some are sep- arated by only 0.05 mm. or twice the diameter of the terminal bulb. All are placed in a very similar situation with regard to the blood lacunae and the nerve cords, always lying near the lumen of the blood space in the angle adjacent and somewhat dorsal to the nerve cord (Fig. 4). In many cases the terminal organ is situated on a low pa- pilla, formed of the endothelium of the blood lacuna and its underlying basement membrane. This papilla projects somewhat into the lumen of the blood space, so that the greater part of the surface of the ne- phridium comes in close proximity to the blood (Figs. 5, 6). 206 WESLEY R. COE Each nephridium consists of three principal parts, (o) the terminal bulb, (b*) the convoluted tubule, and (c) the efferent duct (Figs. 1, 8, 17). -bl tc- bl SLcon con -tc ,-nep FIG. 9. Diagram of nephridium of C. major, showing slender flagella in lumen of convoluted tubule (con) ; 9«, small portion of convoluted tubule with flagella. FIG. 10. Diagram of nephridium of C. spirali-s, showing both nephrostome and convoluted tubule in bulbous projections on wall of blood lacuna (bl). FIG. 11. C. spiralis. Section of nephrostome (tc) and loops of convoluted tubule (con) in single bulbous projection of wall of blood lacuna (bl). (a) The Terminal Bulb {Nephrostome). — This lies in all cases in close proximity to one of the lateral blood lacunae, which are usually much distended throughout the nephridial region (Fig. 4). Sometimes the bulb occupies a small papilla projecting somewhat into the lumen of the blood space and separated from the latter only by a thin covering of parenchyma and the endothelial lining of the blood vessel (Fig. 3). More often it lies deeper in the tissues and separated from the blood vessel and from the surrounding tissues by the mass of gelatinous parenchyma in which it is always imbedded (Fig. 2). TYPES OF NEPHRIDIA IN NEMERTEANS 207 The terminal bulb is a mushroom-shaped structure with an ex- tremely fine tubular stalk and with the free convex surface projecting into a hemispherical chamber (Fig. 3). Occasionally it can be demon- strated beyond question that a number of delicate flagella project freely into the terminal chamber and that the lumen of the latter is continuous with the slender tubule which pierces the stalk of the mushroom-shaped end bulb (Figs. 5, 8). The flagella are obviously projections from the free surface of the end bulb, and their vibration in life doubtless serves to draw into the tubule the fluid which collects in the vacuole. The wall of the chamber consists of a very delicate membrane with one or two oval nuclei on its inner surface (Fig. 5). Two types of cells are found in the terminal organ, (a) those which compose the mushroom-shaped body and bear the flagella and (b) those belonging to the tubule of the stalk. In neither part are there distinct cell boundaries, but in the former the nuclei are much larger than in the latter (Figs. 5, 6). The cytoplasm on the hemispherical free surface of the terminal organ is dense and firm, forming a suitable support for the flagella. In the deeper part of this cytoplasm upwards of 20 oval nuclei are imbedded, six to eight of these being seen in a single longitudinal section (Figs. 5, 6, 8). In a cross section, however, the entire number may be shown (Fig. 7). The canal in the tubular stalk is very slender and only in exceptional cases is the lumen demonstrable, due to the state of contraction at the moment of preservation. The nuclei of this canal are often only half the diameter of those of the terminal organ, or even less (Figs. 7, 8). Their number seldom exceeds a dozen. The size of the terminal bulb varies considerably as may be seen from figures 5-8, which are all drawn to the same scale, the transverse diameter being usually from 0.024 to 0.027 mm., although the smallest are only 0.018 mm. across and the largest as much as 0.03 mm. (b) The Convoluted Tubule. — The slender canal in the stalk passes through the parenchyma surrounding the terminal bulb and then en- larges suddenly into a coiled tubule of much greater diameter and often with a conspicuous lumen (Figs. 1, 3, 9). This part of the nephridium is imbedded in a restricted mass of parenchyma more or less continuous with that surrounding the terminal bulb and extending into the inner portion of the longitudinal muscular layer. The con- figuration of the convolution is quite variable, as a comparison of the various figures will show. Sometimes there is but a single loop, but usually the tubule twists spirally in an irregular manner, parts of it appearing in four or more serial sections. The cytoplasm is coarsely granular, with numerous inclusions, but the nuclei are not separated 208 WESLEY R. COE by distinct cell boundaries (Fig. 9). Tbis part of tbe nephridium closely resembles in structure the main longitudinal canal in those forms having compound nephridia (protonephridia), and its coarsely granular and vacuolated cytoplasm indicates that it has an important excretory function. Long slender cilia project from the inner walls of the convoluted tubule ; giving the appearance of fine threads lying lengthwise in the lumen and extending in the direction of the efferent duct (Figs. 9, 9a). (c) The Efferent Duct. — The convoluted tubule leads directly into an extremely slender efferent duct which passes radially, that is, dorsally and laterally, in one of the connective tissue dissepiments separating the bundles of longitudinal muscles. It then pierces the outer circular musculature, the basement membrane and the integument, to open by a minute pore on the dorsolateral surface of the body (Figs. 1, 4, 9). The course of the duct may be so perfectly straight that nearly the entire length may be contained in one or two of the serial sections, but it is naturally seldom that the plane of the section coincides exactly with that of the duct. The wall of the duct is extremely thin, but the cytoplasm bears numerous oval nuclei throughout its entire length. Even where the duct pierces the integument it has its independent nucleated lining (Fig. 1), as Wijnhoff (1910) has already demonstrated for other species. DISCUSSION Excretory organs of this type have not been described for any of the other groups of Plathelminthes. In some of the Annelids, however, organs of somewhat similar structure are found, each with a ciliated funnel (nephrostome) opening into the body cavity and with a convoluted tubule, often of great complexity. In the nephridium of Cephalothri.v the mushroom-shaped end bulb is apparently homologous with the nephrostome of the annelid and may be so designated. The terminal chambers in Ccplialothri.\- then repre- sent minute ccelomic cavities, the fluid contents of which are in com- munication with the outside world through the nephridia, exactly as in annelids. This type of excretory organ may be designated a metanephridium in order to distinguish it from the more usual type, protonephridium, found in nemerteans (Fig. 17, 5), where each of the numerous end bulbs consists of a single flagellated cell imbedded in the body paren- chyma and with its free border directed toward the efferent duct. TYPES OF NEPHRIDIA IN NEMERTEANS 209 PHYSIOLOGY OF THE METANEPHRIDIUM The process of excretion by this type of nephridium is presumably accomplished by the withdrawal of waste-containing fluids from the surrounding gelatinous parenchyma, and thus indirectly from the nearby blood, by means of the ciliary action of the nephrostome. These fluids then pass to the convoluted tubule, the cells of which are specialized for the excretion of additional waste materials or for the absorption of any contained nutrients, or both ; after which the remaining fluid is forced through the efferent duct to be discharged through the ne- phridiopore. The movement of the fluid in the convoluted tubule is doubtless facilitated by the slender flagella with which it is provided. The numerous granules and minute vacuoles in the cytoplasm of this part of the nephridium are indicative of its excretory function, as Strunk (1930) has recently demonstrated experimentally for Annelids. EXCRETORY SYSTEMS IN CEPHALOTHRIX SPIRALIS In another species of the genus, C. spiralls Coe (formerly considered specifically identical with C. lincaris Oersted of Europe) of the New England coast, the excretory system of the female is likewise of the metanephridial type. In the two sexually mature males of this species which were available for study, however, no metanephridia were found, the excretory system consisting of a pair of clusters of protonephridia situated on the median walls of the cephalic blood lacunae (Figs. 13, 15). The meaning of this apparent sexual dimorphism is by no means clear and will require further investigation on immature forms of both sexes. It may be remembered in this connection, however, that a somewhat similar condition prevails for the reproductive organs of some of the bathypelagic nemerteans, the males of which have only a few pairs of spermaries (and these are situated in the head), while the females are provided with numerous ovaries on each side of the body in the intestinal region (Coe, 1920). It will be recalled also that in the Annelids and other groups of invertebrates the larval excretory system is frequently of the protonephridial type, and is later replaced by the metanephridia. It seems possible that the sexual dimorphism in Ccph- alothrix may be similarly accounted for, assuming that the males have retained the primitive protonephridia, and that these are replaced in the females by the more complicated, and presumably more efficient metanephridia. Studies on immature individuals of both sexes will be made in the near future. 210 WESLEY R. COE METANEPHRIDIUM The metanephridium of the female C. spiralis is similar to that of C. major, but is considerably larger in proportion to the size of the body and is more intimately associated with the lumen of the blood lacuna (Figs. 10, 11, 12, D, E, F). The average diameter of the nephrostome in this smaller species is about 0.023 mm., with some as small as 0.012 mm., as compared with 0.018 to 0.03 mm. in C. major. bl bl '•;"••'••: tei B FIG. 12. Diagrams of the various types of nephridia found in nemerteans, showing the relation of each to the blood lacuna; A, protonephridium, characteristic of most nemerteans, imbedded in parenchyma close beneath blood lacuna ; B, protonephridium of C. spiralis, male, hanging free in blood lacuna ; C, protone- phridium of Geonemertes, imbedded in parenchyma; D, E, F, metanephridia of C. spiral is, female, in successive stages of differentiation. The shape as well as the position of the nephrostome varies consid- erably in the same individual. Only occasionally is the organ circular in surface view, with the opening of the end canal in the center. More often the opening is considerably eccentric, showing more nuclei on one side than on the other in vertical section (Figs. 10, 11, 12, F). In some cases the organ is heart-shaped or distinctly bilobed, with the opening in the indentation (Fig. 12, D, E). In regard to their position relative to the blood lacunae, both the nephrostome and the entire convoluted tubule may lie in the paren- chyma beneath the epithelium and make no encroachment whatever on the lumen of the blood space or both may form bulbous projections into the lumen (Figs. 10, 11, 12, D, E, F). As a general rule, how- ever, the terminal chamber projects freely into the blood space, while considerable gelatinous parenchyma lies between the convoluted tubule and the epithelial lining of the lacuna. In the females a single pair of metanephridia is situated on the TYPES OF NEPHRIDIA IN NEMERTEANS 211 dorsolateral borders of the cephalic lacunae not far anterior to the mouth. The convoluted tubule of this nephriclium is greatly elongated anteroposteriorly, with the slender efferent duct at its posterior end. Anterior to the midgut the nephritlia are widely scattered, increasing in abundance in the anterior portion of the gonad region and becoming less numerous beyond the end of the proboscis sheath. At least a hundred pairs are found in an individual of moderate size. In the mounted sections the distance between adjacent nephriclia is usually 0.1 to 0.2 mm. in the anterior midgut region. The nephrostome is frequently situated on a horizontal level with the lateral nerve cord, with the convoluted tubule either anterior or posterior and slightly dorsal thereto, but sometimes the nephrostome is found much nearer the ventral side of the body. In the latter case the efferent duct passes dorsally above the level of the nerve cord before leading radially to the nephridiopore on the dorsolateral surface of the body (Fig. 14). 13 FIG. 13. C. spiralis. Portion of transverse section through head of male, showing terminal chambers (tc) of nephridium on median wall of blood lacuna (/>/); con, convoluted tubule opening to surface through efferent duct (ncd) ; hn, buccal nerve; cm and hn, circular and longitudinal musculatures; In, lateral nerve cord ; p, proboscis. FIG. 14. C. spiral is. An unusually large nephridium from the intestinal region posterior to the end of the proboscis sheath, showing the terminal chamber (tc) adjacent to the blood lacuna (cbl) and the voluminous convoluted tubule (con) leading dorsally to join the slender efferent duct (ncd) ; in, integument; lu. lateral nerve; hn, ocm, longitudinal and outer circular musculatures. 212 WESLEY R. COE The nephrostome is evidently capable of considerable change of shape by contraction and extension, for the mouth of the end canal joining the terminal chamber may be widely opened (Figs. 10. 11, 14) or it may be almost completely closed. The terminal chamber also may be distended with fluid and thus widely separated from the ciliated surface of the nephrostome (Fig. 10) or the fluid may be withdrawn, allowing the thin wall of the chamber to lie close upon the nephrostome. A few double nephridia were found, and Wijnhoff (1910) observed the same condition in one of the species which she studied. The twinning may involve only the terminal organ and its accompanying end canal or it may include also the entire convoluted tubule. In the latter case two complete nephridia join a single efferent duct. PROTONEPHRIDIUM Mention has been made of the fact that metanephridia have been found thus far only in the females of the several species studied. Only two sexually mature males of C. spiralis with suitable fixation have been available for study and both of these were provided with ex- cretory organs of the protonephridial type (Figs. 13, 15). Each of the two individuals had a single pair of these organs situ- ated on the median borders of the cephalic blood lacunse between the brain and the mouth. Each nephridium consists of a cluster of fifty or more end organs connected with a branched collecting tubule which leads dorsally along the median face of the lacuna (Figs. 13, 15). On the dorsomedian angle of the lacuna the collecting tubule opens into the convoluted tubule, from which the efferent duct leads to the ne- phridiopore on the dorsolateral border of the head (Fig. 13). Each of the end organs consists of a single cylindrical or goblet- shaped cell (flame cell) attached to the wall of the lacuna and more or less completely surrounded by the blood. The cytoplasm of the cell and the cell membrane are extended to form a central oval cavity in which the slender flagella may swing freely (Figs 12, B, 15). The proximal end of the flame cell is narrowed to a slender canal (end canal) which joins with others to form the collecting tubule. Such an intimate association of flame cells with the blood is known for other species of nemerteans, but in no case is there any direct communication between the blood and nephridial systems. In order for fluid to pass from the blood to the excretory canal it must be filtered through the osmotic membranes and cytoplasmic extensions of the excretory cells. TYPES OF NEPHRIDIA IN NEMERTEANS 213 Uebl -Pc cLj Lee -ned FIG. 15. C. spiral is. Diagram of cephalic protonephridium of male, showing the isolated flame cells (fc), with the terminal chambers (tc) leading to the slender end canals (cc) and thence to the collecting tubule (ct), convoluted tubule and efferent duct ; ebl, epithelial lining of cephalic blood lacuna. FIG. 16. Gconcmcrtcs agricola. Diagram of single nephridium with a cluster of slender terminal chambers (tc) and binucleate flame cells (fc) leading by the narrow end canals (cc) to a thick-walled convoluted tubule (con) and thence to the efferent duct (ncd) ; /, tuft of long cilia; ntc, nucleus of terminal chamber; circular bars on wall of terminal chamber. COMPARISON WITH OTHER FORMS With the exception of the metanephridia of the females of species belonging to the family Cephalotrichidae, the excretory systems of all nemerteans in which such organs have been discovered are of the protonephridial type. In the numerous species of hathypelagic nemer- teans, as well as in the littoral Prosadcnoponis, no trace of an excretory system has yet been found. Characteristic of the vast majority of species is a system of simple 214 WESLEY R. COE flame cells (Fig. 12, A; Fig. 17, B) imbedded in gelatinous parenchyma in close proximity to a blood space. Slender end canals from the flame cells lead to profusely branched collecting tubules and thence to a single thick-walled longitudinal canal on each side of the body. One or more slender efferent ducts lead from the longitudinal canal to the exterior of the body (Fig. 17, B). Occasionally, also, some of the efferent ducts open into the esophagus (Coe, 1906). In such a system the ciliary action of the flame cells may withdraw fluids from the surrounding parenchyma and thence from the contiguous A FIG. 17. Diagrams showing comparison between a simple metanephridium (A) of Cephalothriv and the multiple protonephridium (B) more typical for the nemerteans ; con, convoluted tubule; ct, collecting tubule; cbl, epithelial lining of blood lacuna ; Ic, main longitudinal canal ; ned. efferent duct ; ncf>, nephridiopore, to, terminal organ. TYPES OF XEPHRIDIA IN NEMERTEAXS 215 blood space. After passing through the collecting tubules the fluid enters the longitudinal canal with its thick walls of granular and vacuo- lated cytoplasm, indicative of the secretory or excretory function of this part of the system. After receiving the contributions from the cells of the longitudinal canal, and possibly also returning to those cells any nutrient materials that it may contain, the fluid is discharged through the efferent ducts. The movement of fluids through the sys- tem is facilitated by the delicate cilia with which the longitudinal canal is provided (Fig. 17, B). This system is commonly limited to the region of the body lying between the mouth and the midgut, where the blood spaces are vol- uminous and thin-walled, but in some cases it extends through other regions of the body. In the fresh-water Prostoina, for example, ne- phridia extend the entire length of the body, being separated into several independent groups in the adult, but connected together in early life. In the terrestrial nemerteans, Gconeniertcs, there are many isolated groups of flame cells, each group with a convoluted tubule similar to the longitudinal canal in nature, and with its own efferent duct (Fig. 16). The number of such isolated nephridia may be very great and their extent may cover the greater part of the body. They are found not only in the vicinity of the lateral blood vessels but also in the parenchyma beneath the intestine and beside the proboscis sheath (Coe, 1929). As many as 35,000 are estimated to be present in one of the terrestrial forms which has a body length of only 35 mm. (Schroder, 1918). The terminal chamber in these forms is relatively large and its wall of much complexity (Figs. 12, C '; 16). Although the terminal organ of Geoneuiertcs is composed of a binucleate flame cell and a cylindrical collar cell, we know of no transi- tion stage between this protonephridium and the multinucleate metane- phridium of the female Ccphalothri.r. And although the convoluted tubule of the latter is apparently homologous with the longitudinal canal of the protonephridium, the terminal organs of the two types seem to have originated independently, somewhat as have the larval protonephridia and the adult metanephridia of the Annelids. But the question as to whether the metanephridium of Cephalothrix is preceded in the life history by an earlier excretory system of the protonephridial type remains at present unanswered. LITERATURE COE, W. R., 1906. A Peculiar Type of Nephridia in Nemerteans. Biol. Bui!.. 11: 47. COE, W. R., 1920. Sexual Dimorphism in Nemerteans. Biol. Bull.. 39: 36. COE, W. R., 1928. A New Type of Nephridia in Nemerteans. .-Inat. Rcc.. 41: 57. 216 WESLEY R. COE COE, W. R., 1929. The Excretory Organs of Terrestrial Nemerteans. BioL Bull., 56: 306. SCHRODER, O., 1918. Beitrage zur Kenntniss von Geonemertes palaensis Semper. Scnckcn. natur. Gcscllschaft., 35: 155. STRUNK, CARMEN, 1930. Beitrage zur Excretions-Physiologic der Polychaten Arenicola marina und Stylarioides plumosus. Zool. Jahrb., Abt. f. allg. Zool. u. Physio!, d. Tiere., 47: 259. WIJNHOFF, G., 1910. Die Gattung Cephalothrix und ihre Bedeutung fur die Systematik der Nemertinen. Zool. Jahrb., Abt. f. Anat., 30: 427. BLOOD SUGAR AND ACTIVITY IN FISHES WITH NOTES ON THE ACTION OF INSULIN" I. E. GRAY AXD F. G. HALL (From the Zoological Laboratory, Duke University) The blood sugar of fishes has been studied by numerous investigators and great variations in amount have been reported for different species. In most cases a given observer has worked on one or a very few species. and correlations between the amount of sugar and the habits of the fishes have not been attempted. Furthermore, it is difficult to compare the results of different authors, since so many methods of determining blood sugar haAre been employed. Macleod (1926), has suggested that the more active fishes have higher blood sugar than do the more slug- gish forms. One of us (Gray, 1929), has also pointed out a correlation between activity and blood sugar, but detailed data were not given. Among the mammals, Shirley (1928) hints at a tendency for low blood sugar to accompany high activity. She, however, did not make a comparative study, but limited her observations to a single species. In a previous paper (Hall and Gray, 1929), a correlation was pointed out between the habits of marine fishes and their hemoglobin concentration. It was shown that among fifteen species of marine teleosts, in general, the most active had the highest iron values, while the blood of sluggish fishes had low iron content. The fishes with the highest iron content were surface feeding forms with similar habits, and fishermen consider them among the fastest swimmers. The highest hemoglobin was noted among members of the families Scombridse and Clupeidse, examples of which are, respectively, the mackerels and men- haden. These fishes feed largely on plankton and small fishes, which they can only obtain by constantly keeping in motion. At the other extreme are the bottom feeders, such as the goosefish, toadfish, and sand dab, which are very sluggish and have extremely low hemoglobin. These forms remain quiescent on the bottom for long periods of time. Between the two extremes are found the majority of fishes. In the present paper it is shown that correlations similar to those between hemoglobin and activity exist also between blood sugar and activity. This work was carried on at the United States Fisheries Station at Woods Hole, Massachusetts. 217 218 I. E. GRAY AND F. G. HALL MATERIALS AND METHODS , Blood sugar determinations were made on fifteen different SQ^cies of teleosts, representing thirteen families, as shown in Table I.^Blie fishes were obtained from commercial fish traps and were caremlly placed in large floating " live cages," where they were kept free froi asphyxial conditions for at least twenty- four hours before use. T1 importance of keeping the fishes free from asphyxial conditions cannot be overemphasized. In a previous paper (Hall, Gray, and Lepkovsky, 1926), the changes that take place in the concentration of the blood con- stituents of fishes under asphyxia were pointed out. Other workers (McCormick and Macleod, 1925; Simpson, 1926; and Menten. 1927) have noted that asphyxia tends to raise the blood sugar. The time re- quired for fishes brought to the laboratory to recover from the partial asphyxia to which they have been subjected incidental to capture and transportation varies, of course, with the different species and with the methods of handling, both before and after they are placed in the " live cage." McCormick and MacLeod found that it required from two to four days for asphyxial hyperglycemia of the sculpin to subside. With our methods and facilities it was found that one full day was generally enough time to allow for recovery from any asphyxia to which the fishes might have been subjected. Menton (1927), concludes that the variation in sugar content of a species is governed largely by the amount of food ingested. In our experiments the food factor was reduced to a minimum by not using the fishes for a day or more after placing them in the " live cage." The methods of procedure were similar to those employed in pre- vious studies. The puffers, toadfish, and goosefish were bled from the heart with a hypodermic needle. The other fishes were bled In- severing the tail and collecting the blood from the caudal vessels in a small Erlenmeyer flask. Lithium oxalate was used as an anti-coagulant. The blood sugar was determined by Folin's modification of the Folin-\Yu method (Folin, 1926; Folin and Svedberg, 1926). A large percentage of the determinations was made on the same sample of blood used for the iron determinations (Hall and Gray, 1929), to which reference has previously been made. One fish was used for each determination. During the study of the action of insulin the fishes were kept in hatchery boxes, one fish to each box. Insulin from Eli Lilly and Com- pany was used throughout. The insulin was administered by intra- peritoneal injections, in doses of five to fifteen units, depending on the size and species of fish. If the action of insulin in fishes is similar to its action in mammals, overdoses were given in each case. m BLOOD SUGAR AND ACTIVITY IN FISHES TABLE I The Blood Sugar of Marine Fishes 219 No. of Sugai • per 100 c c. of Blood nations Low High Average Group I Bull's eye mackerel (Pneionatophorus colias) Scombridae 10 Mg. 60 ? Mg. 160.0 Mg. 90.7 Butterfish (Poronottis triacanthus) Menhaden (Brevoortia tyrannies) Stromateidae Cltipeidae 8 30 57.5 s? q 113.6 151.5 79.4 75.2 Rudderfish (Palinurichthys perciformis) . . . Common mackerel (Scomber scombrns) Centrolophidae Scombridae 7 9 54.9 48.5 83.3 76.6 67.7 63.5 Eel (Anguilla, rostrata) Anguillidae 4 40.6 67.6 59.0 Bonito (Surda, sarda) Scombridae 3 48 ,S 62.7 55.1 Scup (Stenotomus chrvsops) Sparidae 46 SS ^ 81.4 52.6 Silver hake (Merhiccius InMnearis) Merlucciidae 9 ?S 3 85.4 48.2 Group II Sea robin (Pnonotus carolinus) Triglidae 9 ?08, 60.9 37.4 Sand dab (Lopliopsettu tnaculatd) Pleu ronect idae 4 ?4 6 42.5 31.0 Cunner (Tauto^olabrus adspersus) Puffer (Spheroides maculatits) Labridae Tetraodontidae 4 15 13.4 45 35.1 41.3 25.2 23.1 Toadfish (Opsanus tau) Batrachoididae 6 102 22.3 15.4 Goosefish (Lophius piscatorius) Lophiidae 11 00 10.3 5.6 RESULTS AND DISCUSSION The results of the blood sugar determinations of the fifteen species of marine teleosts are given in Table I. The fishes were kept under conditions approximating the normal as nearly as possible. The high and the low blood sugar values are given together with the average to show the individual variation within the same species. The high and the low values may seem exceedingly far apart in a few cases, and without explanation may be misleading. The great majority of blood 15 220 I. E. GRAY AND F. G. HALL sugar determinations gave results near the average ; it was only occa- sionally that a very high or exceedingly low value was obtained. There appears to be, however, a relatively greater individual variation among fishes of the same species kept under the same conditions, than among mammals. Blood sugar, like hemoglobin, appears to be correlated in a general way with the habits and activity of the fishes. The bull's eye mackerel, butterfish, menhaden, rudderfish, common mackerel, eel, bonito, scup, and silver hake are not only more active fishes than the others given in the table, but also have higher blood sugar. For convenience of discussion the fishes are arbitrarily divided into groups I and II. There is no sharp dividing line, however, between the two groups. Group I consists, for the most part, of aggressive fishes that depend on their own activities in obtaining food. They feed largely on plank- ton, small fishes, or other small animals that require the expenditure of considerable effort to obtain. Members of the Scombridie and • Clupeidas are especially noted for their great activity. Individuals of these families are kept in captivity only with great difficulty even when placed in large " live cages " where they have plenty of room for their constant movements. It is doubtful if the bonito, mackerels, and men- haden ever cease their movements. Some fishes of group I, for example, the scup and hake, might well be classed as intermediate in regard to their activity. They are not always in motion, nor do they pursue their food with the aggressiveness shown by the Scombridse and Clupeidse. The fishes are arranged in the table, not in the order of their activity, but according to their blood sugar content. The correlation between activity and blood sugar is not absolute but occurs in most cases. If arranged according to relative activity the bonito would be at or near the top. The blood sugar of this species may not be strictly comparable with that of the other fishes since it was impossible to keep the bonito alive in captivity. Consequently the only data obtainable were deter- minations made on three small specimens, bled as soon as brought from the traps. Some of group I are excellent migrating fishes that move rapidly through the water in large schools. Mostly they are adapted for fast movement by being " stream-lined " with body-form either fusiform or laterally compressed. In group II are the relatively inactive and sluggish fishes. In con- trast to the majority of group I, these fishes are the less aggressive bottom feeders that are adapted to life on the bottom by having the body- form angular or depressed. The dinner, although having a body- BLOOD SUGAR AND ACTIVITY IN FISHES 221 form resembling members of group I and being found in a variety of habitats, seems to prefer the rocky bottom and does not roam over wide areas in search of food. It will be noted that the average blood sugar of the members of this group is considerably lower than that of group I. The goosefish and the toadfish are two of our most sluggish fishes and have very low blood sugar. Many determinations of the goosefish blood showed merely faint traces of sugar. The goosefish, although it feeds indis- criminately on other fishes, does not as a rule pursue its food. It is one of the anglers and attracts its prey by a lure on one of the dorsal fin-rays. All of the fishes of this group are known to remain quiet on the bottom for long periods of time, which habit is in sharp contrast to the activities of the mackerels and menhaden. TABLE II The Effect of Insulin on the Blood Sugar of Fishes No. of Units of Time for ShnrL' tn Sugar per IOC ) cc. of Blood nations Given Appear Normal After Insulin Group I Menhaden 6 5 Hours 11- 3 Mgs. 75.2 A/gs. 8.6-20.2 Common mack- erel 13 5 1- 4 63.5 9.4-31.2 Bull's eye mack- erel 5 5 3- 6 90.7 9.4-11.1 Scup 20 5-10 10-23 52.6 0.0-15.3 Group II Sea robin . 10 5-15 no shock 37.4 8.8-32.5 Puffer . . . 9 5-15 no shock 23.1 0.0-13.5 Toadfish 15 5-15 no shock 15.4 1.5-22.9 We may say, then, that there appears to be a general correlation between the amount of sugar of the blood, the hemoglobin, the body- form, the activitv, and the habits of marine fishes. Activitv is ex- ^ ' -• pressed here qualitatively. There appears to be a dearth of quantitative determinations of metabolic activity in fishes. The oxygen consumption of the scup, puffer, and toadfish has been studied (Hall, 1929), and the results bear out our estimate of the activity of these fishes. Under the same conditions the oxygen consumption of the puffer was found to be intermediate between the relatively high consumption of the scup and the extremely low oxygen consumption of the toadfish. Because of their great activity a comparable basal oxygen consumption of the Scombridae and Clupeidse, fishes more active than the scup, could not be determined. I. E. GRAY AND F. G. HALL A further interesting relation to activity was noted through a com- parative study of the action of insulin on fishes. At a temperature of 21° C. and under similar conditions, it was found that the very active fishes, menhaden, common mackerel, and bull's eye mackerel, showed insulin shock in a much shorter time than did the moderately active scup. This was perhaps to be expected. Huxley and Fulton (1924), and Olmsted (1924), have pointed out that the rate of action of insulin is dependent upon the metabolic rate of the animal itself. The more sluggish bottom feeders, sea robin, puffer, and toadfish. showed no external evidences of the effects of insulin. As has been previously noted, the normal blood sugar of these sluggish fishes is much lower than that of the more active ones. In some cases, such as the toadfish. the normal sugar concentration is not as high as the insulin-reduced sugar concentration of the more active fishes. A condensed summary of the action of insulin on fishes is given in Table II. Insulin appears to reduce the blood sugar concentration of fishes in much the same manner as in mammals, except that a longer time is required for the action to take place. Although the number is limited, at least some of each species whose blood was analyzed showed reduced sugar concentration following insulin administration. The mackerels, menhaden, and scup, if bled during convulsions, showed reduced sugar content in each case. There is considerable individual variation in the time required for the sugar content to be reduced ; and since the sluggish fishes showed no convulsions, it was difficult to estimate the length of time to allow for insulin action. Puffers, sea robins, and toadfish, bled at various intervals between twenty and forty hours after insulin injection, showed blood sugar values ranging from the normal to mere traces. Since some of each of these species showed reduction of sugar content, it is thought that in those cases wrhere, after insulin administration, the blood sugar was within the normal range of varia- tion, either enough time had not elapsed for the insulin to reduce the sugar concentration, or else too much time elapsed and the fishes re- gained the normal sugar content. The time required for the blood sugar content to be reduced ap- peared to be considerably greater in these sluggish forms than in the more active fishes. It seems improbable that the failure to get insulin shock could be due to insufficient insulin. Toadfish were given re- peated injections of from five to fifteen units of insulin over a period of several days with no visible signs of disturbed metabolism. With the mackerels, menhaden, or scup a single five unit injection usually resulted in death unless glucose was administered. Insulin convulsions in fishes do not necessarily indicate that the BLOOD SUGAR AND ACTIVITY IN FISHES blood sugar concentration is reduced to its lowest level. The rate of reduction of the blood sugar values following insulin injection has been worked out for the scup and will be published later. Here it is sufficient to say that the blood sugar content may be reduced in six to eight hours in this fish. In a few cases mere traces of sugar remained in the blood after eight hours and yet in no case were convulsions ap- parent sooner than ten hours. In other words, a few scup had lower blood sugar before reaching the convulsive stage than did other scup in the midst of convulsions. Furthermore, the fact that sluggish fishes, as the toadfish and puffer, have their blood sugar concentration reduced without showing any shock at all, indicates that insulin shock in fishes does not have as much significance as has been attributed to insulin convulsions in mammals. SUMMARY 1. Correlations between the blood sugar, hemoglobin, body- form, activity and habits of fifteen species of marine teleosts are pointed out. 2. The fishes that show the greatest activity, those that feed at the surface or are aggressively predaceous, have the highest blood sugar concentration. The sluggish bottom feeders have low sugar content in the blood. 3. Insulin shock may be easily produced in active species of fishes. In sluggish forms no external evidence of the action of insulin could be detected. 4. The blood sugar of fishes is reduced by the action of insulin. Less time is required for reduction of sugar content to take place in the active fishes than in the sluggish forms, due probably to differences in the metabolic rate of the different species. In the sluggish forms the sugar content may be reduced without convulsions or shock being apparent. 5. The normal sugar of some of the sluggish fishes is often lower than the insulin-reduced sugar of the more active fishes. STUDIES OF PHOTODYNAMIC ACTION I. HEMOLYSIS BY PREVIOUSLY IRRADIATED FLUORESCEIN DYES HAROLD F. BLUM DEPARTMENT OF PHYSIOLOGY, HARVARD MEDICAL SCHOOL x The hemolysis of red blood cells by the combined action of light and certain photoactive substances was first described by Sacharoff and Sachs in 1905. Such hemolysis occurs in a very short time when red blood cells are exposed to sunlight in dilute concentrations of the photoactive substance. Sunlight alone does not produce hemolysis provided the ultra-violet spectrum is screened out by exposing the cells in glass, nor does the photoactive substance in equal concentration in the dark. This is only one of a wide range of similar phenomena brought about under similar conditions in other cells and tissues, which are generally described collectively under the term photodynamic action or photodynamic sensitisation. The photoactive substances which bring these phenomena about include a large number of compounds, most of which are fluorescent dyes. It is generally assumed that such effects are not produced if the solution of the photodynamic substance is sep- arately irradiated, the erythrocytes or other cells being added subse- quently in the dark (see Clark 1922, p. 288). There are, however, a few recorded experiments which indicate that this is possible. Ledoux-Lebard (1902) found that eosine which had been previously exposed to sunlight killed and cytolyzed paramecia; whereas non- irradiated eosine of the same concentration did not. He suggested, therefore, that photodynamic action is due to the formation of a toxic eosine compound by the action of sunlight. Jodlbauer and Tappeiner (1905) found that this did not occur if the eosine solution was neutral- ized after irradiation and before the addition of the paramecia. They claimed, therefore, that the Ledoux-Lebard effect was due to the forma- tion of acid concomitant with the bleaching of the dye ; the acid being the toxic agent. They did not consider this as a photodynamic effect. Sacharoff and Sachs (1905) described hemolysis by previously irradiated £-(o-nitrophenyl)— /Miydroxyethyl methyl ketone [" o-Nitrophenyl- milchsaureketon "]. They were unable to produce hemolysis with pre- viously irradiated eosine or erythrosine, however, and preferred to 1 Preliminary experiments for these studies were carried out in the Depart- ment of Animal Biology, University of Oregon. 224 HEMOLYSIS BY PREVIOUSLY IRRADIATED DYES 225 consider their one positive result as not belonging to the typical photo- dynamic phenomena. Fabre and Simonnet (1927) were able to pro- duce hemolysis with lecithin which had been irradiated together with hematoporphyrin by light from a mercury vapour arc. Moore (1928) found that previously irradiated cosine killed the eggs of the sea urchin Strongylocentrotus purpuratiis but did not cytolyze them; whereas when the eggs were irradiated together with the dye, they were completely cytolyzed. Moore hypothecates the formation of a toxic cosine com- pound which produces cytolysis upon further irradiation after it has entered the cell. On the other hand, Raab (1900) was unable to pro- duce killing of paramecia by previously irradiated acridine solutions. Hausmann was unable to produce killing of paramecia or hemolysis with previously irradiated solutions of chlorophyll (1909) or hemato- porphyrin (1910) ; although similar solutions produced these effects when irradiated together with the cells. Hasselbach (1909) could not produce hemolysis with previously irradiated solutions of several photo- dynamically active substances including cosine and erythrosine. Pereira (1925) found that Arbacia larvae were not killed by previously irra- diated cosine in sea water. The writer has found that it is possible, under carefully controlled conditions, to bring about hemolysis with previously irradiated solutions of the three fluorescein dyes which he has investigated, fluorescein, eosine and erythrosine. This is of considerable interest because of its bearing on certain theories of photodynamic action which will be discussed later in this paper. EXPERIMENTAL Hemolysis by previously irradiated fluorescein, cosine, and erythro- sine.— The writer's first attempts to produce hemolysis with previously irradiated eosine solutions met with apparent success in only a few in- stances. These were thought at first to be accidental, but with more careful control of conditions it was found possible to obtain consistently reproducible results. The successful technique required the selection of proper hydrogen ion concentration and dye concentration. The hydrogen ion concentration must be carefully buffered, since unbuffered solutions tend to increase in acidity during irradiation. This increase in acidity may inhibit the production of hemolysis by bringing about fixation of the cells as will be pointed out in a later paper. To insure the maximum obtainable buffering capacity, it was found convenient to make up the dye in solutions of primary and sec- ondary sodium phosphate mixtures. In order to insure a medium of proper osmotic pressure for the blood cells, the phosphate mixtures were 226 HAROLD F. BLUM calculated to have the same osmotic pressure as a 0.15 M sodium chloride solution. This was done by assuming that the primary phosphate dis- sociates into two ions, the secondary phosphate into three. The mol fractions of the two salts required for a given hydrogen ion concentra- tion were estimated by the use of Cohn's data for potassium phosphates (see Clark, 1928, pp. 216-220). ~ The hydrogen ion concentrations of the solutions were checked by means of the hydrogen electrode. Such solutions proved rather unsatisfactory in the case of fluorescein, and a solution containing 10 per cent of the phosphate mixture and 90 per cent 0.15 M sodium chloride, was used instead in most experi- ments with this dye. The concentration of phosphate in this solution is still many times that of fluorescein in most of the dye concentrations which were used, and affords an adequate buffer. The optimal concentration of the dye varies with a number of con- ditions; some of which, as for example the intensity of irradiation, it was impossible to control. It was found expedient, therefore, to use a series of dilutions of the dyes ; usually consisting of ten dilutions from 1 per cent to 0.002 per cent." These were exposed to the sunlight for a given period of time. Blood cells were then added to the ir- radiated solutions and also to a control consisting of a corresponding series of non-irradiated dye solutions. Both series were then placed in a dark room where the temperature was in the region of 20° C. Observation of the tubes for hemolysis was made at intervals after the addition of the cells. It was found that in most cases six hours sufficed for the hemolysis to reach a maximum. Since the temperature during irradiation could not be controlled, time was allowed, when necessary, for the irradiated tubes to come to the same temperature as the controls before adding the cells. The solutions were exposed in small test tubes (10x75 mm.), each containing 2 cc. of the solutions. The blood cells were added to each tube in the quantity of 0.02 cc. of a 50 per cent suspension in 0.15 M sodium chloride, by means of a blood pipette. This method avoids any appreciable dilution of the irradiated solution upon the addition of the cells. This precaution has not been observed by most of the investigators who have attempted to produce hemolysis with previously irradiated substances. Human blood cells were used 2 Dr. G. Payling Wright and the writer have found that rabbit blood cells suspended in such solutions show a variation in volume of approximately twelve per cent over the range of hydrogen ion concentration between pH 7.7 and pH 6.0, and have approximately the same volume as cells in serum. 3 The dyes used were Fluorescein, sodium salt (Uranine), from the National Aniline and Chemical Company, Erythrosine B (sodium salt of tetra-iodo- fluorescein) also from the National Aniline and Chemical Company, and Eosine Y (sodium salt of tetra-brom-fluorescein) from Coleman and Bell. HEMOLYSIS BY PREVIOUSLY IRRADIATED DYES 227 in most of the experiments. They were washed by centrifuging three times from suspension in 0.15 M sodium chloride to free them from serum. It is advisable to have the cells as free from serum as possible, since serum is effective in preventing photodynamic hemolysis (Busck, 1906). The intensity of the radiation could not be accurately estimated, but it was found practicable to expose the solutions to bright midday sunlight for one to two hours. Too long continued exposure causes bleaching of the dye, resulting in a lowered concentration of the active dye. TABLE I Hemolysis by Previously Irradiated Fluorescein Solutions exposed to sunlight 90 minutes (2:00-3:30 P.M., September 29, 1929). All solutions contain 10 per cent of sodium phosphate buffer, pH 6.4, isosmotic with 0.15 M NaCl, plus 90 per cent of 0.15 M NaCl. Observations made after 16 hours in dark following addition of red blood cells. H = complete hemolysis, (H) = partial hemolysis, and the dash is used when there is no detectable hemolysis. Concentration of Irradiated solution. Red fluorescein blood cells added after Non-irradiated solution (control) per cent 45 minutes 4 hours 1.0 — — 0.5 — — . — - 0.25 — — . — 0.125 (H) • — — 0.062 (H) (H) — • 0.031 H (H) — 0.015 H H — 0.007 (H) (H) — 0.004 (H) (H) — 0.002 — — — 0.00 — — — Tables I, II, and III show the results of typical experiments with fluorescein, cosine and erythrosine respectively. In these tables, H represents complete hemolysis (i.e. hemochromolysis and stromatolysis) as well as can be judged by the naked eye, (//) represents partial hemolysis, and the dash no detectable hemolysis. These classifications are arbitrary, but since comparison can always be made with the control tubes, there can be no doubt of the general validity of the observations. An examination of Tables I, II, and III demonstrates quite clearly that previously irradiated solutions of these dyes bring about hemolysis in concentrations at which non-irradiated solutions do not. Some bleach- ing of the dye takes place upon irradiation and this raises the question whether the hemolysis may not be due to the products of this bleaching. 228 HAROLD F. BLUM It has been found, however, that completely bleached solutions have no hemolytic action. Non-irradiated cosine and erythrosine produce hemolysis in suf- ficiently high concentration, as is shown in Tables II and III. This was described by Sacharoff and Sachs (1905) and studied by Tappeiner (1908). It is apparently not due to irradiation during the preparation of the solutions ; since in these experiments the results were the same when the solutions were carefully prepared in the dark room under red light, which is outside the absorption range of these dyes, as when TABLE II Hemolysis by Previously Irradiated Eosine Solutions exposed to sunlight for 105 minutes (11:45A.M.-1:30 P.M., September 10, 1929). All solutions contain sodium phosphate buffer, pH 7.0, isosmotic with 0.15 M NaCl. Observations made 5 hours after addition of red blood cells. The symbols are the same as those in Table I. Concentration of eosine Irradiated solution. Red blood cells added after Non-irradiated solution (control). Red blood cells added after per cent 45 minutes 2-'i4 hours 5 hours 45 minutes 5 hours 1.0 H H H H H 0.5 0.25 H H H H H H (H) (H) 0.125 H H H — — . 0.062 H H H — — 0.031 0.015 0.007 H (H) (H) (H) H (H) — — 0.004 . — • — — — — 0.002 — — — — — 0.0 — — — — — prepared with ordinary precautions in the diffuse light of the laboratory. The effect of short exposure to diffuse light is thus within the accuracy of the observations described here. The absence of hemolysis in the higher concentrations of the irradiated dye in Table I is probably due to fixation of the cells. This phenomenon will be discussed in a later paper. The marked effect of hydrogen ion concentration upon the hemolytic activity of irradiated and non-irradiated dyes will also be discussed in that paper ; the hydrogen ion concentrations for the ex- periments here described have been chosen as those at which the differ- ence in hemolytic activity between previously irradiated and non-irradi- ated solutions could be most clearly demonstrated. It will be noted in Tables I, II, and III that the results are changed HEMOLYSIS BY PREVIOUSLY IRRADIATED DYES 229 very little when the exposed solutions are allowed to remain in the dark for as much as four or five hours after irradiation, before the addition of the cells. This shows very conclusively that the increased hemolytic activity of the irradiated solutions cannot he due to their having a greater temperature than the controls because of the absorption of heat during the period of exposure, since ample time is allowed for the two series of solutions to come to the same temperature. It also demonstrates that whatever change occurs in the course of irradiation is not rapidly reversible in the dark. Moore (1928) observed, similarly, TABLE III Hemolysis by Previously Irradiated Erythrosine Solutions exposed to sunlight for one hour (11:00 A.M.-12:00 M., September 28, 19291. All solutions contain sodium phosphate buffer, pH 6.5, isosmotic with 0.15 M NaCl. Observations made 6 hours after addition of red blood cells. The symbols are the same as those in Tables I and II. Concentration of Irradiated solution. Red blood erythrosine cells added after Non-irradiated solution (control) per cent 45 minutes 1M hours 5 hours 1.0 H H H H 0.5 H H H H 0.25 H H H H 0.125 H H H H 0.062 H H H (H) 0.031 H H H (H) 0.015 (H) (H) (H) — 0.007 • — • — — — 0.004 — . — — — 0.002 . — — — — 0.00 — — — — that in the case of the killing of sea urchin's eggs by previously irradiated cosine, the solution retained its toxic properties after six hours in the dark. DISCUSSION Numerous hypotheses have been developed to explain the mechanism of photodynamic action, most of which contain the assumption that the photodynamic substance and substrate (e.g. cells) must be irradiated together. This is true of the theory of Tappeiner (1909) which he outlines as follows : The presence of the photodynamic substance merely accelerates the action of visible light. The split products of this reac- tion are removed through oxidation by molecular oxygen. Ordinarily LIB RARY! / • 230 HAROLD F. BLUM these products accumulate and inhibit the reaction, but the combined action of light and a photodynamic substance accelerates their removal and consequently the total reaction. Another conception, based on the fact that most of the photodynamic substances are fluorescent, is that the photodynamic effects are due to the action of fluoresced radiation upon the protoplasm. Since the fluoresced light is only a more or less polarized radiation from a particular region of the visible spectrum characteristic of the substance concerned (Pringshein 1928, p. 195), it can hardly be expected to have such destructive effects. Moreover, Raab (1900) showed that paramecia are not damaged when exposed to the fluoresced radiation from a solution of fluorescent substance with which they are not in contact ; and likewise, Sacharoff and Sachs (1905) showed that red blood cells exposed under the same conditions are not hemolyzed. Nevertheless, this concept remains current to a certain extent. Schanz (1921) suggests from studies on the photo- electric effect in albumin, and albumin plus fluorescein dyes, that the changes brought about in the cell constituents are due to the absorption of electrons emitted by the dye during irradiation. Clark (1922, pp. 302-303) suggests that the photodynamic substance shifts the photo- electric threshold of the cell constituents from the ultra-violet into longer wave lengths. Metzner (1924) claims that the photodynamic effects are brought about by an action within the cell dependent upon the com- bination (adsorption) of the dye with the protoplasm. Jodlbauer ( 1926) assumes that the dye must be adsorbed by the cell, and that only those dyes are photodynamically active which retain their ability to be activated by light while in combination with the cell substance. Such theories demonstrate how firmly the idea is established that the photo- dynamic substance and substrate must be irradiated together. Obvi- ously all such explanations of photodynamic action must be discarded or modified, in light of the fact that hemolysis may be brought about by previously irradiated photodynamic substances. EXPERIMENTAL Evidence that Oxidation Is a Factor in Photodynamic Hemolysis. — A theory of direct oxidation of cell constituents by the action of light and the photodynamic substance was put forward by Straub (1904a). His hypothesis was founded principally upon the analogy between the photodynamic action of cosine upon cells and its ability to oxidize iodide ion in the presence of light. He found (1904^) that, in proper concen- tration, cosine may oxidize many times its equivalency of iodide when the two substances are exposed to sunlight together in solution. He conceived that the cosine is changed to an cosine peroxide by the action HEMOLYSIS BY PREVIOUSLY IRRADIATED DYES 231 of light; and that this peroxide brings about the oxidation of an equivalent amount of iodide ion, being returned in so doing to the original cosine form. The cosine may then proceed to the oxidation of another quantity of iodide, thus acting in a sense as a catalyst. He could not, however, demonstrate the existence of an intermediate per- oxide, being unable to obtain conclusive evidence of the oxidation of iodide ion by the action of previously irradiated cosine (1904a). The writer finds that previously irradiated cosine will oxidize iodide ion, as shown by a positive starch reaction after adding potassium iodide in the dark. The oxidation proceeds rather slowly immediately after the addition of the potassium iodide, which may account for Straub's failure to observe it in his experiments. Table IV" presents some quantitative results obtained when (1) fluorescein dyes and po- tassium iodide were irradiated in solution together, and (2) when the iodide was added to the previously irradiated dyes. The determinations were made by titration of the free iodine formed due to the oxidation of iodide ion, with 0.001 N sodium thiosulfate against starch indicator. When potassium iodide is added to the previously irradiated dye and the mixture placed in the dark, the oxidation takes place quite slowly, reaching a maximum after about three hours. The titrations were, therefore, performed after the elapse of this time. The accuracy of determination of iodine in such small concentrations is, of course, sub- ject to some error. In order to determine the magnitude of this error, solutions containing quantities of iodine of the same order as those represented in Table IV were titrated. The solutions were of the same volume, contained the same concentration of dye and of potassium iodide, and were buffered at the same hydrogen ion concentration as the experimental solutions. With concentrations of iodine correspond- ing to the lowest values in Table IV, the determinations were con- sistently 10 to 15 per cent lower than the theoretical. Writh quantities of iodine corresponding to the highest values the error was not greater than one per cent. The 0.001 N thiosulfate solution was always freshly prepared by dilution from a 0.1 N stock solution. The experiments described in Table IV represent conditions in the region of the optimal for the reaction of the iodide with each clye. The extent of these reactions seems to be greatly affected by the hy- drogen ion concentration, and by other factors, which will not be dis- cussed here. Controls containing the same concentration of potassium iodide, but no dye, never showed more than a trace of free iodine when exposed to sunlight simultaneously with the potassium iodide-dye mix- tures. Likewise, solutions of the dye containing potassium iodide showed no trace of free iodine after many hours in the dark. 232 HAROLD F. BLUM TABLE IV Oxidation of Potassium Iodide by Irradiated Fluorescein, Eosine, and Erythrosine Fluorescein KI PH Volume of Solution Duration of Irradiation KI Added after Irradiation Volume of Na2S2O3 Mols of Dye Mols of Iodide Oxidized 0.00 IN per cent Cc. Hours per cent Cc. 0.0005 M 1.0 6.0 6.0 8 0.0 18.6 3 X 10-6 18.6 X 10~6 0.0005 M 1.0 6.0 6.0 8 0.0 17.1 * 3 X 10-6 17.1 X 10-" 0.0005M 1.0 6.0 6.0 0 0.0 0.0 3 X IQ-o 0.0 0.0005M 0.0 6.0 6.0 8 3.0 1.0 f 3 X 10-« 1.0 X 10-6 0.0005 M 0.0 6.0 6.0 8 3.0 0.5 | 3 X lO-6 0.5 X 10-« 0.0 1.0 6.0 6.0 8 0.0 0.4 0.0 0.4 X 10~6 Eosine per cent Cc. Hours per cent Cc. 0.001 M 3.0 6.0 6.0 6 0.0 14.5 6 X 10-6 14.5 X 10-6 0.00 1M 3.0 6.0 6.0 6 0.0 14.5 * 6 X 10-6 14.5 X lO-6 0.001 M 3.0 6.0 6.0 0 0.0 0.0 6 X 10-« 0.0 0.001 M 0.0 6.0 6.0 6 3.0 3.2 f 6 X lO-6 3.2 X 10~6 0.001 M 0.0 6.0 6.0 6 3.0 3.0 } 6 X 10-6 3.0 X 10-6 0.001 M 0.0 6.0 6.0 0 3.0 0.0 6 X 10-6 0.0 Erythrosine per cent Cc. Hours per cent Cc. 0.001M 3.0 6.0 6.0 6 0.0 19.1 6 X 10~6 19.1 X 10-fi 0.001 M 3.0 6.0 6.0 6 0.0 19.5 * 6 X 10-6 19.5 X 10-6 0.00 1M 3.0 6.0 6.0 0 0.0 0.0 6 X 10~6 0.0 0.001M 0.0 6.0 6.0 6 3.0 1.3 t 6 X 10-6 1.3 X 10-6 0.001 M 0.0 6.0 6.0 6 3.0 1.0} 6 X 10-6 1.0 X lO-6 0.001 M 0.0 6.0 6.0 0 3.0 0.0 6 X 10~6 0.0 * Titration after 3 hours in dark. f KI added immediately after irradiation with titration after 3 hours in dark. J KI added after 3 hours in dark following irradiation; titration 3 hours later. Table IV shows that iodide ion equivalent to several times the quan- tity of dye present may be oxidized when exposed together with the dye (equivalency considered as one mol of iodide ion per mol of dye). Straub (1904b) was able, in fact, to oxidize a quantity of iodide ion sixty-five times as great as the quantity of dye present. On the other hand, when the dye alone is irradiated and the potassium iodide added subsequently in the dark, the quantity of iodide ion oxidized is always less than that equivalent to the dye present. In the latter case it was never found possible, in a considerable number of experiments under varving conditions, to oxidize more iodide than a quantity equivalent HEMOLYSIS BY PREVIOUSLY IRRADIATED DYES 233 to the quantity of dye present. When irradiated in the absence of a readily oxidizable substance, such as iodide ion, a certain amount of the dye is oxidized, as is indicated by bleaching. Thus the transforma- tion of all the dye to the active form cannot be expected, and we should expect that less iodide would be oxidized than a quantity equivalent to the quantity of the dye originally present. This appears to be the case. When the dye is exposed with a readily oxidizable substance, no bleach- ing occurs, indicating that this substance is oxidized instead of the dye. All these facts lend support to Straub's hypothesis. They demonstrate at least that a substance is formed upon irradiation of the dye solution which is capable of oxidizing substances which the non-irradiated dye cannot, and indicate that this is an intermediate substance in the oxida- tions brought about by the action of the dye and light. The quantity of iodide ion oxidized is not greatly altered if the dye is allowed to remain in the dark for several hours after irradiation before potassium iodide is added. This shows that the change brought about by irradiation is not rapidly reversible in the dark. This is ex- actly parallel to the case of hemolysis where, as we have seen, hemolysis is brought about by previously irradiated dye solutions which have re- mained in the dark for several hours after irradiation before addition of the cells. Substances produced in the bleaching of the dye are not responsible for the oxidation of iodide ion, since completely bleached solutions do not bring about this oxidation. This is again parallel to the case of hemolysis, since as stated above, hemolysis is not produced by completely bleached dyes. These latter facts suggest very definitely that the substance in irradiated solutions of a fluorescein dye which brings about hemolysis is the same as that which brings about the oxidation of iodide ion ; and that, therefore, the former process is prob- ably dependent upon an oxidation. If it is true that the hemolysis of blood cells by irradiated dyes in- volves the oxidation of cell constituents in a manner similar to the oxi- dation of iodide ion, we should expect, parallel to the above observations, more extensive oxidation and thus greater hemolysis when the dye is irradiated together with the cells than when previously irradiated. In the former case the dye may, presumably, act in a catalytic sense, thus oxidizing several times its molecular equivalency of cell constituents ; whereas in the latter case the amount of oxidation is limited by the quantity of dye present. The data presented in Tables V, VI and VII, appears to confirm this prediction ; the hemolytic action seems to be quantitatively much greater when the dye and cells are irradiated to- gether than when the dye is irradiated alone and the cells added later in the dark. The statement that hemolvsis is more readily produced 234 HAROLD F. BLUM TABLE V Comparison of Hemolytic Activity of Fluorescein Irradiated With and Without Blood Cells Solutions exposed to sunlight for one hour and 30 minutes. All solutions contain sodium phos- phate buffer, pH 6.5,isosmotic with 0.15 M NaCl. Observations made after 20 hours in dark following addition of blood cells. Symbols as in preceding tables. P = precipitate. Concentration Fluorescein Solution Fluorescein Irradiated Fluorescein of Irradiated Alone. Not Fluorescein with Cells Cells Added in Dark Irradiated per cent 1.0 P (H) — 0.5 P (H) — 0.25 P (H) — 0.125 H (H) — 0.062 H (H) — 0.031 H (H) — 0.015 H (H) — 0.007 H — — • 0.004 H — — 0.002 H — — 0.0 — — — when the dye is irradiated together with the cells than when irradiated separately is a generalization to which many exceptions occur, due chiefly to the complicating factor of fixation which will be considered in a later paper. That hemolysis may proceed farther in the former case than in the latter, in conditions where fixation is not a complicating TABLE VI Comparison of Hemolytic Activity of Eosine Irradiated With and Without Blood Cells Solutions exposed to sunligh tfor one hour and 30 minutes. All solutions contain sodium phos- phate buffer, pH 6.9, isosmotic with 0.15 M NaCl. Observations made after 7 hours in dark following irradiation. Symbols as in the preceding tables. Concentration of Eosine Eosine Solution Irradiated with Cells Eosine Irradiated Alone. Cells Added in Dark Eosine Not Irradiated per cent 1.0 H H H 0.5 (H) H (H) 0.25 H H — 0.125 H H — 0.062 H H — 0.031 H H — 0.015 H (H) — 0.007 H — — 0.004 H • — — 0.002 H — — 0.0 • — — ' HEMOLYSIS BY PREVIOUSLY IRRADIATED DYES 235 factor, seems justified by all the writer's observations on red blood cells under the two conditions. It would be, of course, absurd to at- tempt an exact quantitative comparison between the results in Table IV and those in Tables V, VI, and VII, since we do not know in the case of the blood cells, what substances may be subject to oxidation, or what their oxidation-reduction potentials may be. The action of non-irradiated dyes, previously mentioned, is in all probability not an oxidative process, since oxidation of iodide ion by these dyes does not take place in the dark. Whatever the nature of this process, however, when hemolysis occurs after irradiation in a concentration of dye which does not produce hemolysis when not ir- radiated, we are justified in the assumption that the changes bringing about hemolysis may be oxidative, since we know that the oxidizing power of the dye solution has been increased by irradiation. TABLE VII Comparison of Hemolytic Activity of Erythrosine Irradiated With and Without Blood Cells Solutions exposed to sunlight for one hour. All solutions contain sodium phosphate buffer. pH 7.0, isosmotic with 0.15 M NaCl. Observations made after 6 hours and 20 minutes in dark following irradiation. Symbols as in preceding tables. Concentration of Erythrosine Erythrosine Solution Irradiated with Cells Erythrosine Irradiated Alone. Cells Added in Dark Ervthrosine " Not Irradiated per cent 1.0 H H H 0.5 H H H 0.25 H H H 0.125 H H H 0.062 H H H 0.031 H H — 0.015 0.007 H H (H) — 0.004 H — — 0.002 H — — 0.0 — — • — DISCUSSION Further evidence that oxidation is an important factor in photo- dynamic processes is not lacking. Oxygen is known to be necessary for a number of photodynamic effects (Straub, 1904a; Jodlbauer and Tappeiner, 1905), since they do not take place in its absence. Spe- cifically as regards hemolysis, Hasselbach (1909) found that hemolysis by light and certain photodynamic substances, including cosine and 16 236 HAROLD F. BLUM erythrosine, did not take place in a vacuum, and Schmidt and Norman (1922) found that hemolysis by cosine and sunlight did not occur in hydrogen. Sacharoff and Sachs (1905) showed that the presence of the reducing substance sodium sulfate may prevent hemolysis by ir- radiated erythrosine. Noack (1920) showed that a number of in- organic reducing agents may inhibit photodynamic effects, and Schmidt and Xorman (1922) found that a number of readily oxidizable organic and inorganic substances will prevent hemolysis by cosine and light. Noack (1920) has also shown quite definitely that certain plant pig- ments can be oxidized by various photodynamic substances and light, and gives evidence that these phenomena involve the formation of in- termediate peroxides. CONCLUSIONS The demonstration of the formation of an intermediate substance in the process of photodynamic hemolysis by fluorescein dyes offers quite conclusive evidence against the sensitization theory of Tappeiner and other theories which assume that photodynamic substance and sub- strate must be irradiated together. The demonstration that a definite increase in the oxidizing power of solutions of these dyes is brought about by irradiation, together with the accumulation of other evidence pointing toward an oxidative process, makes it necessary to consider the oxidation of cell constituents as a probable underlying factor in photodynamic hemolysis. Likewise, such oxidations must be considered as a possible factor in all photodynamic processes. SUMMARY 1. Hemolysis may be produced by previously irradiated fluorescein, cosine and erythrosine. 2. Similarly, previously irradiated fluorescein, cosine and erythrosine oxidize iodide ion. 3. These findings render untenable the sensitization theory of Tap- peiner and other theories which necessitate the simultaneous action of light and the photodynamic substance, while supporting Straub's theory of direct oxidation of cell constituents. 4. Oxidation must be considered as a probable underlying cause in photodynamic hemolysis and all other photodynamic phenomena. BIBLIOGRAPHY BUSCK. G., 1906. Die Photobiologischen Sensibilisatoren und ihre Eiweissverbin- dungen. Biochcm. Zcitschr., 1: 425. CLARK, J. H., 1922. The Physiological Action of Light. PJiysioL Rer., 2: 277. HEMOLYSIS BY PREVIOUSLY IRRADIATED DYES 237 CLARK, W. M., 1928. The Determination of Hydrogen Ions. Third Edition. Baltimore. FABRE, R., AND SIMONNET, H., 1927. Contribution a 1'etude de 1'hemolyse par action photosensibilisatrice de l'hematoporphyrine. Compt. rend. Acad. Sci, 184: 707. HASSELBACH, K. A., 1909. Untersuchungen iiber die Wirkung des Lichtes auf Blutfarbstoffe und rote Blutkorperchen wie auch iiber optische Sensi- bilisation fur diese Lichtwirkungen. Biochem. Zcitschr., 19: 435. HAUSMANN, W., 1909. Die photodynamische Wirkung des Chlorophylls und ihre Beziehung zur photosynthetischen Assimilation der Pflanzen. Jahrb. f. wiss. Botanik., 46: 599.' HAUSMANN, W., 1910. Die sensibilisierende Wirkung des Hamatoporphyrins. Biochem. Zcitschr., 30: 276. JODLBAUER, A., 1926. Die physiologischen Wirkungen des Lichtes. Handbuch. dcr norm. it. path. Physio!., 17: 305. JODLBAUER, A., AND TAPPEINER, H., 1905. Die Beteiligung des Sauerstoffs bei der Wirkung fluorescierender Stoffe. Dcntschcs Arch. f. klin. Med., 82: 520. LEDOUX-LEBARD, 1902. Action de la Lumiere sur la To.xicite de 1'Eosine. Ann. lust. Pasteur., 16: 587. METZNER, P., 1924. Zur Kenntnis der photodynamischen Erscheinung. III. Biochem. Zcitschr., 148: 498. MOORE, A. R., 1928. Photodynamic Effects of Eosine on the Eggs of the Sea L'rchin, Strongylocentrotus purpuratus. Arch. di. Sci. BioL. 12: 231. NOACK, K., 1920. Untersuchungen iiber lichtkatalytische Vorgange. Zcitschr. f. Botanik, 12: 273. PEREIRA, J. R., 1925. On the Combined Toxic Action of Light and Eosin. Jour. E.vper. Zool, 42: 257. PRINGSHEIN, P., 1928. Fluorescenz u. Phosporescenz. Berlin. RAAB,' O., 1900. Ueber die Wirkung fluorescirender Stoffe auf Infusorien. Zcitschr. f. BioL, 39: 524. SACHAROFF, G., AND SACHS, H., 1905. Ueber die hamolytische Wirkung der photodynamischen Stoffe. Miiiich. Med. Woch., 52: 297. SCHANZ, F., 1921. Die physikalischen Vorgange bei der optischen Sensibilisation. Pfliiger's Arch., 190: 311. SCHMIDT, C. L. A., AND NORMAX, G. F., 1922. Further Studies on Eosin Hemolysis. Jour. Gen. Physiol.. 4: 681. STRAUB, W.. 1904<7. Ueber chemische Vorgange bei der Einwirkung voti Licht auf fluoreszierende Substanzen. Miinch. Med. Woch.. 51: 1093. STRAUB, W., 1904??. Uber den Chemismus der Wirkung belichteter Eosinlosung auf oxydable Substanzen. Arch, e.vper. Path. u. PhannakoL, 51: 383. TAPPEIXER, H., 1908. Untersuchungen u'ber den Angriffsort der fluorescierenden Substanzen auf rote Blutkorperchen. Biochem. Zcitschr., 13: 1. TAPPEINER, H., 1909. Die photodynamische Erscheinung. Ergebnisse der Physi- ologic, 8: 698. THE EQUILIBRIUM OF OXYGEN WITH THE HEMOCY- ANIN OF LIMULUS POLYPHEMUS DETERMINED BY A SPECTROPHOTOMETRIC METHOD ALFRED C. REDFIELD (From the Department of Physiology. Harvard Medical School, Boston, and the Marine Biological Laboratory, Woods Hole) The respiratory proteins, including hemoglobin, hemocyanin, chlo- rocrurin and hemerythryn, are unique in combining with and dissociat- ing from oxygen at pressures which fit them for the physiological trans- portation of this gas. The factors which determine the condition of equilibrium between oxygen and the pigment are of interest not only because of the evident physiological relationship between the charac- teristics of the oxygen dissociation curves of the blood of various or- ganisms and the pressures of oxygen in the environment, but because of the interesting physico-chemical problem which the phenomena pre- sent. The hemocyanins appear to possess certain advantages for the study of these problems. Not only do these proteins exist naturally in solution in the blood so that the complications which arise from dealing with corpuscles are avoided, but they are relatively stable compounds which lend themselves without difficulty to purification and preserva- tion. The hemocyanins of different species appear, in addition, to ex- hibit very considerable differences in their physical and chemical prop- erties; and consequently, one has the advantage in their study of being able to resort to the comparative method in testing generalizations. Finally, from the technical point of view, the hemocyanins which are essentially colorless when reduced become strongly colored in the oxy- genated state and consequently lend themselves to the employment of colorimetric methods for the determination of the degree of oxygenation of the solutions. The present paper contains an account of a spectrophotometric method for the determination of the degree of oxygenation of hemo- cyanin solutions. The method is applied to an examination of the equilibrium between oxygen and a purified salt-free preparation of the hemocyanin of the horse-shoe crab, Linmlns [>oly[>licuu, is given by the expression p = (B - P - aq) /, where B is the barometric pressure, P is the pressure in the tonometer recorded at the end of equilibration, aq is the tension of aqueous vapor at the temperature of the water bath, and / is the fraction of oxygen in the gas mixture. To obtain complete oxygenation, the tonometer is simply evacuated and filled with pure oxygen gas prior to equilibration. To obtain com- plete reduction is difficult under those circumstances in which the affinity of hemocyanin for oxygen is great. We have not found the employ- ment of chemical reducing agents satisfactory, as certain of these tend to influence the color of the solution and others must be employed in such concentrations that they may affect the scattering of light on which the absorption by the reduced solution depends. The most satisfactory procedure is to employ hydrogen to wash out the tonometer after the oxygen is freed from the hemocyanin under low pressure. The solu- tions are accordingly evacuated, equilibrated for twenty minutes, filled with hydrogen, re-evacuated, again equilibrated, allowed to settle, and then measured. Further repetition of the process does not lead to lower readings, although it is doubtful whether, under certain circumstances, this process removes the last traces of oxy hemocyanin. The reason for this is that, in the process of evacuation and equilibration, small quan- tities of denatured material are formed which fail to settle out com- pletely when the solutions are allowed to stand. The formation of pre- cipitates of this sort, which goes on more readily in the reduced solu- tions, constitutes the principal limit to the precision of the method. We have recently constructed tonometers which can be placed in the cups of a large centrifuge and which make it possible to remove these troublesome precipitates. Such tonometers have not been employed in the experiments described in this paper. The Preparation of the Hemocyanin Solutions. — The hemocyanin employed in the present investigation was prepared from material ob- tained during the summer of 1928. It was preserved in the precipitated state by adding 350 grams of ammonium sulphate to each liter of serum. The material was purified some months later by repeated salting out 246 ALFRED C. REDFIELD followed by dialysis, against dilute sodium bydroxide, as described by Redfield, Coolidge and Sbotts (1928). Tbree preparations were ob- tained, having' the following characteristics : Specimen 18 A, dry weight 0.1031 gram per cc., combined base 19.4 X 10~5 mols per gram; Speci- men 18 B, dry weight 0.1255 gram per cc., copper 0.0208 milligram per cc. or 0.168 gram per 100 grams dry substance, combined base 19.1 ; ' 10 5 mols per gram; Specimen 18 C, dry weight 0.097 gram per cc.. copper 0.16 milligram per cc. or 0.165 gram per 100 grams dry weight, combined base 21.6 X 10^5 mols per gram. These solutions were preserved with toluene at a low temperature. The day before measurements were to be made, they were further diluted by the addition of distilled water containing amounts of hydrochloric acid or sodium hydroxide appropriate to secure the desired hydrogen ion activity and to reduce the hemocyanin to a concentration favorable for the measure- ments; that is, to about 2.5 per cent. After standing all night, the solutions were filtered and then employed for the determination of the oxygen dissociation curves. A portion of the solution was also reduced by equilibration with hydrogen and used for the determination of the hydrogen ion concentration by means of the hydrogen electrode.1 Measurements were made upon solutions at several hydrogen ion activities between pH 7.4 and pH 10.4, and upon a solution at pH 4.5. At hydrogen ion activities intermediate between pH 4.5 and about pH 6.8, Liundiis hemocyanin is insoluble in distilled water, and solu- tions of sufficient clarity cannot be obtained. At reactions more acid than pH 4.5, a colorless modification of L'unulus hemocyanin is formed (Redfield and Mason, 1928). The characteristics of the oxygen dis- sociation curve at these hydrogen ion activities will be dealt with in a subsequent paper.2 DATA ON OXYGEN DISSOCIATION The results of the series of measurements which have been made upon purified solutions of Limulus hemocyanin are recorded in Table III. The first column contains a description of the material employed in each case ; the second column, the partial pressure of oxygen in the tonometer at the completion of equilibration ; the third column, the value of the extinction coefficient of the solution (2/d log tan a,/), as meas- ured with the spectrophotometer, employing light of the wave-length 1 In the case of the two solutions prepared from Specimen 18 C, the pH value was somewhat less than that to be expected from the amount of NaOH .added, as judged from the titration curve published by Redfield, Humphreys and Ingalls (1929). In the other solutions the agreement is good. 2 I am indebted to Miss Elizabeth Ingalls for technical assistance in con- ducting the experiments and for preparing the hemocyanin solutions employed. EQUILIBRIUM OF OXYGEN WITH HEMOCYANIN 247 590 m/j.. This measurement is not corrected for absorption by the solvent. The fourth column records the extinction coefficient of the oxygenated chromatic groups, 2/W(log tan a,, --log tan a-,-) ; the fifth column, the value of v as defined in equation (7). The equilibration was carried out in a water bath at 25° C. The length of the T-tube of the tonometer in which the absorption of light was measured, d, was usually 3.3 centimeters. In the case of a few measurements, tubes were employed which differed slightly from this length (3.15 to 3.60 cm.). THEORY OF OXYGEN EQUILIBRIUM In oxyhemocyanin, one molecule of oxygen is bound by a quantity of hemocyanin containing two atoms of copper. The reversible reac- tion may consequently be indicated by the equation where n represents the number of mols of oxygen bound by each mol of hemocyanin. In treating the equilibrium according to the mass law, as was done by Hiifner (1901) and later by Hill (1910), in the case of hemoglobin, the result is in which k is the equilibrium constant of the reaction. If v is the fraction of hemocyanin in the oxygenated condition, 1 - -y is the reduced fraction and, putting p, the partial pressure of oxygen in mm. of mer- cury, in place of the oxygen concentration, equation (8) may be written (9) 1-v or, log -— = log K + n log p. (10) In this form the equation is convenient for graphical solution for n and K. In Fig. 2 is reproduced the data recorded in Table I, arranged in the form indicated by equation (10). The lines drawn through the points in each case are straight lines indicating the linear relationship demanded by the equation. The slope of the lines drawn through the points, determining the value of n, is 1.0. The values of K correspond- ing to the positions of the lines drawn in Fig. 2 are indicated in Table III. Employing these values of K and taking ;/ as equal to 1.0 in each case, the values of v may be calculated and are indicated in column 6 of 248 ALFRED C. REDFIELD Table III for comparison with the observed values. It appears that the theoretical treatment from which equation (10) is derived is ade- quate to account for the shape of the oxygen dissociation curve at least pn 743 pn pn gsi 109 /i FIG. 2. Logarithmic plot of data of oxygen dissociation curve of hemocyanin of Ln;n, oxygen pressure in mm. Hg. as a first approximation, and to provide a single series of constants to define the effect of hydrogen ion activity upon the equilibrium.3 Careful scrutiny of the data in Table III reveals a tendency for the low values of v to be slightly greater than the calculated values and high values to be slightly less than the theoretical. In order to make vivid the adequacy of the theory for treating the entire set of observations, in Fig. 3 the values of y obtained at each pH value are plotted against Kf> in the usual form of the oxygen dissociation curve, and a line corre- sponding to the theoretical treatment is drawn through the points, « again being taken as 1.0. 3 It should be emphasized that the pH values are determined on reduced solu- tions. No account has been taken of possible change in pH with oxygenation. According to Redfield, Humphreys and Ingalls (1929), the effect may be expected to be small. EQUILIBRIUM OF OXYGEN WITH HEMOCYANIN 249 TABLE III Data of Oxygen Dissociation Curves of Limn! us Hemocyanin. Temperature, 25° C.; Wave-length, 590 HIM- Description P 2 -7 log tan ay - loe tan a" y y d '°g tan ar mm. Hg (observed) (calculated) Specimen 18 A 0 0.034 0 0 0 0.16 0.068 0.034 0.117 0.074 Concentration: 0.56 0.106 0.072 0.248 0.218 0.0258 grams per 0.80 0.133 0.099 0.342 0.286 cc. 1.24 0.149 0.115 0.397 0.382 Combined acid : 1.65 0.166 0.132 0.455 0.452 20 X 10~5 mols 2.34 0.194 0.160 0.552 0.540 per gram 2.74 0.204 0.170 0.586 0.578 pH 4.52 3.92 0.225 0.191 0.659 0.662 6.20 0.252 0.218 0.752 0.757 K = 0.500 7.28 0.259 0.225 0.776 0.784 8.65 0.269 0.235 0.810 0.812 11.9 0.281 0.247 0.852 0.856 14.8 0.283 0.249 0.859 0.881 17.6 0.291 0.257 0.886 0.898 20.6 0.298 0.264 0.910 0.913 24.5 0.299 0.265 0.914 0.924 27.4 0.299 0.265 0.914 0.933 39.0 0.305 0.271 0.935 0.952 744 0.324 0.290 1.00 1.00 Specimen 18 A 0 0.034 0 0 0 0.39 0.092 0.058 (0.265) 0.171 Concentration: 0.80 0.119 0.085 0.305 0.276 0.0258 grams per 1.49 0.158 0.124 0.441 0.415 cc. 2.61 0.188 0.154 0.548 0.554 Combined base: 3.74 0.207 0.173 0.616 0.640 19 X 10~5 mols 4.78 0.223 0.189 0.672 0.695 per gram 5.96 0.239 0.205 0.730 0.739 pH 7.43 6.83 0.253 0.219 0.780 0.766 8.38 0.258 0.224 0.797 0.800 £.' = 0.476 16.6 0.282 0.248 0.882 0.888 26.4 0.300 0.266 0.946 0.927 37.7 0.297 0.263 0.936 0.948 152 0.310 0.276 0.982 0.987 740 0.315 0.281 1.00 1.00 250 ALFRED C. REDFIELD TABLE III (continued) Data of Oxygen Dissociation Curves of Limulns Hemocyanin. Temperature, 25° C. ; Wave-length, 590 m/i. Description P -j log tan av 2 10B tan ^ y y d tan ar mm. Hg (observed) (calculated) Specimen 18 B 0 0.046 0 0 0 0.84 0.078 0.032 0.166 0.213 Concentration: 1.22 0.103 0.057 0.295 0.282 0.0208 grams per 1.98 0.123 0.077 0.399 0.390 cc. 3.15 0.150 0.104 0.539 0.504 Combined base: 4.40 0.158 0.112 0.580 0.587 39 X 10~5 mols 9.30 0.185 0.139 0.720 0.750 per gram 12.80 0.201 0.155 0.803 0.805 pH 8.51 17.7 0.204 0.158 0.818 0.852 28.1 0.217 0.171 0.886 0.900 K = 0.322 39.0 0.219 0.173 0.902 0.927 751 0.239 0.193 1.00 1.00 Specimen 18 C 0 0.047 0 0 0 0.63 0.082 0.035 0.155 0.177 Concentration: 0.97 0.093 0.046 0.204 0.217 0.0242 grams per 1.64 0.128 0.081 0.358 0.319 cc. 4.12 0.167 0.120 0.531 0.541 Combined base: 6.38 0.197 0.150 0.654 0.646 63 X lO"5 mols 8.94 0.209 0.162 0.717 0.719 per gram 11.5 0.216 0.169 0.748 0.767 pH 9.71 16.8 0.230 0.183 0.810 0.828 21.8 0.228 0.181 (0.801) 0.862 K = 0.286 26.7 0.250 0.203 0.898 0.885 724 0.273 0.226 1.00 1.00 Specimen 18 C 0 0.066 0 0 0 0.61 0.084 0.018 0.092 0.098 Concentration : 1.10 0.088 0.022 0.112 0.164 0.0242 grams per 1.79 0.124 0.058 0.296 0.242 cc. 4.15 0.155 0.089 0.454 0.426 Combined base: 6.53 0.177 0.111 0.566 0.538 77 X 10~5 mols 9.04 0.183 0.117 0.597 0.618 per gram 12.2 0.202 0.136 0.694 0.686 pH 10.42 17.1 0.242 0.176 (0.898) 0.754 22.6 0.225 0.159 0.812 0.802 K = 0.178 27.2 0.224 0.158 0.806 0.830 751 0.262 0.196 1.00 1.00 EQUILIBRIUM OF OXYGEN WITH HEMOCYANIN 251 DISCUSSION In the forty years since Hiifner suggested the application of the mass law to the equilibrium between oxygen and hemoglobin, numerous investigations have indicated that equations similar to those employed in the present treatment are more or less adequate to describe the data in hemoglobin solutions free of electrolytes (Barcroft, 192cS). Un- certainty has sometimes accompanied the results of such investigations because of the instability of purified hemoglobin solutions (Ferry, 1924; Hecht, Morgan and Forbes cited by Barcroft, 1928) . In the presence of electrolytes and in blood, the dissociation curves of hemoglobin in- variably have a sigmoid shape, requiring some additional assumptions for their explanation. /o y ,y»< t • -pM -ass K-OSOO Q-ptl 743 K-O475 &-f>H B3/ K-G3SS + -pn 3 71 K-OSB6 y.-prt ID terns consists, insofar as the description of the living and fixed egg is concerned, merely in a confirmation of the observations of Lillie (1906). In addition, the egg has been studied by means of the cardioid condenser, which enables the investigator to trace the cor- tical, or as it is called in the above paper, the ectoplasmic portion of the egg. and to distinguish it from the endoplasmic portion during the maturation stages. The observations are entirely in accord with those cited above, but will be described here in order that they may be re- ferred to in the discussion. FIG. 2. Photograph of a section of a Cluctoptcnts egg at the mesophase of the first maturation division, showing the " ectoplasmic defect " and the arrange- ment of adjacent parts. FIG. 3. Sketch of an egg of Chcctoptcnis at the mesophase of the first matura- tion division as it appears by dark field illumination with the cardioid condenser to show the " ectoplasmic defect " as the polar elevation is forming. As is well known, the egg of Ch(ctoptcnts is penetrated by tin spermatozoon before the formation of the polocytes. Thus activated, the cortical or ectoplasmic portion withdraws from the animal pole of the egg, and the spindle of the first oocytic division approaches the cell membrane remaining surrounded by the endoplasmic cytoplasm. It becomes attached to the outer zone in the region of the ectoplasmic 260 LEIGH HOADLEY defect and there forms the first polar body. This has been described by Lillie (cf. Fig. 2) and may be seen in living eggs by means of the cardioid condenser (Fig. 3). It is composed entirely of endoplasmic substance without deutoplasm. The polar body must, therefore, repre- sent ground substance plus some of the residual substance of the germinal vesicle. In addition there is nuclear chromatin from the first maturation division of the egg nucleus. The polocyte is entirely re- leased from the egg, which immediately begins the second maturation division, the ectoplasmic defect remaining till this is complete. During this time the first polar body does not divide, as is the case in Loligo, but remains inactive adjacent to the outer membrane covering the egg. As in the case of the first polar body, the second contains none of the cortical ectoplasmic material but only endoplasmic cytoplasm and nu- cleus. The course as described here may be traced with ease by use of the cardioid dark field apparatus, as the egg is relatively small and there is an evident optical difference in the appearance of the cortical ectoplasmic and the deeper endoplasmic layers. Associated with this lack of the cortical layer is the fact that neither of the polocytes of Chcctoptcrus divide as is the case of those produced by the egg of Loligo. Discussion As is well known, the relation between the formation of the polar bodies and the penetration of the spermatozoon varies in different groups of animals. In some forms the spermatozoon penetrates the egg before either of the polar bodies is formed. In other cases the first polar body is completed before penetration, but the second is de- pendent upon the activation of the egg at fertilization. In still another group the second polar body is also formed before penetration. It is evident that conditions controlling the development of the polar body must vary according to the conditions existing in different animals. In the case of the first group, the polocyte is a part of the egg at the time of the activation and therefore must initially be activated in the same way. In the second group the second polar body must be activated while the first is not. The first, in order to simulate the conditions existing in the polar body of the first group, must be penetrated by a sperm cell or be activated in some other fashion. In this group the second polar body might conceivably develop save for the fact that the cell organs introduced by the sperm are not present (see below). In the third group of eggs, neither of the polar bodies is a part of the egg at the time of sperm penetration and hence has fulfilled none of the conditions attendant on its activation. In order to show any de- POLOCYTE FORMATION AND CLEAVAGE OF POLAR BODY 261 velopment, therefore, it would have to be activated. In addition, as has already been mentioned by Conklin (1915) both of the polar bodies of the first group would be protected against further activation by sperm because of cortical modifications attendant upon the previous insemination of the egg. The same would be true of the second polar body in the second group, In order to explain the lack of development in these, it would be necessary to demonstrate either a consistent lack of some significant part of the nuclear or cytoplasmic portion of the entering spermatozoon or of the egg at the time of polocyte formation. In the same way, development of the second polocyte of the second group and both of the polocytes of the third group must be dependent on subsequent activation. The last statement is dependent on another condition which would seem in the light of the present observations to be of primary importance. The polocyte must contain all of the constituents of the egg essential both to its own activation, and to the activation of the spermatozoon. Unfortunately, there is little information as to the actual constitution of the cytoplasmic portion of the polar body available in literature on the subject. These structures have been considered as aborted ova by Mark in his monograph on Liina.v coinpcstris (1881 ). In that form the polar bodies do not develop. It would now be of great interest to know the actual constitution of the polocyte. In isolated cases the polar bodies have been observed to develop to some extent at least. Lefevre (1907) cites the case of the polar body of Thalasscina mclllta, which is formed in response to the acid activation (artificial parthenogenesis) of the egg. Subsequent cleavage of the extruded cell gives rise to a ' morula,' which later dies. In this case also there is some confusion in that the actual composition of the polocytes which develop is not known. In no case did they develop beyond this early stage, however, though that may conceivably be due to quantitative deficiencies which do not immediately concern us here. The polar nuclei such as are formed in some insects and Crustacea do not come under the realm of our discussion, but. inasmuch as they sometimes cleave, 'we should mention them. The polar nuclei may fuse and in some cases they divide to form what appear to be accessory em- bryonic structures. This is the case in the hymenopter, Litomastix (Silvestri, 1908). The most extensive development of the polocyte is to be found recorded in the paper of Francotte (1898) on the polyclads. in which he records the formation of the gastrula by fertilized first polar bodies of ProstheccrcEus vittatits. The argument is slightly different from that in other cases, though it will concern us in a moment. We shall 262 LEIGH HOADLEY therefore consider it here. The egg of this form produces a polar body at the first maturation division, which may be as much as one- fourth as large as the remaining portion of the egg. Subsequently this may be fertilized, as is the egg itself, and then both of the units give rise to one polar body and continue in their development. Fran- cotte described the development of these forms only as far as the gastrula. It may be assumed, however, that all of the conditions for cell division and development of any other than a quantitative nature are fulfilled. We shall return to a consideration of this case in con- nection with the fate of the so-called giant polar bodies of other forms. In connection with the impregnation of the polar bodies after their extrusion, mention should be made of certain observations of Fol. This investigator has reported sperm penetration of polocytes in echino- derms. No mention is made of the reaction of the sperm or of the egg, so that it is not known as to whether the conditions of activation, as expressed morphologically, are complete or not. It is conceivable that in this case the very small size of the polar body would preclude its further development. It is also possible that the polar body of the echinoderm egg does not include all of the parts of the egg requisite to development. It is in this last statement that we find a possible explanation of the difference between the behavior of the polar bodies in Loligo and Chcc- toptents. It can be demonstrated that in the formation of the polar bodies in Chcctoptcrns, while the majority of the deutoplasm remains within the egg and the polocyte consists mainly of hyaloplasm, only the deeper endoplasmic portion of the cytoplasm is extruded, and the polocyte does not contain all of the zones characteristic of the egg which develops. It might be emphasized here that hyaloplasm and ectoplasm are not identical. In complete accord with this lack of ectoplasm, there is no second division of the primary polar body. In Loligo, on the other hand, the extruded polar body contains not only the endoplasmic region, but also some of the outer cortical part of the cytoplasm. In this case the polocytes develop for a little while. A striking; difference between the two types of polar body is to be found in the presence or absence of the cortical ectoplasmic portion, which may be assumed, therefore, to play an important role in future events. Certain observations made by Conklin (1915, 1917) on the behavior of artificial giant polocytes in Crepidula are very important in the con- sideration of the phenomenon. In this paper, Conklin changes his previous views on the subject and states that in such cases the lack of the sperm aster may be the important factor in the non-development of the large polar body. These were obtained by centrifuging the eggs POLOCYTE FORMATION AND CLEAVAGE OF POLAR BODY 263 during the formation of the polar body. In those cases in which the rotation of the egg was such that the spindle was in the line of the gravitational pull and at the outer end, huge polar bodies were formed which appeared to contain all of the egg substances and yet did not develop. Following certain previous conceptions. Conklin differentiates between those phenomena of fertilization leading to activation of the cytoplasm and to the development of the egg. The polocyte. having been activated with the rest of the egg by insemination, forms, but. lacking the aster which he considers as the part essential to further de- velopment, fails to divide. In view of the experiments of Lillie (1906) on Chcetopterus eggs during maturation phases it would appear that there may be a question as to the composition of the polar body. It" one is to examine Fig. 24 (p. 185) in Lillie's report, one sees that when the Chcetopterus egg is centrifuged, certain substances of the egg are more or less free and that one of these is the endoplasmic material. As a result, the endoplasmic material takes its position according to the force of gravity. The ectoplasmic cortical material, on the other hand, is fixed, not being displaced by the centrifugal force. If, then, the force of gravity were applied in such a way that the endoplasmic material were all pushed against the region of the ectoplasmic defect or pole at which the spindle is attached, the result would be that it would bulge outside of the ectoplasmic portion at this point and that, when polar body formation was completed, a great quantity of this material would be separated from the egg and follow the polar body. This would not necessarily involve the inclusion of any of the ecto- plasmic material in its formation. If this should prove to be the case, the non-development of the polar body of Crcpidula, when present as the giant polar body, might be due to the same factors which seem at present to be responsible for the non-development of the pnlocyte of Chcetopterus. In this connection it is of interest to note that the first polocyte of Crcpidnla may divide once by mitosis and subsequently several times by amitosis (Conklin, 1902, page 21, Fig. 41, etc.). This does not occur in the second polocyte. There may be a discrepancy in their constitution or in the quantitative relationships between the amounts of substances of cytoplasmic nature present. In this way the first polocyte may resemble that of Loligo, while the second may be like that of Chcetopterus. In any event, the conditions present in Cluctop- tcnis do not obtain for Loligo and the results differ accordingly. The conception of the great importance of the cortical portion of the egg to the phenomena of early development is not new as it has been stressed in a number of places by several investigators, notably by Chambers (1921) and Just (1923). These authors show by their experiments 264 LEIGH HOADLEY that there is a real distinction morphologically and functionally between cortex and endoplasm. In conclusion I would like to suggest that in Loligo and Chcetopterus and possibly in other forms, the development of the polar body is de- pendent on the presence of cortical ectoplasmic material in that organ or at least on the ratio between the amount of the ectoplasmic substance and the remaining cytoplasmic and nuclear material. LITERATURE CITED CHAMBERS, ROBERT, 1921. Studies on the Organization of the Starfish Egg. Jour. Gen. Physiol., 4: 41. CONKLIX, E. G., 1902. Karyokinesis and Cytokinesis in the Maturation, Fertili- zation, and Cleavage of Crepidula and other Gastropods. Jour. Phila. Acad. Nat. Sci, 12: part 1. CONKLIX, E. G., 1915. Why Polar Bodies do not Develop. Proc. Nat. Acad. Scl., 1: 491. CONKLIX, E. G., 1917. Effects of Centrifugal Force on the Structure and De- velopment of the Eggs of Crepidula. Jour. Ex per. Zodl., 22: 311. FRANCOTTE, P., 1898. Recherches sur la maturation, la fecondation et la seg- mentation chez les Polyclades. Arch. d. Zodl., 6: 189. JUST, E. E., 1923. The Fertilization-Reaction in Echinarachnius parma. VI. The Necessity of the Egg Cortex for Fertilization. Biol. Bull., 44: 1. LEFEVRE, G., 1907. Artificial Parthenogenesis in Thalassema Mellita. Jour. Ex- per. Zodl., 4: 91. LILLIE, F. R., 1906. Observations and Experiments Concerning the Elementary Phenomena of Embryonic Development in Chaetopterus. Jour. Exper. Zodl., 3: 153. MARK, E. L., 1881. Maturation, Fecundation, and Segmentation of Lima.v cam- pcstris Binney. Bull. Mus. Compar. Zodl., Cambridge, Mass., 6: 173. WATASE, S., 1891. Studies on Cephalopods. I. Cleavage of the Ovum. Jour. .Mor ph., 4: 247. THE DISTRIBUTION OF PIGMENT AND OTHER MORPHO- LOGICAL CONCOMITANTS OF THE METABOLIC GRADIENT IN OLIGOCH^TS GRACE EVELYN PICKFORD OSBORN ZOOLOGICAL LABORATORY, YALE UNIVERSITY INTRODUCTION The form of the antero-posterior metabolic gradient of the Oligo- cluets has now been well established by many workers and by almost as many different methods. Hyman (1916) investigated the gradient of susceptibility to KCN in many lower Oligochaets and distinguished two types, a primary gradient found only in the primitive /Eolosomatidse and young zooids of the Naididae in which the susceptibility decreased progressively from head to tail, and a widely distributed secondary type in which the susceptibility again rose at the posterior end. In Lumbriculus rarians measurements of the oxygen intake of different regions of the body by the Winkler technique (Hyman and Galigher, 1921) showed a secondary metabolic gradient; the accurate manometric determinations of Shearer ( 1924) on the " earthworm " confirm the primary but throw no light on the secondary gradient, since the experi- ments were only made on head and tail portions. The early work of Morgan and Dimon (1904) on the potential gradient showed that in Lituibn'cns tcrrestris and Allolobophora foetida the head and tail were electronegative to the middle region, while Moore and Kellogg (1914) found that in an electric field Lumbricus oriented itself in the form of a U with head and tail towards the cathode. Hyman and Bellamy (1922) confirmed these results and correlated them with the metabolic gradient. Hatai ( 1924) showed that in two Japanese species of Phcrc- t'una (incorrectly named Perich&ta) the amount of heat required to produce initial heat rigor in the muscles of the body wall was greatest at the anterior and posterior ends and least in the middle region of the body. He correlated these results with the percentage water content of the body wall, which is inversely proportional to the temperature required to produce initial heat rigor. Watanabe (1928) found that in P. coiniininissiiiia the potential gradient is on the average of the secondary type, although dorsally it is perhaps of the primary type. Recently Perkins (1929) has published a short note in which he 265 266 GRACE EVELYN PICKFORD claims that in earthworms the gradient of extractable reduced sulphydryl reaches a maximum in the mid anterior region of the body. Perkins summarizes his results as follows, " In earthworms I find that the gradient of growth corresponds with the gradients of total iodine equivalence, extractable sulphydryl, and total sulphur (gravimetric) and not with the gradient of total metabolism observed by the oxygen uptake ; the last, therefore, includes other oxidation systems which it is legitimate to suppose result in katabolism rather than the anabolism of growth. It is interesting to find that gradients in the earthworm have a summit at about that point whence a divided worm grows for- wards or backwards according to the aspect of the cut surface." x As regards the dorso-ventral gradient very little work has been done, although Hatai (1924) states that the temperature necessary to produce initial heat rigor is greater for dorsal than for ventral and intermediate for lateral portions of the body wall. Little attention has been paid to the morphological concomitants of the metabolic gradient. Hess (1924) showed that the sensitivity of Linnbrictts terrcstris to light is greatest at the anterior end and least in the mid region of the body and that except on the first five and last two segments it is confined to the dorso-lateral regions ; he also noticed that the distribution of pigment corresponds rather closely to the light sensitivity. In a later paper Hess (1925) showed that the distribution of the photo-receptor organs coincides with the distribution of the photo- sensitive regions, thus putting the gradient of sensitivity to light on a morphological basis. Nomura (1926) has extended the work of Hess, showing that in the ventral nerve cord of Allolobophora foctlda Sav. there is an axially graded distribution of photic response ; negative orientation, which also characterizes the brain, increasing posteriorlv and positive orientation anteriorly, while the supposed neurones causing backward crawling are apparently restricted to the anterior end op- posing the brain, which controls forward crawling. DISTRIBUTION OF PIGMENT Many species of earthworm are pallid and others may be colored green, blackish, or yellow by as yet uninvestigated pigments, but by far the most commonly occurring coloration is due to a reddish or purplish-brown pigment which has been shown in some species (see Kobayashi, 1928) to be a porphyrin allied to some derived from chloro- phyl. This reddish pigment is characteristically distributed on the dorsal side and is most intense at the anterior end. A typical case can be found in the well-known species Luinbricus terrcstris Linn. Indi- : References to text figure omitted. METABOLIC GRADIENT IN OLIGOCH^TS 267 vicluals of this species will be found to vary somewhat in the intensity and exact extent of pigmentation, but the following description taken from a specimen recently caught near this laboratory will serve as an example: "Intensely pigmented dorsally at the anterior end, the pig- mentation extending laterally to about cd (the line of the lateral setse), the first three segments also slightly pigmented ventrally; posterior to the clitellum the lateral extent of the dorsal pigmentation becomes re- duced until only a mid-dorsal line is left which persists throughout the posterior half of the body; at the extreme posterior end there is again an increase in intensity and extent of pigmentation (except on the terminal segment which is small and pale) which extends laterally to below the setal line cd on the seventh to the second last segments and even faintly on the ventral side of the second, third, and fourth seg- ments from the end." In this case the distribution of pigment follows the secondary type of gradient, and it may be said in general that when- ever a species of earthworm exhibits this red-brown pigmentation (pre- sumed, but of course not proved in most genera to be due to a porphyrin allied to that of Lnmbricus and Allolobophora (Eiscnia) foctida Sav.. it will be distributed according to the primary if not the secondary type of gradient. Hatai (1924) noticed that in " Perichata " niegas- colidioides Goto et Hatai the dorsal side was more pigmented than the ventral, though curiously enough he did not correlate this with the dorso-ventral gradient. My own investigations have so far been con- fined to a systematic examination of South African species of the genera Chilota and Acanthodrilus. In these genera every gradation from total pallor to intense pigmentation can be found ; some of the most interesting cases are those species, or varieties of otherwise pallid or pigmented species, in which pigmentation is only found on the first or last few segments. For example, the Cape Flats species Acantho- drilns antiidinis Bedd. is pigmented dorsally on the first and last four or five segments but more intensely on the latter, while in many unde- scribed species of Cliilota only a few of the anterior segments are pig- mented. When the dorsal pigmentation is intense and occurs along the whole length of the body it is usual to find that the first five to ten segments are deeply pigmented ventrally, while in many cases pigment is deposited on the thickened septae and generally on the inner side of the body wall at the anterior end. A more complete discussion will be given in my forthcoming paper on the South African Acanthodrilinae. The distribution of pigment in Oligochsets may be compared with that described by Faris (1924) for Amblystoma embryos. In this case the pigment is apparently a melanin and is deposited in regions of tissue differentiation as opposed to regions of proliferation. If the intensity 18 268 GRACE EVELYN PICKFORD of pigmentation in Oligochcets is really a function of the metabolic rate, it seems possible that highly pigmented species would have a higher oxygen intake than pallid ones. It is hoped to investigate this point shortly on a large number of species. If this view is correct, and it is supported by the fact that pallid species are more sluggish in their movements and less irritable to handling than pigmented ones (com- pare Allolobophora (Eisenia) rosea Sav. with species of Luinbricns), it would seem unlikely that the porphyrin is merely derived from the food of the worm, as has been suggested, and more probable that it is a breakdown product of the worm's own haemoglobin. MULTIPLICATION AND REDUCTION OF SET.E As regards the more specifically morphological concomitants of the axial gradient, certain stages in the reduction and multiplication of setal numbers are significant. In the primitive lumbricine condition there are two pairs of setae per segment except on the first, which never has setae; a reduction in numbers sometimes takes place as in species of the Microchcutiis bciihaini group where setse are absent on the first six or seven segments of the adult (frequently only the lateral pair are absent on segment 6). This trend to reduction finds an extreme case in Tritoyciiia crassa Mchlsn., in which only the ventral setae of the clitellar region persist. In the Enchytrseidae parallel cases can be found; in the genus Distichopus only ventral setae are present. In the genus Michaelsena transitional species occur from M. inangcri Mchlsn., in which dorsal and ventral setae are present throughout and M. principissce Mchlsn., in which the ventral setae commence on segment 3 and the dorsal on segment 14, to M. iionnani Mchlsn., which has ventral setae from segment 3 onwards but dorsal setae only on segments 4-6, and M. subtilis Ude., in which dorsal setae are absent and ventral setae occur only on segments 4-6. In the genus Achccia setae are totally absent. These cases may be compared with the phenomenon of cephalization in the Naididse (Stephenson, 1912 and 1923), in which certain anterior segments are devoid of dorsal (i.e. lateral) setse. Hyman (1916) found a very peculiar gradient in the Naid Chcetogastcr diaphamis, in which the susceptibility was least at the head end. In this genus dorsal setae are totally absent and ventral setae though present on seg- ment 2 are absent on segments 3, 4, and 5. The tendency to setal multiplication is a very widely distributed phenomenon, and the perichaetine condition has apparently arisen in- dependently many times in various families of the terrestrial or Neo- Oligochaets (see Stephenson 1921, 1923 for a discussion of this and METABOLIC GRADIENT IN OLIGOCH^ETS 269 other trends in the evolution of the Indian Oligochaeta). The multi- plication of setae varies from a condition in which six or eight pairs occur instead of four per segment to the purely perichaetine condition in which each segment has a complete ring, but the most interesting cases are those in which a transitional condition exists. In Mcgascolc.\~ zvilleyi Mchlsn. there are eight setae per segment at the anterior end and twelve in the middle and posterior regions; in M. vilpattiensis Mchlsn. there are eight setae in four pairs on segments 2 and 3, eight or nine on segment 4, circa 11 on segment 13, circa 24 on segment 26, and circa 26 at the posterior end. In general in transitional species the smaller number and 1 or more primitive paired condition persist at the an- terior end. Sufficient data are unfortunately not available as to the extreme posterior end, so that it is not possible to state whether the smaller number also persists there in these intermediate forms. Hatai (1924) has investigated the setal numbers in the purely perichaetine species " Pe ricliccta " (Plicrctiina) inegascolidioidcs Goto et Hatai. He finds that the number of segments is extremely constant and bears no relation to the size of the worm and that the total number of setae per worm does not vary very greatly. The number of setae per segment increases from segment 2-25, remains about constant up to segment 100 and then decreases again, thus exhibiting a curve comparable with the secondary type of oligochaet gradient. From a survey of the avail- able data it would thus seem as if setal multiplication were correlated with a lower and setal reduction with a higher metabolic rate. The case of Acantlwbdclla pclcdina Grube, an aberrant parasitic form re- garded until recently as a leech, must not be overlooked, although the evidence (c.f. Clicctogastcr) cannot be interpreted until the form of the metabolic gradient has been investigated. In this species setae are present only ventrally on the first five segments. MULTIPLICATION AND REDUCTION OF NEPHRTDIA The trend to setal multiplication is paralleled and usually accom- panied by the multiplication of the nephridia, primitively one pair per segment. Unfortunately the whole subject of nephridial multiplication stands in need of a thorough revision since the publication of Bahl's admirable series of studies on Phcrctiuia (1919 and 1922), Lanipito (1924) and Woodwardia (1926). The brief descriptions of system- atists who classified their species as '' micronephridial," " megane- phridial," and " mixed mega-and-micronephridial " are now shown to be totally inadequate. Nevertheless, what little can be judged from the existing knowledge yields points of considerable interest. In the first place, loss or reduction of nephridia when it occurs seems to take 270 GRACE EVELYN PICKFORD place at the anterior end, e.g. in Pontodrttus, Sparganophilus and Diporochceta pcllitcida Bourne (re last species see Stephenson, 1925). Bahl considers that the first step in nephridial multiplication was the separation of the nephrostome, which then either disappeared or formed with accompanying nephridial cells a separate septal meganephridium opening into the gut, while the main mass of the nephridium hroke up to form funnel-less integumentary nephridia. In Pheret'una the septal nephridia have also undergone multiplication to the micronephridial condition. If this view be provisionally accepted, the two trends, sep- aration of the nephrostome and multiplication, may be considered in- dependently. As regards the former, numerous cases can be found in the literature in which " meganephridia " occur only in the middle and posterior regions of the body. In "Lampito" (Mcgascolex} trilobata Steph. and " L." niauritii Kinb., Bahl found that the septal meganephridia commenced in segment 19, while in Woodwardia bahli Steph. they commence at 24/25. Benham (1905) describes two species of Spenceriella, — " Diporochata " gigantca and " D." shakes pearl, which are " micronephric " but retain large paired nephrostomes in each seg- ment. Unfortunately he does not say how far forward these occurred. In Comarodrilus grai'd\i Steph. " micronephridia " occur in the anterior part of the body as far back as segment 12; behind this "megane- phridia " only. In the development of Octochcrtus multiponts, Beddard found (1892) that the nephrostomes degenerate after their separation from the nephridial mass, but that they may persist in the posterior segments. These cases appear to be merely examples of a very gen- eral phenomenon, viz., the tendency for the nephrostomes to disappear anteriorly. An interesting case is that of Hou>ascolc.r corethntnis Mchlsn., a species which is transitional both for perichaetine and micro, nephridial conditions. The setae are lumbricine in the anterior and middle regions and perichaetine posteriorly, while " meganephridia " displace the " micronephridia " posteriorly. The case of nephridial multiplication seusu strict o requires a sta- tistical investigation, but observations such as those of Bahl on " Lampito " and Pheret'una spp. and of Stephenson on Hoplochatella kiuneari Steph. indicate that a great multiplication in numbers of micro- nephridia in the clitellar region may be a general phenomenon. While there is thus considerable evidence that nephridial and neph- rostomal reduction follows the primary metabolic gradient, occurring first at the anterior end, the case of nephridial multiplication is not at all clear cut and the issue is frequently confused by the occurrence of pharyngeal nephridia (tufts of funnel-less nephridia opening into the pharynx) in the most anterior segments. The clitellar region, which METABOLIC GRADIENT IN OLIGOCH^TS 271 is sometimes the region of greatest multiplication (ride supra}, is not known to be the region of lowest metabolism, since the physiological gradient has not been investigated for the species concerned, but evi- dence from other species suggests that the clitellar region is too far forward to coincide with the region of lowest metabolism. If Perkins' (1929) speculations as to the anabolic gradient are well founded, it is possible that certain morphological features such as nephriclial multi- plication in the clitellar region might be interpreted more readily by a correlation with this rather than with the total metabolic gradient. Ex- amples have been cited above in which " micronephridia " are replaced by or co-exist with " meganephridia " in the posterior part of the body. Sometimes, c.. Univ., 1: 293. PERKINS, M., 1929. Growth-gradients and the Axial Relations of the Animal Body. Nature, 124: 299. SHEARER, C, 1924. On the Oxygen Consumption Rate of Parts of the Chick Embryo and Fragments of the Earthworm. Proc. Roy. Soc. London, 96: 146. STEPHENSON, J., 1912. On a New Species of Branchiodrilus and Certain Other Aquatic Oligochaeta, with Remarks on Cephalization in the Naididse. Rcc. Ind. Mus., 7: 219. STEPHENSON, J., 1921. Contributions to the Morphology, Classification, and Zoogeography of Indian Oligochseta II. On polyphyly in the Oligochreta. Proc. Zo'ol. Soc. London, Part 1, p. 103. STEPHENSON, J., 1923. The Fauna of British India ; Oligoch.Tta. Taylor and Francis, London. STEFHENSON. J., 1924. On some Indian Oligochsta, with a Description of Two New Genera of Ocnerodrilins. Rcc. Ind. Mus., 26: 317. STEPHENSON, J., 1925. On some Oligochjeta Mainly from Assam, South India, and the Andaman Islands. Rcc. Ind. Mus., 27: 43. WATANABE, Y., 1926. On the Electrical Polarity in the Earthworm, Pcrichccta communissima Goto et Hatai. Sci. Rep. Tohoku /;;;/>. Unir., 3. No. 2. p. 139. DISTRIBUTION OF SETsE IN THE EARTHWORM, PHERE- TINIA BENGUETENSIS BEDDARD 1 P. B. SIVICKIS (From the Zoological Laboratories, University of the Philippines, Manila and Lietuvos Universitctas, Kaunas, Lithuania.) The oligochaet genus, Phcretima, which occurs abundantly in the Philippines and other oriental countries, is characterized by the pres- ence of a large number of seta? on each segment except the most an- terior. Taxonomists have regarded the distribution and number of setae as specific characteristics, but apparently have observed that the number varies on different segments, since they usually specify the segment for which the number of setae is given (Michaelsen, 1900; Stephenson, 1923). No data have been found, however, concerning variation in number of setae on a particular segment. Counts of setae made by the writer show a considerable range of variation, both in the number of setae on corresponding segments of different individuals and on different segments of the same individual. Moreover, the numbers of setse on different segments of the same individual vary along the axis in a way which suggests a relation to the longitudinal physiological gradients. Data are given below concerning these variations. MATERIAL AND METHODS Pherftima benguetensis Beddard, the species on which the counts were made, is common in the Philippines. During the greater part of the rainy season the worms are found in large numbers near or on the surface of the ground. By the end of the rainy reason they become heavily parasitized by gregarines and later disappear almost completely, but whether the disappearance is due to death or to movement away from the surface of the ground is not known. Counts of setae were made on one hundred animals. Fifty of these were collected on the campus of the University of the Philippines and fifty from the town of Pasig near Manila. The latter were somewhat larger than the former, but their general specific characteristics indi- cated that both lots belonged to the same species. 1 The data presented in this paper were obtained while the writer was a member of the Department of Zoology of the University of the Philippines. Acknowledgments are due to Miss Paz Lorenzo, Mr. D. Quajunco and Mr. G. T. Lantin for assistance. My thanks are due to Prof. C. M. Child for critical review of this paper. 274 SET.E IN EARTHWORM 275 The counts were made on animals preserved in formalin. For counting they were opened along the mid-dorsal line, the internal organs were removed, and the body wall was cut into pieces of a size con- venient for microscopic examination between two slides. Counts of such pieces were either made at once or the two slides with the piece between them were tied together and placed in a hot one per cent solu- tion of KOH for five hours or more, until they became transparent, but were removed before maceration had proceeded so far that the setae were freed from the tissue. A section along the dorsal mid- line is more satisfactory for such preparations than a section elsewhere because the dorsal wall is thicker than in other regions, and since the KOH attacks the edges of the preparation first, the thicker dorsal wall is not destroyed before the other parts have become sufficiently trans- parent. After maceration the pieces were mounted in glycerol and all the setae on the segments selected were counted under a low power of the compound microscope with the aid of a mechanical stage. Par- ticular care was taken to make certain that all setae on each segment selected were included in the counts. In the region of the clitellum counts are less readily made than elsewhere because the thickening of the body wall in this region makes it difficult to see the setae. Since there are no setae on the first segment, counts were begun with the second, and further counts were made on the fifth, tenth, fif- teenth, etc., that is, on every fifth segment up to the sixtieth. In order to minimize possible errors which, however, proved to be less than was feared, in counts on the fifteenth segment, a segment of the clitellum. counts were made on the segment next anterior (13) and the segment next posterior (17) to the clitellum. Counts from the posterior di- rection began with the last posterior segment and were made on every fifth segment until the sixty-fifth segment from the posterior end was reached. This procedure leaves a short middle region uncounted in some animals with a large number of segments, but since the mean number of segments can readily be extrapolated in this region, the results are not seriously affected. The method of making counts in two directions from each end of the body is regarded as preferable to that of making counts from an- terior to posterior end, because by the latter method the most pos- terior segment counted is rarely the last segment of the body and rep- resents different levels in different cases. DISTRIBUTION AND SIZE OF SET.£ Each segment except the most anterior possesses a large number of setae more or less uniformlv distributed about the circumference, 276 P. B. SIVICKIS but with occasional gaps and occasional duplications. The setae are less than a millimeter in length, and taper slightly from the base to a blunt tip. 1 io H « 5i Jo J! Jo v! ;o 55 to £5 to 5! 50 ~ ^o 35 ' jo 25 Jo is 10 5 FIG. 1. Graph from the data of Table I showing the variation in numbers of setse along the main axis of the body in Phcretima bcnguctcnsis. Ordinates represent the mean numbers of setse (M') on particular segments; abscissae repre- sent segment numbers. The anterior end is at the left. Setse from three regions of the body have been isolated by boiling in KOH pieces of the body wall from the selected regions and have been measured with an ocular micrometer. The data of such measure- ments are as follows : On segments 2-5, length 0.6 mm. ; diameter 0.08 mm. 36-40, 0.3 mm.; 0.05 mm. Last ten segments, 0.5 mm. ; 0.03 mm. These measurements indicate the variation in size of the setae. The longest setae are found on the anterior and posterior segments, the shortest in the middle regions. From the anterior end the seta- very gradually decrease in size to the clitellum. For some distance posterior to the clitellum the setae are only about half the length and little more than half the diameter of those on the anterior segments. Pos- terior to the middle of the body they begin to increase in size and for the most posterior segments they are almost as long, though less in diameter than those at the anterior end. In general the length of the setae varies inverselv as their number. IN EARTHWORM 277 w ca H o u O (O rt !"« 's u ST3 K C O £ ^^ tuO be 'S O O C a« s I O M || 3 V -l CN w-l ~* — < 1^1 ^H O ~- 'C O oo O O °O o vo O r^ O *"-1 OO lO' — i OO O u"5 10 O O O O O OOOOO < — O O «o 1'— i IO t^- 00 ^f OO "5 fN t — ^ ^ *^ ^* O OO O \o if~> \o t^- \O so t~^- sO sO ^ — - £^3 I- r- o c > 3 sO sO so so so t^. sO so OO so t^» sO t^ OO t^ OO 10 sO 10 '— 100 so 10 sO »"c. fO IO "to ^ c -f e>* IO-* ceo- sOio' o. — 1'°° 00 •* sO^. SOIO 00 SO tC«! O 10 oo'"? SO IO 10 O> &*>. sOio ^00 a^ "O i/l 1^3 O1 06^ O so 00 ^ 0^. SO 10 0°. 278 p. B. SIVICKIS COUNTS OF SETVE The numerical data for the first ten and the last ten animals of the hundred counted are recorded in full in Table I as a sample indicating how the counts run. Animals 1-10 of the table are from those col- lected on the University campus, animals 91-100 from those collected at Pasig. The first vertical column gives the number of the individual in the series, the following column the number of setae counted on cor- responding segments. The first vertical section of the table gives counts from the anterior end to the sixtieth segment, the second sec- tion, counts from the posterior end to the sixty-fifth segment from that end. The last column gives the number of segments in each animal. In the last two horizontal lines of the table are given the mean values (-M) and the standard deviation (o-) as calculated by the standard formulae for corresponding segments of all animals counted, that is, each value of M and a given is the value for one hundred corresponding segments. The variations of M at the differ- ent body levels are plotted in the graph (Fig. 1). Examination of the data recorded in the table shows that in spite of a considerable variation in the number of setae per segment of any individual, the general course of the variations in different regions is well expressed by the means. The number of setae is relatively small on the anterior segments, but increases rapidly to the twentieth seg- ment, beyond this more slowly to the thirtieth segment, where the maximum number of setae per segment is attained. Posterior to this segment the number of setae decreases gradually to the posterior end of the body. DISCUSSION The very definite course of variation in number of setae along the body of Phcrctiina suggests that it must be correlated with regional physiological differences of some sort, and since it is gradual and in opposite directions in anterior and posterior regions, the possibility that it may be in some way correlated with the longitudinal physiological gradient in the body is also suggested. Nothing is known concerning the gradients in Pherctima, but in most other oligochaets examined a double gradient has been found. Hyman (1916) has found in most of the microdrilous oligochaets a decrease in susceptibility from the anterior end posteriorly to a certain level and an increase from this level to the pos- terior end. Hyman and Galigher (1921) found a similar double gradi- ent in oxygen consumption in Lumbnciihis and Nereis. Perkins ( 1929), investigating oxygen consumption, total iodine equivalence, amount of glutathione and total sulphur content in different regions of the body IN EARTHWORM 279 of an earthworm (unnamed), also finds differentials which vary in two directions. If such a double gradient exists in Phcrctinia, as is prob- able, the smaller numbers of setae occur at the higher, and the larger numbers at the lower levels of the respiratory gradient. We know nothing at present concerning the nature of the relations between gradients and setae, but it may be provisionally suggested that a develop- ing seta sac in the regions of more intense metabolism inhibits the de- velopment of other seta sacs over a greater distance than in regions of lower metabolism, consequently at the higher gradient levels fewer setae develop on the circumference of the segment than at lower levels. Such an inhibiting action of a developing part or organ on other similar organs within a certain distance from it is very generally recognized by both botanists and zoologists, and in various cases the range of this effect appears to be very definitely associated with the intensity of metabolism in the part concerned. Whether this suggestion of a possible relation between the numbers of setae on different regions of the body is correct must remain for further investigation to determine. In addition to the regional variations in numbers of setae, individual variations in number on corresponding segments appear in the table. The standard deviation a is lowest in the anterior region of the body. This is particularly evident anterior to the tenth segment. The highest value of a appears in the posterior region, particularly in the ten pos- terior segments. Between these extremes a fluctuates between 4.62 and 6.00. The relatively low . Zool.. 20: 99. HYMAN, L. H., AND GALIGHER, A. E., 1921. Direct Demonstration of the Ex- istence of a Metabolic Gradient in Annelids. Jour. Exp. Zool., 34: 1. MICHAELSEN, W., 1900. OHgochaeta. Das Tierreich, No. 10. PERKINS, M., 1929. Growth-Gradients and the Axial Relations of the Animal Body. Nature. 124: 299. STEPHENSON, J. F., 1923. Oligochaeta. Fauna of British India. London. STUDIES ON THE PHYSIOLOGY OF THE EUGLENOID FLAGELLATES II. THE AUTOCATALYTIC EQUATION AND THE QUESTION OF AN AUTOCATALYST IN GROWTH OF Eliglcna THEODORE L. JAHN BIOLOGICAL LABORATORY, UNIVERSITY COLLEGE, NEW YORK UNIVERSITY The theory of a catalyst of growth, as proposed by Robertson, has been the stimulus for a number of investigations to determine the pres- ence or absence of an autocatalyst in protozoan cultures. The earlier investigations have been reviewed previously (Jahn, 1929). and it was shown experimentally at this time that the growth rate of Eitglcna in mass cultures of high concentrations of organisms was not higher than in cultures of relatively low concentrations, but that in most cases the reverse was true. It is the purpose of the present paper to reanalyze the experimental data previously obtained (Jahn, 1929) from the point of view of rela- tive rate of division at various times during the period of observation. It will be shown, first, that the division rate, as calculated from the autocatalytic formula used by Robertson in his work on ciliates, is a progressively decreasing quantity, and hence that this autocatalytic equation can not be interpreted to involve an autocatalyst which effects an increase in division rate ; and second, that, on the basis of experi- mental evidence, the autocatalytic equation may fit the growth curve of Eliglcna cultures. On the basis of the experimental evidence, it is believed that Robertson's theory of an autocatalyst of growth is un- necessary to an interpretation of the experimental data obtained in the case of Eliglcna. The writer is deeply indebted to Professor R. P. Hall for sug- gestions offered during the preparation of this paper. THE DIVISION RATE AS DERIVED FROM THE AUTOCATALYTIC EQUATION The equation for an autocatalytic chemical reaction has been ap- plied by Robertson (1923) to the rate of growth of ciliates in isolation cultures. The differential form of the equation is doc -j- = Kx(A • x), (Equation 1) 281 THEODORE L. JAHN where .r is the number of organisms, A the maximum number attainable in a given amount of medium in question, and t is time. When in- tegrated this becomes log -j-2— = AK(t - /,), (Equation 2) A X where t\ is the time when x = A/2. The differential equation states that the rate of increase in number of organisms at any time is proportional to the number present at that time and to the difference between that number and the maximum number attainable. Or one may let A equal the original (and also total) food supply. If this is measured in units, such that one organism utilizes on the average exactly one unit of food between divisions, then at any time the amount of food consumed will be equal to the number of divisions that have taken place. After the first few divisions the number of divisions which have occurred is approximately equal to the number of organisms present. Therefore A-x may be regarded as the available food supply, and in this sense the equation means that the rate of increase in number is proportional to the number present and to the amount of free food material. In either case dx/dt is the rate of increase of the total number of organisms as related to time. This, however, is interpreted incorrectly by Robertson as the rate of division of the organisms. The actual rate of division, that is, the average frequency of division (or fission) per unit time, is not d.\'/dl but — — , or the rate of increase of the total number of organisms x divided by the number (x) present at any given time (t). This we may represent by D, and then we may restate the division rate as D = dx/dt = KX(A - X] = _ x} (Equation 3) x x The division rate therefore varies directly as A-x. Since A is constant, and .r is continually increasing, and A-x therefore decreasing, it can readily be seen that the division rate as derived from the autocatalytic equation is a decreasing linear function of x. If the division rate is plotted against time, the result will be a sigmoid curve with a negative slope, practically the same as the original integral curve except that the abscissa will be shifted and the ordinate sign reversed (Fig. 1). Since the division rate is continually decreasing, the intervals between divisions will become progressively longer as the culture is continued. More recently Robertson (1928) has proposed a new equation for STUDIES ON EUGLENOID FLAGELLATES 283 the growth of Metazoa. The differential form of this equation is dx _ kp , . dt 1 +p V)(A •-*), (Equation 4) where p is the constant proportion between nuclear and cytoplasmic increment and b/p is the excess of nuclear material which is present at the initiation of development, that is at the moment of fertilization, Graph A Graph C GrapK B D FIG. 1. Type curves computed from the autocatalytic equation. A. Differential curve showing dx/dt, the rate of increase in numbers, plotted against time (t). B. Integral curve showing x, the number of organisms, plotted against t; .v approaches the value x = a as an asymptote. C. Division rate (D) plotted against time. D approaches zero as an asymptote. D. Division rate plotted against numbers of organisms (.r). D is a decreasing linear function of x, becoming zero when x is equal to a. The scale of t is the same in graphs A, B, and C. b being the same quantity translated into its cytoplasmic equivalent through multiplication by the proportionality factor p. Whatever meaning p could assume in a protozoan culture is difficult to state, but the division rate as calculated from this new equation is also a de- creasing quantity as in the previous set of equations. D = dxfdt kpA b kp x (1 + p)* 1 + p (A • - b - - .T) . (Equation 5) The division rate as calculated from the new equation is equal to the sum of two quantities, one of which varies as the reciprocal of .v and 19 284 THEODORE L. JAHN the other as a decreasing linear function of x. The sum of these two quantities is, of course, a decreasing function. Therefore, the division rate, whether calculated from the new or from the old equation, is a decreasing function. The above modification (equation 3) of the autocatalytic equation has been expressed by Brody (1927), who states, "It signifies that the relative rate of growth is directly proportional to the growth im- pulse," (a-x}. Since Brody was considering the application of the formula to Metazoa, he did not express the idea that the modification might be used to represent division rate of cells, or that under such conditions it would indicate a decrease in division rate. Snell (1929) points out that since the volume of a growing organism changes, equations derived from the law of mass action can not be ap- plied to growth without considerable modification. The value of — Jv calculated from the modification he proposes is also a decreasing func- tion as in the preceding equations. A SenesV .6 .2 0 T >eries days FIG. 2. Graphs showing the division rate (D) plotted against time for the three cultures of Series V and a composite curve for the three cultures of Series VI. Values of D were computed from the equation dx/dt The rate of increase of the total, dx/dt was determined by the graphical differ- entiation of the experimental growth curves. These values were then divided by the corresponding values of x to give the values of D for the times (0 under consideration. It is to be noted that this is a descending sigmoid curve such as is to be expected if the autocatalytic equation is applicable to the case. (See also graph C, Fig. 1.) STUDIES ON EUGLENOID FLAGELLATES 285 THE DIVISION RATE OF Euglena The growth curves of Euglena from four series (I, III, V, and VI) previously described by Jahn (1929) have been differentiated graph- ically to give values of d.r/dt for various values of t. If the values of dtf/dt are divided by corresponding values of x, one may arrive at values of - -, or D, for the values of t considered. If D is plotted X against t, the result is a descending sigmoid curve. The division rate curves for Series V and VI are shown in Fig. 2; the curves for the other series are similar in form. In Table I are shown the values of the division rate as computed from analysable data. TABLE I . dx/dt X Culture number Day i 2 3 4 5 I, 1. 63 36 33 31 Ill, 1 80 59 44 V, 1. 1 00 80 50 40 V, 2 80 73 50 43 34 V, 3 .72 .69 55 40 37 VI, 1, 2, and 3 (averaged). . .83 79 66 55 50 DISCUSSION The results of the above analyses demonstrate a decreasing sigmoid division-rate curve for cultures of Euglena. This indicates that the growth rate of Euglena closely simulates the reaction rate expressed by the autocatalytic curve, and that the periods between cell divisions in a single line of cells become progressively longer as the culture is continued. The writer's observations on Euglena thus differ from those of Robertson (summary, 1924), who maintains that his experiments also show the growth of Infusoria (Enchdys) to be autocatalytic, since in isolation cultures the division rate is low at first but becomes pro- gressively higher with each successive division. The autocatalytic formula, as stated, can be adopted only on the assumption that the food supply is limited from the beginning and is therefore continuously being decreased by the growth of the organisms. In Robertson's ex- 286 THEODORE L. JAHN periments the available food supply was not decreasing during the period of observation but was increasing due to bacterial growth, and Rob- ertson's cultures also show an increase in division rate and not a de- crease as required for the application of the autocatalytic equation. Hence, it is obvious that the autocatalytic growth curve cannot be ap- plied to such experiments with ciliates. The experiments of the writer with Euglcna were conducted under more readily controlled conditions than were previous experiments with ciliates. Since bacteria are not a source of food for Euglcna, it is safe to assume that the few bacteria present did not accelerate appreciably the division rate of the organisms. Therefore, the food of the flagellates was limited to the inorganic salts initially present in the medium and the carbon dioxide dissolved in the water. Since the primary physical factors (light and temperature) affecting growth were constant, and the chemical substances (carbon dioxide and inorganic salts) entering the reaction were continuously decreasing as the reaction progressed, the experiment may be considered as more nearly re- sembling a closed system — such as that to which the autocatalytic equa- tion is applied in chemistry — in which the variables are food, flagellates, and waste products of flagellates, the food being converted into more flagellates and waste products. The available food material was con- tinually decreasing as the organisms increased in number. Therefore, the autocatalytic equation may be applied to the experiments of the writer; whereas, it cannot, as previously explained, be applied to ex- periments of other workers with ciliates. Richards (1928) has shown that the division rate of yeast cells in a limited volume of medium is a decreasing quantity; and further- more, that when the medium was changed frequently, the division rate remained practically constant. Hence, neither his results nor those of the writer furnish a basis for the assumption of an autocatalyst capable of accelerating division rate. Robertson's concept of autocatalysis in Protozoa has, of course, grown out of his numerous applications of the autocatalytic equation to growth curves of plants, of man, and of other animals. As pointed out above, the division rate of Protozoa in cultures, as calculated from the autocatalytic equation, is an ever decreasing rather than a progres- sively increasing quantity. In metazoan growth Robertson was not measuring division rate of cells, but rather the increase in weight (or increase in total number of cells') of a many-celled body. In Protozoa, on the other hand, it was the rate of cell division as well as the rate of increase of total number which he measured, and he assumed that an increase in the latter necessarily involved an increase in the former. STUDIES ON EUGLENOID FLAGELLATES The rate of increase of the total number of cells in a metazoan or in a protozoan culture is accelerated during the early phase or phases of growth, but if the growth is autocatalytic, the rate of cell division is continually decreasing. In either case, the rate of increase in total number of cells (provided the increase follows the autocatalytic curve) is accelerated because the number of growing units is increasing— not be- cause of an acceleration of the growth rate of the individual units, but in spite of a decrease in the growth rate of these units. The rate of increase of the total number of cells and the division rate of the in- dividual cells are two distinct conceptions which should not be confused. BIBLIOGRAPHY BRODY, SAMUEL, 1927. Growth and Development with Special Reference to Do- mestic Animals. III. Growth rates, their evaluation and significance. University of Missouri, Agr. E.r[>. Station. Bulletin 97.. CUTLER, D. W., AND CRUMP, L. M., 1924. The Rate of Reproduction in Artificial Culture of Col Indium col fro da. Part III. Blochem. Jour., 18: 905. JAHX, T. L., 1929. Studies on the Physiology of the Euglenoid Flagellates. I. The relation of the density of population to the growth rate of Euglena. Biol. Bull, 57: 81. JAHX, T. L., 1929. The Autocatalytic Equation and the Question of an Auto- catalyst in the Growth of Euglena. Anat. Rcc., 44: 224. RICHARDS, OSCAR W., 1928. The Rate of Multiplication of Yeast at Different Temperatures. Jour. Phys. Chcm., 32: 1865. ROBERTSOX, T. B., 1923. The Chemical Basis of Growth and Senescence. Phila- delphia and London. ROBERTSOX, T. B., 1924. Principles of Biochemistry. Philadelphia and New York. ROBERTSOX, T. B., 1928. The Dynamics of Growth and Differentiation. Arch. Sci, Biol. (Napoli), 12: 235. SNELL, GEORGE D., 1929. An Inherent Defect in the Theory that Growth Rate is Controlled by an Autocatalytic Process. Proc. Nat. Acad. Sci., 15: 274. THE EFFECT OF LACK OF OXYGEN ON THE SPERM AND UNFERTILIZED EGGS OF ARBACIA PUNCTULATA, AND ON FERTILIZATION ETHEL BROWNE HARVEY (From the Washington Square College, New York University and the Marine Biological Laboratory, Woods Hole) It has been shown in a former paper (Harvey, 1927) that when fertilized eggs are deprived of oxygen, development is arrested, and the eggs remain in whatever phase of division they were in when oxygen was taken away ; they gradually resume development and pass through subsequent phases of division when oxygen is readmitted. The experi- ments were performed on two species of sea-urchin occurring at Naples, Strongylocentrotiis (Paracentrotus) I'nidus and Echinus microtubcr- culatus. Some of these experiments have been repeated on the Woods Hole species, Arbacia punctulata, and have given the same results. The present paper deals with the effect of lack of oxygen on the un- fertilised eggs and the sperm of Arbacia punctulata, and on the fertiliza- tion process in these eggs. The work was done during the summer of 1929 at the Marine Biological Laboratory of Woods Hole. I wish to thank the Director for the facilities of the laboratory. The experiments on unfertilized eggs and sperm were carried out for the most part by bubbling hydrogen through a suspension of eggs or sperm in sea-water in a closed glass vessel, from which they could be drawn off at desired intervals for observation. The connection be- tween the hydrogen tank and the glass vessel included a quartz tube containing platinized asbestos which was kept heated to redness to remove the last traces of oxygen ; from here to the glass vessel, the connection was entirely of metal and glass, sealed with De Khotinsky cement, to avoid the leakage of air which takes place through rubber connections. The length of time for complete removal of air and re- placement by hydrogen, of course, depends on size of vessel, amount of sea-water, rate of bubbling, etc., but under the conditions of the experiments it required approximately twenty minutes. That a state of complete anaerobiosis obtained was shown by the fact that under similar conditions the luminescence of luminous bacteria was stopped, as ascertained by E. N. Harvey. When unfertilized eggs are thus kept without oxygen, they are very 288 EFFECT OF OXYGEN LACK ON EGGS OF ARBAC1A 289 little affected. During a period of exposure of 8 hours, one can ob- serve no difference in appearance between the eggs when drawn from the hydrogen chamber and the control unfertilized eggs ; and the ex- posed eggs can be fertilized and develop normally. For the first 3 hours, the eggs when withdrawn from the hydrogen chamber can be fertilized with as much ease and as rapidly as eggs kept in air ; the fertilization membrane comes off at the same time (1-2 minutes) and the first cleavage plane comes in at exactly the same time (about 50 minutes) as in the control lots. When, however, eggs which have been exposed over 3 hours to hydrogen are withdrawn and fertilized, there is a slight lag (^r^Vz minutes) in the formation of the fertilization membrane, a tendency of the membrane to adhere to the egg, a slight crenulation of the egg surface, and a lag of from 2 to 5 minutes in occurrence of the first cleavage. This was not due to the bubbling, for when air in place of hydrogen was bubbled for the same length of time through the same amount and concentration of eggs, these eggs when fertilized showed no lag in the formation of the fertilization membrane nor in time of cleavage over eggs kept at the same time undisturbed in watch glasses and fertilized. When eggs, which have been kept in hydrogen for three or more hours, are withdrawn and left in air unfertilized for 45 minutes and are then fertilized, they show no lag in membrane formation or in time of cleavage. The lag evidently represents the recovery time from exposure to the oxygen - free atmosphere. The unfertilized eggs have therefore a very short recovery period after a prolonged exposure to hydrogen, and recover instantly after a shorter exposure. They are thus in marked contrast to fertilized eggs, which require a comparatively long period (!/o hour to 1 hour) for recovery from exposure to hydrogen lief ore resuming development. It may be that the longer recovery period of the fer- tilized eggs from the effects of lack of oxygen is related to their greater oxygen consumption as compared with that of the unfertilized eggs. After exposure for 6 or 8 hours to either hydrogen or air (in the ap- paratus used) some of the eggs become cytolyzed, owing probably to the mechanical disturbance of the bubbling; the effect increases with time until, after about ten hours, practically all the eggs are cytolyzed. Whether, therefore, the life of the unfertilized egg is prolonged by lack of oxygen could not be determined by these experiments. Loeb and Lewis (1902) found that unfertilized eggs would live somewhat longer in absence of oxygen (64 hours) than in air (48 hours), and very much longer in a weak concentration (N/1000) of KCN (112 hours). This latter effect may, however, be due to destruction of harmful bacteria by the KCN as pointed out by Gorham and Tower (1902). 290 ETHEL BROWNE HARVEY For experiments on sperm cells of Arbacia, a fairly concentrated suspension was used, one drop of fresh undiluted sperm to 10 cc. of sea-water {i.e., about .6 per cent), giving a decidedly milky appearance. In such a concentration sperm live longer and retain their fertilizing power for a longer time than in a more dilute suspension, probably owing to COL> production as shown by Cohn (1918). When hydrogen is bubbled through the sperm suspension for about two hours, the sperm are motile immediately on withdrawal from the hydrogen cham- ber, or at least as quickly as they can be observed under the micro- scope. The lots of eggs into which they are immediately drawn form fertilization membranes and cleave at the same time as the controls. After an exposure of 2 to 3 hours, the sperm recover motility within a few seconds and fertilize eggs with a very slight lag over the controls. After an exposure of more than 3 hours, some of the sperm do not recover motility and only a fraction of the eggs to which they are added are fertilized. After 4 hours, the sperm are all inactive, do not fertilize the eggs and never recover. A control experiment in which air in place of hydrogen was bubbled through a similar amount and concen- tration of sperm showed that the deleterious effect is due to lack of oxygen and not to the mechanical agitation, since these sperm were just as active and potent for fertilization even after 9 hours of bubbling as are fresh sperm. It is interesting to note that the prevention of oxidations by means of a hydrogen atmosphere gives a different result from that obtained by the use of cyanides. Drzewina and Bohn ( 1912) found that the sperm of Stronglyocentrotus would survive and remain potent for 48 hours in KCN (1 : 1,000,000), and that when they were subjected to KCN for long periods (1 to 10 hours) they caused a more normal development of eggs than when subjected for a short period (30 minutes to 1 hour). Cohn (1918) found that KCN rendered Arbacia sperm inactive and prolonged their life, and in fact suggested that " whatever decreases the activity increases the length of their life." This is certainly not true for hydrogen. It may be, however, that some other factor associated with the absence of oxygen, such as the lack also of CO, is responsible for the death of the sperm in my ex- periments. A study was made of individual sperm cells in the absence of oxygen by using a modified Engelmann chamber to which hydrogen was ad- mitted and the sperm kept in a hanging drop (see Harvey, 1927). It was found that in many cases enough oxygen leaked through the vaseline seal with which the cover was mounted on the chamber to enable the sperm to keep their motility for several hours. By entangling the sperm in platinized asbestos threads, it was possible in some cases to keep EFFECT OF OXYGEN LACK ON EGGS OF ARBACIA 291 them absolutely oxygen- free, and they became motionless within a half hour. If air was then admitted, the sperm immediately became motile. Even if the bubbling of hydrogen was stopped, within a very few min- utes the sperm became active. It apparently requires a very minute amount of oxygen for motility of the sperm. When sperm are kept in an Engelmann chamber without oxygen for two hours, they clo not recover motility on admission of air. They are killed by the absence i if oxygen even more quickly than when the experiments are done in bulk. The most interesting question in connection with lack of oxygen on eggs and sperm is whether fertilization can take place and the fertilization membrane be thrown off during complete absence of oxy- gen. An attempt to answer the question was made by keeping un- fertilized eggs in one drop and sperm in another drop very close together in an Engelmann chamber. Hydrogen was sent through for a half hour, then the chamber was shaken so as to make the drops coalesce and the sperm come in contact with the eggs, still keeping hydrogen passing through the chamber and the seal intact. It was found that when the sperm are completely immotile, they do not fertilize the eggs, probably because they cannot get to the surface of the egg; they go in currents around and past the eggs ; in no case is a fertilization mem- brane thrown off. On admission of air the sperm become motile and the membranes of the eggs are thrown off in 1 to 2 minutes as normally. If there is the slightest trace of air leaking in the chamber, sufficient for a few only of the sperm to be very slightly motile, some of the eggs are fertilized on mixing the drops, and fertilization membranes are thrown off, but no further development takes place until more air is admitted. The question, therefore, whether oxygen is necessary for membrane formation has not been answered. If there is absolutely no oxygen, the sperm are absolutely immotile and cannot fertilize the eggs, probably owing to mechanical difficulties, and no membranes are given off. Loeb also found that if the sperm cells of Strongylocentrotus were made immotile by NaCN, they were unable to fertilize the eggs even when squirted on eggs with jelly removed. If in my experiments, there is the slightest trace of oxygen, a few sperm remain motile and fertilize eggs which throw off membranes. If membrane formation does require oxygen, it is in an almost infinitesimal amount. It requires more oxygen for the development of fertilized eggs than it does for motility of sperm, fertilization of the egg and the formation of the fertilization membrane. ETHEL BROWNE HARVEY SUMMARY 1. Unfertilized eggs of Arbacia are not visibly affected by complete lack of oxygen for a period of 8 hours. After an exposure of 3 hours they recover immediately on admission of air ; after a longer exposure, when air is readmitted and the eggs are fertilized, there is a slight lag in the formation of the fertilization membrane and in time of cleavage. 2. Sperm of Arbacia are rendered motionless by lack of oxygen, but are otherwise unaffected for 2 hours. They recover immediately on admission of air. After 3 hours some of the sperm are irreversibly injured, and after 4 hours they are all killed. 3. When sperm are added to unfertilized eggs, both being in com- plete absence of oxygen, fertilization does not take place, and the fertili- zation membrane is not thrown off because the sperm are not motile, and cannot get to the surface of the egg. The membrane is thrown oft" immediately on admission of air. If there is the slightest trace of air, which may leak through the vaseline seal to the chamber, sufficient for only a few sperm to be very slightly motile, the eggs with which they come in contact throw off fertilization membranes, but do not develop further until more air is admitted. If oxygen is necessary for mem- brane formation, it is in an almost infinitesimal amount. LITERATURE COHN, E. J., 1918. Studies in the Physiology of Spermatozoa. Biol. Bull, 34: 167. DRZEWINA, A., AND BOHN, G., 1912. Effets de 1'inhibition des oxydations sur les spermatozo'ides d'oursin et, par leur intermediare, sur le developpement. Compt. rend. Acad. Sci., 154: 1639. GORHAM, F. P., AND TOWER, R. W., 1902. Does Potassium Cyanide Prolong the Life of the Unfertilized Egg of the Sea Urchin? Am. Jour. PhysioL, 8: 175. HARVEY, E. B., 1927. The Effect of Lack of Oxygen on Sea Urchin Eggs. Biol. Bull., 52: 147. LOEB, J., 1915. On the Nature of the Conditions which Determine or Prevent the Entrance of the Spermatozoon into the Egg. Am. Nat., 49: 257. LOEB, J., AND LEWIS, W., 1902. On the Prolongation of the Life of the Un- fertilized Eggs of Sea Urchins by Potassium Cyanide. Am. Jour. Physio!., 6: 305. THE EFFECT OF CONJUGATION WITHIN A CLONE OF PARAMECIUM AURELIA DANIEL RAFFEL (From the Zoological Laboratory of the Johns Hopkins University) INTRODUCTION On the effects of conjugation in paramecium, particularly in re- lation to the production of inherited variations, the results of investi- gators are in conflict. Jennings (1913), working with both Para- mecium aurelia and Paramecium caudatum, reported that conjugation increased inherited variations : that it caused the production of diverse biotypes. The members of a clone — a population derived by fission from a single individual, whether an ex-conjugant or not — remained nearly or quite uniform in their inherited characteristics so long as conjugation did not occur among them. But after conjugation within such a clone, the inherited characteristics of descendants of the different ex-conjugants had become diverse. Thus by conjugation many dif- ferent biotypes had been produced, the descendants of each ex-conjugant constituting a single uniform biotype. Calkins and Gregory (1913), on the other hand, reported that there is in Paramecium caudatum as much variation among the de- scendants of the four individuals produced by the first two fissions of a single ex-conjugant as was found between the progeny of different ex- conjugants. They conclude that, " The results of this study show that physiological and morphological variations in the progeny of a single ex-conjugant of Paramecium caudatum are fully as extensive as the variation between the progenies from different ex-conjugants " (p. 523). Jennings (1916, p. 528, and 1929, p. 188) has tried to show that the results of Calkins and Gregory are invalidated by uncontrolled sources of error. On the one hand, he holds that their method of culture permitted continuing environmental differences between their different populations, such as would give rise to differences that would appear to be hereditary, although they were not. On the other hand, he notes the occurrence of conjugation within some of their cultures and the fact that this might readily have occurred undetected. This would vitiate their conclusions. 293 294 DANIEL RAFFEL Obviously, the situation calls for a new investigation of the matter, in which such methods shall be employed as shall certainly exclude the possibility that environmental differences affect the results, while at the same time the occurrence of unobserved conjugation is ex- cluded. It is such an investigation that is here presented. In order to assure a uniform environment for all the lines of descent an elab- orate technique was employed. This is described on later pages. The method involved, first, the use of a synthetic culture medium of known composition, with pure cultures of food organisms and uniform glassware; second, continuation of the uniform conditions by the cultivation of the paramecia under aseptic conditions; third, frequent testing of the culture fluid in which the organisms have lived in order to ascertain whether the uniformity of the environment has been maintained. In addition, the organisms are cultured singly and transferred daily to new drops of culture fluid, so that it is im- possible for conjugation to occur. Continuing diversities between lines cultivated simultaneously under such conditions can be inter- preted only as caused by constitutional differences among the organ- isms, not as due to diversities in food or cultural conditions, or other extrinsic factors. Taking these precautions, two comparisons are made. First, a population descended from different ex-con jugants is compared with a population derived by fission from non-con jugants of the parent clone. Second, four lines descended from each ex-con jugant are compared with one another, and the several such different clones are similarly compared. In this way it is possible to determine whether increased hereditary variation and differentiation into diverse bio- types are produced by conjugation. The investigation was suggested to me by Professor H. S. Jen- nings, and my thanks are clue to him for assistance throughout the work. I am also indebted to Rose Mahr Raffel, who assisted in the carrying out of the experiment, and without whose aid cultures of this magnitude, using the elaborate technique here employed, could not have been carried through. MATERIALS AND METHODS In this investigation an elaborate technique was used in order to subject all of the lines to identical environmental conditions. Great care was taken to eliminate any possible sources of variation. To this end the culture fluid, the food supply and the glassware used were standardized to as great an extent as was possible. The work which has been carried on for several years by Hartmann and his associates EFFECT OF CONJUGATION OF PARAMECIUM 295 at the Kaiser- Wilhelm Institut fur Biologic has made possible the use of synthetic culture fluids and pure cultures of food organisms for the cultivation of protozoa. The use of pure cultures of unicellular algae as food organisms appears first in the work of Luntz (1926) on the rotifer Ptcrodina clliptica and more recently in the work of Adolph (1929) with the ciliate Colpoda. The results of the work of Hartmann and his associates are given in a recent paper of Belar (1928). The following pages contain a detailed account of the methods used to obtain uniformity in the environmental conditions throughout this experiment. 1. Culture Fluid The culture medium used was a physiological salt solution of known composition. After many attempts to find a solution in which the race of Paramcciuui aurclia which was used would live, it was found that if the solution described by Pringsheim (1928) for the cultivation of algoe was altered so as to be neutral, it furnished an excellent medium for this organism. This modification was obtained by replacing the KH2PO4 used by Pringsheim by an equal molar concentration of K2HPO4. The composition of the solution was KNO, 0.5 gram, K2HPO4 0.06 gram, MgSO4 0.02 gram, FeCL, 0.001 gram, water 1000 grams. The water used in making this solution was redistilled from a still made of Pyrex glass and had in all cases a conductivity less than 1.05 X 10 6 mho. This solution was made up in quantities of one liter. It was then divided into portions of approximately 15 cc. in test tubes. These test tubes were plugged with non-absorbent cotton and the solution was sterilized in the autoclave for 15 minutes under 15 pounds of steam pressure. The solution was kept in this way for periods varying from a few days to two months before it was used. Tubes tested at intervals showed no bacteria and no measurable altera- tions in composition. 2. Food Organism The food organism used was a unicellular green alga, Stichococcus bacillaris.1 This was cultivated on 0.05 per cent Benecke's agar com- posed of water, 1000 grams; Agar-Agar, 15 grams; NH4NO3, 0.2 gram; CaCL, 0.1 gram; MgSO4.7H2O, 0.1 gram; and K,HPO4, 0.1 gram. The components of this agar were boiled together until the agar-agar was all dissolved. Five cc. portions were poured into test- tubes which were then sterilized in the autoclave under 15 pounds of steam pressure for fifteen minutes. These tubes were then " slanted " 1 1 am indebted to Professor W. R. Taylor of the University of Pennsylvania for the identification of this organism. 296 DANIEL RAFFEL in order to obtain a large, easily accessible surface. Twenty of these tubes were seeded from a pure culture of the alga on successive days. After this the slants were used in the order in which they had been seeded and as they were used they were replaced by new tubes seeded from them. The tubes in which the alga was cultivated were kept constantly before a north window in order to obtain sufficient light. Each da}- the tube of Stichococctis to be used that day and a fresh tube of the culture fluid were opened close to a flame into which their open ends were immediately thrust. Then a small quantity, approx- imately 5 cm., was scraped from the agar with a platinum loop which had just been sterilized in the flame. This small quantity of the alga was then quickly suspended in the solution and both tubes were im- mediately restoppered. Then a new tube of agar was seeded from the same tube and replaced in its proper place in the rack. Many tests of the suspension were made from time to time and in no case was any bacterial contamination found. An effort was made to have the suspension of alga in the solution always of the same density. How- ever, no method more accurate than a comparison of the appearance of the tubes was found for determining the success of this effort. For this reason, preliminary experiments were performed in order to de- termine whether or not the quantity of algae used affected the rate of reproduction of the paramecia. It was found that sufficient algae to produce a slight greenish tinge in the suspension furnished enough food for these organisms. Greater densities than this had no effect on the rate of reproduction even when they were far in excess of any used in the actual experimental work. At all times an excess of algae was assured and the drops containing the paramecia always showed a large number of the algae at the end of the period during which the organisms remained in them. It was found, however, that if the paramecia were kept in this so- lution with this single food organism, they were unable to live and re- produce. If a very slight trace of a Bacillus candicans was present, this difficulty was eliminated.2 Attempts were made to cultivate the par- amecia on this bacillus in the absence of the alga. All such attempts failed, and when a mixture of the two food organisms was used, the food vacuoles were dark green in color — indicating that the food supply was composed mainly of the alga. After a slight trace of this bacterium was once introduced into a culture of paramecium, it was perpetuated in the transfers of the organisms. As far as it was possible to de- 2 I am indebted to Professor W. W. Ford, Professor of Bacteriology in the School of Hygiene and Public Health of the Johns Hopkins University, for the identification of this bacillus. EFFECT OF CONJUGATION OF PARAMECIUM 297 termine by plating in the usual way, this bacterium was present in ap- proximately the same quantity from day to day in all of the many cases tested at random. It was thought advisable, however, to determine whether or not differences in the quantity of this organism present af- fected the rate of fission of the paramecia. There was no difference in the effect produced by the presence of any quantity of the bacterium less than that required to make the drops of culture fluid appear milky. At no time during the course of this investigation was this condition approached. 3. Glassware The various lines of paramecium used in this investigation were cultivated on slides with two concavities. It was found from prelim- inary work that different slides affected the paramecia differently. On some slides representatives of all the lines tested reproduced more rapidly than did other representatives of the same lines on other slides. The pH of drops of culture fluid which had remained on the different slides was tested and was found to vary greatly. Drops of fluid which had been identical when placed on the slides were found to vary by a whole pH unit within twenty- four hours. This showed that the glass of the various slides differs in solubility. New slides were then ob- tained, all of the same kind of glass. These slides were of French origin. After two days the organisms grown on these slides died out. No amount of washing the slides with various kinds of solvents made it possible to cultivate organisms on them. Investigation disclosed that this French glass is made by a process involving the use of lead. It appears that the presence of this element was responsible for the toxic effects of these slides on the organisms. When this was discovered, new slides were obtained which were of white glass and were all pro- duced by the same manufacturer. These slides were the only ones used in this investigation. Before they were used they were thoroughly washed in running water. Then they were washed in ether and 95 per cent alcohol in order to remove any organic matter with which they might have been contaminated. They were again thoroughly washed with running water, rinsed in several changes of tap water and finally rinsed in hot distilled water. Each day the slides were thoroughly washed in the following manner. First they were held, individually, in running tap water and the depressions were rubbed well with the thumb. They were then placed in a receptacle containing clean tap water. In this receptacle they were rinsed three times. Then, after the last tap water was thoroughly drained off, the slides were covered with hot distilled water. Thev were then dried on racks. 298 DANIEL RAFFEL In order to prevent contamination of the cultures by bacteria in the air, Petri dishes 100 mm. in diameter and 15 mm. deep were used as moist chambers. This made it possible to transfer the organisms with a minimum of exposure to the air. The dishes contained water at the bottom ; the two slides to each dish were supported above this on strips of glass. After the Petri dishes, the slides, and the glass plates were assembled, they were heated in the hot air sterilizer for one hour at 150° C. In order to facilitate the handling of the numerous dishes which were used, baskets were made from ^ inch wire netting which held a dozen Petri dishes in four tiers of three dishes each. The organisms were transferred by means of capillary pipettes. Each of these contained a plug of cotton inserted into its wide end. This is a precaution necessary to prevent contamination of the cultures by microorganisms which would otherwise be introduced by the rubber bulbs used on the ends of the pipettes. The glass part of the pipettes with their cotton plugs were kept in large museum jars, in which they were heated in the hot air sterilizer for one hour at 150° C. before each time they were used. 4. Method of Transferring Organisms Before the daily transfers were made, the Petri dishes were removed from the hot air sterilizer. Then two drops of the culture suspension were dropped into each concavity. Large pipettes which were drawn out until the ends were 2 mm. in diameter were used for this purpose. These pipettes, like the ones used for transferring the paramecia, were protected by cotton plugs and were sterilized before each time that they were used. The mouth of the test tube containing the suspension of culture fluid was sterilized in the Bunsen flame each time that it was opened. The tops of the successive Petri dishes were then raised on one side, the pipette was introduced and two drops were allowed to fall into each concavity. Four dozen dishes were prepared in this manner at one time. From time to time bacteriological plates were prepared from culture medium which was treated in the manner de- scribed above, after it was left for twenty-four hours. In every case the plates were negative, thus indicating that the technique was ab- solutely dependable. In transferring the animals a Petri dish containing the two slides was placed on the stage of the binocular microscope. Another dish containing new culture fluid was placed at the experimenter's right. One organism was then taken from each concavity and transferred to the corresponding concavity of the new dish. This was done very rap- idly, using a clean pipette that had just been removed from the jar of EFFECT OF CONJUGATION OF PARAMECIUM 299 sterile pipettes. A separate pipette was used for the organisms of each dish. The new dishes were then removed to the constant temperature chamber, in which they were left at a temperature of approximately 24° C. (There was in the history of the cultures variation in temper- ature from 22.2 °-26.2° C.) 5. Isolation and Sterilization of the Clone The various lines of Parainecinui aurelia used in this investigation are the descendants of a single individual which was isolated from a mass culture in the laboratory on July 29, 1929. Parpart (1928) has shown that spores of bacteria may be, and often are, carried within paramecium and that in washing these organisms, precaution must be taken to eliminate these spores as well as the bacteria external to the paramecium. For this reason, when the individual which was used to start the clone for this investigation was washed, the precautions suggested by Parpart were observed. The individual was first washed successively in five concavities containing sterile culture fluid. Then at intervals of one hour it was washed through five more similar quantities of fresh culture fluid. It was then placed in a con- cavity containing the regular culture suspension described above in which there was a slight trace of the Bacillus caudicans. No bacteria were added at any later time. From time to time throughout the course of the experiment bacteriological plates were made from drops from which the paramecia had been removed. Several dishes containing both ex-con jugant and non-con jugant lines were taken at random for this purpose. At no time did any plate made in this way indicate the pres- ence of any bacteria except the bacterium which had been introduced at the beginning. THE EXPERIMENT 1. Plan The plan of the experiment was as follows : A clone was obtained by allowing a single individual of Paramecium aurelia to multiply. A portion of the clone was induced to conjugate, while another portion was kept without conjugation. The former, after the separation of the pairs, yields lines of descent that constitute the ex-conjugant population, the latter the non-conjugant population. These two populations are later compared as to their mortality, fission rate, variation and the inheritance of the variations. For comparison with the results of Calkins and Gregory, a method similar to theirs was employed for the grouping and subdivision of 20 300 DANIEL RAFFEL the ex-conjugant lines. Each of the two members of a pair was al- lowed after separation to divide twice, yielding four individuals of common origin, the four quadrants. From each quadrant a line of descent was obtained. Each set of four quadrants derived from a single ex-conjugant is called, for convenience, a tetrad. The variation within single tetrads is compared with the variation among lines belong- ing to different tetrads (and so derived from different ex-con jugants). This tests whether the diversity among the descendants of a single ex- conjugant is as great as that between those of different ex-con jugants (as is maintained by Calkins and Gregory). 2. Description The experiment was begun with the isolation of a single organism on July 29, 1929. The progeny of this individual were propagated on slides by daily transfer until August 5, 1929. By this time there were approximately 1500 individuals present. On the morning of August 5, all of the individuals, except one, from each concavity, were transferred to two small sterile culture dishes contained within Petri dishes. No fresh culture fluid was added to those culture dishes and the least possible quantity was carried over with the organisms. The other organisms were transferred to clean slides in the usual manner. These latter ones were the source from which the non-con jugant lines used in this experiment were obtained. The process of transferring this number of animals occupied several hours. Before all the organisms had been transferred conjugation had begun in the two culture dishes. One hundred and twenty pairs of con jugants were removed from the culture dishes and numbered in order of their removal. The next morn- ing the pairs had separated. The two members of each pair were trans- ferred to the two concavities of a clean slide. The non-con jugants, one from each dish which had been transferred to slides on August 5, were transferred to clean slides until 112 non-con jugants had been trans- ferred. The number of fissions was recorded in the case of the non- conjugants. On August 7-8 the ex-conjugants completed their first two divisions, giving rise to the four lines or quadrants from each of the ex-conjugants which were to be propagated in this experiment. On August 7th and 8th the non-con jugant lines and the ex-conjugant lines were so distributed that no two non-con jugant lines or lines from the same tetrad were cultivated in the same Petri dish. This was done so that if any correlation was found between the quadrants of a tetrad or between lines of the non-con jugant population, it could not be the result of cultivation on the same slides or in the same dishes. EFFECT OF CONJUGATION OF PARAMECIUM 301 From August 6th to September 10th inclusive, each line was trans- ferred each day (except on August 8th and 10th as described below). On August 8th and August 10th the amount of work was so great that it was not possible to complete the transferring until after midnight. The lines which were not transferred until after midnight on these days were not transferred again for approximately 36 hours. On August llth all of the lines which were incomplete because of losses were dis- carded. When this was done, the number of lines retained was the maximum number that two persons could transfer once daily, using this involved technique. From August llth to the close of the experiment on September 10th, all the lines surviving were transferred daily. The actual numbers isolated at the beginning of the experiment were for the non-con jugants 112; for the ex-conjugants 405 lines or " quad- rants " derived from 105 different ex-conjugants, belonging to 58 dif- ferent pairs. The numbers were reduced by accident or death of lines, so that the actual numbers of lines available for comparison were, for the first ten days of the experiment, 66 non-con jugants, 324 ex-con- jugants ; for the first twenty days, 64 non-con jugants, 295 ex-conjugants ; for the entire period of 36 days, 46 non-conjugants, 115 ex-conjugants. During the first week following the beginning of the experiment a rather large number of deaths occurred among the non-con jugant lines. After this period there occurred a period of about three weeks during which deaths among the non-con jugant lines were rare. Many of the deaths which occurred during the early part of this period were lines that had stopped dividing during the earlier period. On the twenty-fifth day of the experiment (August 29th), the rate of mortality among the ex-con jugant lines increased rapidly. This was followed two days later by an increase in the rate of mortality among the non- conjugants. This high rate of mortality continued for nearly ten days. The occurrence of this high rate of mortality in the ex-conjugant lines beginning twenty-five days after conjugation was accompanied by a general depression in all the lines. This fact and the occurrence of two such periods in the non-con jugant lines, separated by a period of about twenty-five days, led to the suspicion that these were periods of endomixis. On September 6th many of the excess animals from the non-con iugant and ex-conjugant lines were stained and mounted for stud}'. The individuals from the ex-conjugant lines showed in many cases the conditions of late stages of endomixis. Numerous frag- ments of macronuclei were present, and in one case the organism was found to be at the climax of the endomictic process. The represen- tatives of the non-conjugant lines showed on the whole the conditions 302 DANIEL RAFFEL typical of earlier stages of endomixis. Large irregular macronuclei were found, often accompanied by large fragments. It seems clear, therefore, that the periods of high mortality were periods of endomixis : a relation which other investigators have observed. On September 10th, thirty-six days after conjugation had occurred, the experiment was discontinued. At this time 46 lines of non- conjugants and 115 lines of ex-conjugants were still in existence. 3. Results The experiment was designed to supply data mainly upon the rate of reproduction, its variability and the inheritance of the variations, in the ex-conjugants and non-conjugants. It yields also certain data on comparative mortality, which will be given first. A. Mortality A considerable number of the lines of both the non-conjugants and ex-conjugants died out during the thirty-six days of culture. The percentages surviving in each group at certain periods after the be- ginning of the experiment, are the following: After 20 days 25 days 35 days Non-con jugant lines 73.0 68.6 51.7 Ex-conjugant lines 79.2 67.9 30.8 Thus on the whole the mortality is much greater among the lines descended from the ex-conjugants. At the end 51.7 per cent of the non-conjugant lines were alive as against 30.8 per cent of the ex- conjugant lines. B. Rate of Reproduction The basic data as to comparative rate of reproduction in the non- conjugants and ex-conjugants are given in Table I. The number of fissions for both groups is reckoned from August 7th, on which day all of the ex-conjugants divided once or twice. Thus the statistics are not affected by any delay in fission due to the process of conjugation itself. Table I shows at A the number of fissions for the different lines for the 20 days of culture beginning August 7th and ending August 26th; throughout this period there were 64 lines of non-conjugants and 295 lines of ex-conjugants. the latter derived from 99 different ex- conjugants and so forming 99 tetrads. At B are shown the distribution EFFECT OF CONJUGATION OF PARAMECIUM 303 o « •5 £ <— o o £ II c v g.S te — ~c •*-• u 2 O ^— ' L m . , U" 11 tn £ 4> *X* .E -5 TJ ^ u c •£ > 4) 'to u u 4> ^ "u £ ^L e 4> "O (S •^ .2 o 4-* U V f; ^2 a 60 - C 2 tn • — 1— 1 4) rt CO TABLE i* tn — tn — t^ C _>, « rt Q o tN E "! 8 ^o -C 03 CO •a S C nj _o 'to to « tn tn 4) E |s < tn C ; p; Q •_ 'tn tn _Q C £ ^ 3 C C O — 'tn 03 tn •^ C ^ V G O _2 « tn L., bo C •• 3 .2 'a +-1 +J O •9 ^ >< 'u £ 4) tn 'n 4i 5 £-1? 4) '35 i- SUOISSI J rs »- 1 fN O O O -H 4i 41 o. H. to D 3 4 OJ n ^.^ o -f O s ^O Tf ^O £ ^ rr, r* S? tN IO ON tN t-*- ro O •-1 «4< CN C»J CN IO t— IO -H « O IO O IO Q o (N ^S10 c (ft 00 tN CN (T) c _o 'to to s t^ CN s s IO '-' u 1 s r^~ 3 s «s s - s - s -H^ 0 (N CN 00 - - 0 - Non-conjugancs Ex-conjugants Means of tetrads 5 s a. V I SUOISSI J O 0 4141 ID »O 5 o. H. V 3 4 ^S 3 C*«N o vO core O iO O 00 vo^ t-- ^^ V^ £ .00. E ^.VO 25 tN OO en 10 esf-5 Q CM 10 ~,o H) a £ T^* ^^ CO 0 •o -> to E 5 N 0 00 -o, J2 3 5 --« z 5 r^4 IO *t rH 5 - 5 5! - 5 - o - % 00 N Non-conjugants Ex-conjugants rt Q o c rt "8 c J W ~ •o o a o [X. s< u S: ( U rt 2 3 *— ' »-H »— 1 O O O O C) O -H-H-H OO OO f- IO •* T}< ^— t »-H »^ s o. H. c« V 3 4 VO -* — 1 \O CM O — i •o 00 »— * o 00 IO r^ f*> •* O i^ O "* ^— 1 ^^ IO vO ro •* r^ >-i CN O -O in o^ <^> *-> CN —i IO IO OO -H VO J^ »-H 0 10 rvl O ON IT> T-H x I/-. •* t-~ VO tN , 'is to m •* O X3 CN Q rt o 10 c» -^ ^-H S "O n »-i IO <*5 4—4 o (N r^ -H 10 ^f O CN «-H IO q >— 1 -H o q r-t «H »o CK 0 O; CN ^ >o 00 ^— i Non-conjugants Ex-conjugants Tetrads 304 DANIEL RAFFEL of the numbers of fissions for the 46 surviving non-conjugant and 115 ex-con jugant lines that survived throughout the entire period of 36 days, August 6th to September 10th. At C in Table I are shown the mean fission rates for the total period of survival, for all the lines that lived more than 10 days. Mean Rate. — As Table I shows, the mean rate of reproduction for the non-conjugant and con jugant groups did not differ greatly, although in every case the mean rate for the non-con jugants is higher by a small but significant amount. The mean fission rate for all non-conjugant lines is 1.58 ± 0.01 ; for all ex-conjugant lines, 1.48 =iz 0.01. C. Variation in Fission Rate But it is in the variation of the fission rate that the difference be- tween the non-con jugants and ex-conjugants is striking. An inspection of Table I shows at once that the variation in the ex-conjugant lines is much greater than that in the non-conjugant lines. For the first twenty days, the number of fissions in the non-conjugant lines varies from 28 to 37, a range of 10. For the ex-conjugant lines, in the same period the range is from 16 to 37, a range of 22, more than double that for the non-con jugants. For the entire 35 days, the non-conjugant lines range from 47 to 61, the ex-conjugant lines from 38 to 61. The mean daily fission rates (C, Table I) vary in the non-con jugants from 1.25 to 1.75; in the ex-conjugants from 0.85 to 1.85. The range for the former is 0.55; for the latter 1.05, or nearly double that for the non- conjugants. The fission rate for the lowest lines of ex-conjugants is far below that for the lowest non-con jugants, and the highest ex- conjugant line is above the highest non-conjugant. Conjugation within the clone has caused a wide spreading out of the fission rates ; it has produced stocks with lower, and with higher, rates than any found in the clone before it has conjugated. Computation of the standard deviations and coefficients of variation shows the same great increase in variation as a consequence of conjuga- tion. The means, standard deviations and coefficients of variation, computed from the data shown in Table I, are given in Table II. As Table II shows, the coefficient of variation of the ex-conjugant lines is for the first 20 days 158 per cent of that of the non-con jugants ; for the entire 35 days it is 187 per cent of that of the non-conjugants. For the mean daily fission rates of the different lines, the coefficient of variation for the ex-conjugants (10.14) is 139 per cent of that of the non-conjugants (7.28). EFFECT OF CONJUGATION OF PARAMECIUM 305 The comparative distribution of the fission rates of non-con jugants and ex-con jugants, as shown in Table I, are worthy of notice. In the first 20 days (A. Table I) 22 lines of ex-conjugants, or 7.4 per cent of all, show fewer fissions than any of the non-con jugants. In the entire 35 days (B, Table I), the proportion is nearly the same: 7.8 per cent of the ex-conjugant lines lie below all of the non-conjugant lines. At the opposite extreme the two sets are alike ; the highest lines have the same number of fissions in the two cases. In mean daily fission rates, 18 lines of ex-conjugants, or 5.6 per cent of all, lie below all of the non-conjugant lines, while one ex-conjugant line lies above all the non- conjugant lines. TABLE II Means, standard deviations and coefficients of variation of non-conjugant and ex-conjugant lines, for the numbers of fissions during certain periods; and for the mean daily fission rates of the different lines. Based on the data given in Table I. A. Numbers of Fissions First 20 Days Total 35 Days Mean Stan. Dev. Coef. Var. Mean Stan. Dev. Coef. Var. Non-conjug'ts Ex-conjugants 32.9±0.2 31.3±0.1 1.88 ±0.11 2.83 ±0.08 5.70±0.34 9.02 ±0.25 56.1 ±0.3 53.3±0.3 3.19±0.22 4.49±0.20 5.68±0.40 8.42 ±0.38 B. Mean Daily Fission Rates of the Different Lines Mean Stan. Dev. Coef. Var. Non-conjugants. . . Ex-conjugants. . . . 1.58 ±0.01 1.48 ± 0.01 0.12 ±0.01 0.15 ±0.00 7.28 ± 0.43 10.14 ±0.27 It is clear, therefore, that conjugation within the clone has much increased the variability of the fission rate, and that one of the factors in the increased variability is the production of a considerable number of ex-conjugant lines that have a lower fission rate than any lines among the non-conjugants. D. Variation among Quadrants Derived from a Single Ex-conjugant, Compared with Variation Between Lines Derived from Different Ex-conjugants Calkins and Gregory (1913) reached the conclusion that the varia- tion between different quadrants (the four lines derived from a single ex-conjugant) was as great as that between lines derived from diverse ex-conjugants. Lines derived from a single ex-conjugant constitute, 306 DANIEL RAFFEL of course, a clone within which conjugation has not occurred; so that according to this result, there is no increase of variation in consequence of conjugation within the clone. To test this particular matter, the variation between the different quadrants of the same tetrads (each tetrad derived from a single ex-conjugant) was compared with the variation among progeny of the different ex-con jugants. For each tetrad records of only two to four lines are available, so that the coef- ficients of variation within the tetrad are not statistically adequate, but the general result is of interest. For the number of fissions during the first 20 days of the experiment, the mean coefficient of variation for the lines constituting a single tetrad was 4.53 ; for the means of the diverse tetrads (progeny of the diverse ex-con jugants), the coef- ficient of variation was 8.32. For the average daily fission rate, the mean coefficient of variation for the lines constituting a single tetrad was 5.22; for the diverse tetrads it was 8.57. As will be seen by comparison with Table II, the mean variation within tetrads (4.53) is of a similar order to the mean variation for non-conjugant lines of a clone (5.70) (that is. to the variation within a clone in which conjugation has not occurred). On the other hand, the variation when the different tetrads are compared (8.32) is much greater, and is similar to the variation (9.02) when all the lines derived from ex-conjugants are compared. This indicates strongly that the four quadrants produced by the first two divisions of an ex-conjugant do not differ in any general way from any other products of fission of a single individual. Further, the similarity between the coefficients of variation for all ex-conjugant lines taken separately, and that for the means of the diverse tetrads, indicates that the variation among the ex-conjugant lines is due mainly to the inherent differences between the ex-conjugants. The higher variation among diverse tetrads, as compared with less variation between the quadrants belonging to the same tetrads, may be further shown by comparing the maximum differences found ( 1 ) between any two lines of the original non-conjugant population ; (2) between quadrants belonging to a single tetrad; (3) between the means of diverse tetrads ; and (4) between any two ex-conjugant lines. These comparisons are given in Table III. This table shows that the maximum difference within any of the 99 tetrads and the maximum difference between any two non-conjugant lines of the original population are very nearly identical. On the other hand, the maximum differences between any two ex-conjugant lines are only slightly greater than the maximum differences between EFFECT OF CONJUGATION OF PARAMECIUM 307 any two of the tetrads. (It is to he expected that the maximum dif- ferences between two tetrads would he slightly smaller than that be- tween the two ex-conjugant lines which differ most, since the fissions for tetrads are usually the means of two to four lines.) Thus the single tetrads do not significantly differ in their variability from the general non-conjugant population, while the variation between the dif- ferent tetrads is much greater than that within the tetrads. TABLE III Maximum Differences Bet-ween Lines Having Different Relations to Each Other uith Respect to Conjugation Total Fissions First 20 Days Average Daily Fission Rate Maximum difference between two non-con- jugant lines of the original population ... 9 0.50 Maximum difference within any tetrad .... 10 0.58 Maximum difference between two means of tetrads 17.75 0.84 Maximum difference between two ex-con- jugant lines 21 1.00 The matter may be tested further by determining whether there is correlation in fission rates between the members of the tetrads. If the different quadrants within the tetrads differ as much as do the members of different tetrads, there should be no correlation between the members of the tetrads. If, on the other hand, the different quadrants of the single tetrads show a significant correlation, this will demonstrate that such quadrants are not so unlike as are different lines of the ex-con j ugar.t population taken at random. The data for this comparison are shown in Table IV, based on the numbers of fissions during the first 20 days of the experiment. The fissions for each quadrant of each tetrad are entered as X against the fissions for each other quadrant of that same tetrad as Y. As some of the tetrads had but two surviving lines, others three or four, the total number of entries in the table is 332 pairs. Since the members of the tetrads are like variates, the correlation must be computed as for a symmetrical table in which each pair is entered twice, in reverse order (see Jennings, 1911. for the method of computation). The coefficient of correlation between the fission rates of the quad- rants of the same tetrad, obtained from this table, is very high, amount- ing to 0.854 ± 0.007. Beyond question, therefore, the quadrants de- rived from a single ex-con jugant are much more alike in their fission rates than quadrants derived from diverse ex-conjugants. 308 DANIEL RAFFEL All the four lines of evidence thus agree in showing- clearly that a population composed of different ex-conjugants of a clone has a higher variation in fission rate than do the offspring of single ex-conjugants. (1) The coefficient of variation is much higher for the ex-conjugant population than for the non-conjugant. (2) The coefficients of varia- tion for quadrants belonging to single tetrads is much less than the coefficient of variation for the means of diverse tetrads. (3) The max- imum differences between lines within tetrads are much less than the maximum differences between means of different tetrads. (4) There is a very high correlation (0.854) between the lines or quadrants de- rived from the same ex-conjugant. These four lines of evidence es- tablish firmly the fact that conjugation within a clone causes increase of variation. TABLE IV Paramecium aurelia. Correlation between total number of fissions, August 7-26, of each member of the tetrad with every other member. 18 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 16 1 1 22 1 1 2 23 1 1 24 1 1 1 3 25 2 2 26 1 1 2 27 1 3 2 1 7 28 2 2 1 1 3 2 1 1 13 29 1 1 2 3 6 7 3 2 3 28 30 1 2 6 6 8 6 3 1 1 1 35 31 1 4 2 5 7 14 5 5 7 3 4 57 32 1 1 4 5 6 7 7 10 5 2 1 49 33 1 1 6 8 10 11 10 10 1 58 34 1 3 1 9 3 16 6 3 1 43 35 3 7 3 4 1 1 19 36 1 1 1 1 1 4 1 10 37 1 1 2 11226 11 7 10 28 41 61 48 50 39 14 10 1 332 r = +0 854 ± 0.007 E. Inheritance of the Diverse Fission Rates In order to determine whether the diverse fission rates observed are hereditary, the number of fissions which occurred in each line during the first ten days after the ex-conjugants began to divide was cor- related with the number of fissions during the following ten days. These coefficients of correlation were obtained (a) for all of the non- conjugant lines which lived from August 6th-26th, (b) for all of the EFFECT OF CONJUGATION OF PARAMECIUM 309 ex-con jugant lines which lived through the same period, and (c) for the means of all the tetrads of which one or more '' quadrants " sur- vived until August 26th. The correlation tables from which these coefficients of correlation were calculated are given in Tables V, VI and VII. No correlation was found among the non-con jugant population (Table V). The coefficient of correlation obtained was -f- 0.016 ± 0.084. Thus the differences in fission rate (which are not very great, as Table V shows) are not inherited differences. TABLE V Paramecium aurelia. Non-conjugant lines. Correlation of number of fissions, August 7-16 and August 17-26. 13 14 15 16 18 19 20 13 1 1 14 1 1 2 4 15 1 1 3 3 8 16 1 1 8 6 1 17 17 1 4 2 5 2 14 18 2 2 7 3 1 15 19 1 2 3 20 1 1 2 1 4 10 22 20 7 64 r = + 0 016 ± .084 TABLE VI Paramecium aurelia. Ex-conjugant lines. Correlation of number of fissions, August 7-16 and August 17-26. 6 9 10 11 12 13 14 15 16 17 18 19 10 1 1 1 3 11 1 1 1 3 12 1 1 2 2 6 13 2 1 3 3 4 1 14 14 1 2 3 9 7 6 28 15 2 1 11 9 16 11 2 1 53 16 1 1 1 3 8 24 20 19 6 83 17 1 4 8 10 17 18 2 1 61 18 1 1 1 6 10 10 3 32 19 1 1 1 1 5 1 10 20 1 1 2 2115 5 15 38 63 78 71 14 2 295 r = + 0.327 ± 0.035 310 DANIEL RAFFEL For all the ex-conjugant lines, each taken separately (Table VI), there is a well-defined positive correlation of -f- 0.327 ± 0.035. Thus many of the diversities between these are inherited. When, however, the means of the separate tetrads are taken, and the fissions for the first ten days are tabulated against those of the second ten days (Table VII), the correlation rises to +0.651 ±0.039. In such a table, the process of averaging the different quadrants of the tetrad smoothes out in large measure the accidental differences between the diverse lines, leaving mainly the intrinsic differences, which are inherited ; hence the high coefficient of correlation. TABLE VII Paramecium aurelia. Means of tetrads. Correlation of number of fissions, August 7-16 and August 17-26. o o o o o *•>• if~- 10.50 1 11.00 1 12.00 1 13.50 1 14.00 1 1 1 1 2 14.25 1 14.33 1 14.50 1 1 14.67 1 1 14.75 1 15.00 1 121 2 1 15.25 1 1 1 15.33 1 1 1 15.50 1 1 1 1 1 15.67 1 2 1 1 15.75 1 1 1 1 16.00 1 112 1 5 1 2 16.25 1 11 1 1 16.33 1 1 16.50 2 2 1 1 1 16.67 1 1 1 1 1 16.75 1 1 17.00 1 2 2 1 17.25 1 2 1 17.33 1 1 1 17.50 1 1 1 17.67 1 1 1 112111171212 11 53636 164 103 2 5 1 1 2 r = +0.651 ±0.039 1 1 1 1 6 1 1 2 2 1 8 3 3 5 5 4 14 5 2 7 5 2 6 4 3 3 3 99 Thus in a non-con jugant population there is no indication of hered- itary diversities between the different lines, while among the lines de- EFFECT OF CONJUGATION OF PARAMECIUM 311 rived from different ex-conjugants, hereditary diversities are clearly present. A population of ex-conjugant lines consists of diverse bio- types produced by conjugation from a homogeneous clone. F. Similarity Between the Lines Derived from the Two Members of a Pair of Conjugants Jennings (1913) and Jennings and Lashley (1913) found that the lines derived from the two members of a pair of conjugants resembled each other more than do the progeny of ex-conjugants derived from different pairs. An attempt was made to determine whether this re- lation holds for the population studied in this investigation. This was done by correlating the mean number of fissions for twenty days, and the mean average daily fission rate of all lines which lived for more than ten days, of the two tetrads derived from each pair of conjugants. The coefficients of correlation which were obtained were -4- 0.102 ± 0.073 and -- 0.188 ± 0.070. Because of the small number of pairs in- volved and the large probable error obtained, these coefficients of cor- relation are of uncertain significance. Another experiment using a large number of pairs is planned, in order to investigate this matter further. SUMMARY AND DISCUSSION This paper gives an account of an investigation designed to test critically the question whether conjugation produces inherited variation within a clone of Pamnieciuui aurclia. An elaborate technique was devised and carried through, to prevent the occurrence of environmental diversities among the lines of descent tested : a synthetic culture fluid was employed ; pure cultures of food organisms used, and the glassware standardized to the highest possible degree. Cultivated in this way, ex-conjugant lines of descent were compared with non-con jugant lines from the same parent clone, with respect to the rates of fission. The results are : (1) Conjugation greatly increased the variation in fission rate. The population composed of lines descended from ex-conjugants showed a much greater range of variation and a much greater coefficient of variation than did the population derived from non-conjugants. The variation was extended by conjugation mainly in the direction of lowered fission rate. A considerable proportion of the ex-conjugant lines had a lower daily fission rate than any of the non-conjugant lines. Others had as high a fission rate as the non-conjugants (see Table I). (2) The four quadrants derived from the first two divisions of single ex-conjugants showed when compared only such variation as is 312 DANIEL RAFFEL found in non-con jugants ; not at all such extreme variations as are found between lines derived from diverse ex-conjugants. The four quadrants from a single ex-con jugant are highly correlated in their fission rates, showing a correlation coefficient of 0.854 dr 0.007. Such quadrants derived from a single ex- con jugant are thus much more alike in their fission rates than are lines derived from diverse ex- conjugants. There is no indication that the first two fissions occurring after conjugation have any effect in segregating diverse lines, or that they differ in their effects from any other fissions. (3) The diverse fission rates of lines or populations derived from different ex-conjugants are in large measure inherited, while the differ- ing rates of non-conjugant lines are not inherited. The work therefore leads to the following conclusions: Conjugation within a clone of Parameciuni aurelia produces diverse biotypes, having different inherited fission rates. The fissions of a single ex-conjugant do not give origin to diverse biotypes ; this is as true of the first two fissions after conjugation as of later fissions. LITERATURE CITED ADOLF, E. F., 1929. The Regulation of Adult Body Size in the Protozoan Colpoda. Jour. Expcr. Zool., 53: 269. BELAR, K., 1928. Untersuchung der Protozoen. Methodik iviss. BloL, 1: 735. CALKINS, G. N., AND GREGORY, L. H., 1913. Variations in the Progeny of a Single Ex-conjugant of Paramecium caudatnm. Jour. Ex per. Zool., 15: 467.^ JENNINGS, H. S., 1911. Computing Correlation in Cases where Symmetrical Tables are Commonly Used. Am. Nat., 45: 123. JENNINGS, H. S., 1913. The Effect of Conjugation in Paramecium. Jour. Exper. Zool., 14: 279. JENNINGS, H. S., 1916. Heredity, Variation and the Results of Selection in the Uniparental Reproduction of Difflugia corona. Genetics, 1: 407. JENNINGS, H. S., 1929. Genetics of the Protozoa. Bibliographic a Gcnctica, 5: 105. JENNINGS, H. S., AND LASHLEY, K. S., 1913. Biparental Inheritance and the Question of Sexuality in Paramecium. Jour. Ex per. Zool., 14: 393. LUNTZ, A., 1926. Untersuchungen iiber den Generationwechsel der Rotatorien. Biol Zcntralb., 46: 233. PARPART, A. K., 1928. The Bacteriological Sterilization of Paramecium. Biol. Bull.. 55: 113. PRINGSHEIM, E. G.. 1928. Algenreinkulturen. Eine Liste der Stamme, welche auf Wunsch abgegeben wurden. Arch. f. Protisk., 63: 255. A MECHANISM OF INTAKE AND EXPULSION OF COL- ORED FLUIDS BY THE LATERAL LINE CANALS AS SEEN EXPERIMENTALLY IN THE GOLDFISH (CARASSIUS AURATUS) GEORGE MILTON SMITH ANATOMICAL LABORATORY, SCHOOL OF MEDICINE, YALE UNIVERSITY In the course of studies of lateral line canals of the goldfish, it seemed advisable to observe possible reactions of the canals of the lateral line organs to absorption of coloring substances held in sus- pension by water. To accomplish this purpose, goldfishes were im- mersed in various weak solutions containing lampblack, India ink, ver- milion and Berlin blue, and allowed to live over periods of time vary- ing from a week to two months. From time to time the fishes were examined and it was found that actually small amounts of these col- oring substances had been taken up by the lateral line canals of the head or trunk. Such small patches of absorbed pigment occasion- ally found caught in the lumen of the lateral line canals gave, how- ever, unsatisfactory evidence of any mechanism of absorption or ex- pulsion of fluids into the system of canals. Finally, by using more highly concentrated solutions of some of these same substances, in which the fishes, temporarily, were allowed to swim, a very striking outline of the lateral line canal system filled with coloring substance was obtained ; and there was also offered an opportunity of directly observing the intake and expulsion of these colored fluids through the pores dis- tributed along the canal system. An illustrative experiment is as follows : goldfish, length 5 cm. from tip of snout to base of tail, whitish color. Solution : India ink 20 cc., water 200 cc., temperature 20° C. Preliminary examination of fish showed normal-looking lateral line canals of head and of body. The fish was placed in the India ink solution for thirty seconds, rinsed in water, and changed to a shallow dish of water for examination under the dissecting microscope. The canal system of each side of the head was sharply outlined in black, the supra and infraorbital, the hyo- mandibular and the supra-temporal canals were deeply injected, and appeared as sharp black lines. The absorbed India ink extended to about one-fourth of the adjacent region of the lateral line of the. trunk. After the lapse of over one half minute, there was noted black coloring 313 314 GEORGE MILTON SMITH matter, stringy as if mixed with mucus, first at one and then another of the pores of either side of the head. The canals a few minutes later, began to assume here and there a clearer, grayish appearance. Bits of coloring matter were wiped away with cotton swabs from the pores and were followed by fresh extrusion of delicate shreds of darkly colored mucus. The fish was now allowed to swim in a large jar of clear water at room temperature. At the end of 10 minutes the hyomandibular, supra and infraorbital canals were clear of India ink. At the end of 15 minutes the lateral canals of the trunk had nearly cleared. At the completion of 30 minutes, only the supra-temporal canal showed the remains of India ink in the form of a faint gray line. The supra-temporal canal was cleared of the remaining India ink when 35 minutes had elapsed, so that all canals now contained a clear, limpid, normal-looking mucus with no evidence of previous staining (Figs. 1-6). The immediate penetration of colored fluids into the canal system may be observed under a dissecting microscope by applying drops of India ink, by means of a finely drawn pipette, over the pores of any part of the canal system of the head or trunk. There follows a rapid intermingling of India ink with the mucous contents of the canals and a consequent spread of India ink along the canals in either direction from the point of application of ink at the pores of the surface. If the application of India ink is continued, adjacent communicating branches of the canals soon become injected with the black coloring- substance. When the application of India ink is discontinued, expulsion of the India ink, mixed with mucus, begins and can be seen leaving the canals at the. pores which furnish communicating passages between the canals and surface. Elimination of India ink, mingled with mucus, continues until the canals are entirely cleared and appear normal. It is essential to employ healthy, active goldfishes for experiments of this character. Dying fish take up coloring substances in an ir- regular manner. It was found that in the dead goldfish a penetration of coloring substances occurred to some extent. This seemed to be less intense and more irregular and patchy than in the living fish and, of course, there was not the immediate elimination of coloring sub- stance by the flow of mucus from the canals. At times no penetration of the coloring substances occurred in the case of the dead fish, possibly on account of the lack of mucus in the canals. Experiments such as these mentioned above were repeated many times in different ways with evidence of intake and expulsion whenever coloring substances were brought into contact with the lateral line canals. This evidence occurred also in the experimentally-blinded fish and in fishes with nares destroyed by cautery. INTAKE AND EXPULSION OF COLORED FLUIDS 315 " D A 5 / ^ 4. 5. 6. FIGS. 1, 2, 3, 4, 5 AND 6. Diagrammatic drawings of lateral line canals of goldfish as seen from above, illustrating intake and expulsion of a solution of India ink, 20 cc. ; water, 200 cc. Fig. 1, A. Lateral line canal of trunk; B. C. D, and E, supra-temporal, hyomandibular, infraorbital and supraorbital canals, re- spectively, previous to intake. Fig. 2. Filling of canals, after 30 seconds of im- mersion in India ink solution indicated by black dots in canals. Fig. 3. Clearing of supraorbital, infraorbital, and hyomandibular canals 10 minutes after fish was placed in clear water. Figs. 4 and 5 show progress of clearing after 15 and 30 minutes respectively. Fig. 6 shows canal system entirely cleared after 35 minutes. 21 316 GEORGE MILTON SMITH India ink and Berlin blue acted as coloring agents most favorable for the experiments. Vermilion in suspension in water was useful for the studies over longer periods of time when certain symmetrical dis- tributions of absorbed coloring matter occurred. Lampblack was not found satisfactory. It rarely gained entrance into the canals, possibly because the conglomerate and adherent particles formed were too large to permit of entrance into the pores of the canal system. The complete elimination of absorbed coloring substances from the lateral canals varied in different animals over a considerable range of time. Such a difference in the elimination of India ink from the lateral canals was noted in the following experiments, carried on simul- taneously with two fishes of different size: Two goldfishes, A and B, 4l/2 cm. and 7 cm. respectively in length ; fluid for immersion : India ink, 100 cc. ; water, 500 cc. ; temperature, 20° C. 1 :27 P.M. Both fishes placed in India ink solution. 1 :30 P.M. Both removed and examined. In both, all branches of lateral line system of head were colored black. The lateral lines of the 'trunk were black in the proximal or cephalic third in both fishes. 1 :31 P.M. Placed in tanks of fresh water. Both fishes, from now on, examined under the binocular microscope every 10 minutes. 1 :41 P.M. In both fishes the lateral canals of the body were cleared of black color, and in both the nasal parts of the supraorbital canals and the submaxillary parts of the hyomandibular canals were clear. 2:11 P.M. Clearing of canals had proceeded to a point where A showed only a moderate amount of India ink in the supra- temporal canal ; and B showed a very slight staining of the supra- temporal, both infraorbitals and the posterior part of the hyo- mandibular on the right and left sides. 2:31 P.M. Fish A showed only a slight amount of staining in the occipital canal, while fish B had all canals perfectly cleared and translucent. 2:51 P.M. Fish A had canals now entirely cleared of India ink, hav- ing taken twenty minutes longer than fish B. Apparently, with a change in environment, the lateral canals of the goldfish were placed in operation as forms of testing apparatus. If the fish was changed from one colored solution to another of different color, directly, or with an opportunity of cleaning the canals in fresh water, the lateral line canals took up the colored fluid of the new environment. INTAKE AND EXPULSION OF COLORED FLUIDS 317 An experiment illustrating a change involving intake and expulsion of three different colored solutions is the following-: o Goldfish, 514 cm. in length ; markings : whitish with slight black pigment above eyes. 6:41 P.M. Placed in a dish containing vermilion, 20 grams; water, 500 cc. 6:54 P.M. Left supraorbital canal was brilliantly injected with ver- milion and there was a small amount of vermilion in the right hyomandibular canal near the angle of the jaw. 7 :00 P.M. Same distribution of vermilion as at previous reading. Fish changed to clear water. 7:15 P.M. Left supraorbital canal clear of vermilion. Minute plug of vermilion in right submaxillary region. 7:15 P.M. Fish placed in a dish containing Berlin blue. 5 grams; water, 500 cc. 7:18 P.M. After 3 minutes taken out of Berlin blue solution. Both infraorbital canal and submaxillary parts of hyomandibular canal showed as bright blue. 7 :19 P.M. Placed in fresh water. 7 :29 P.M. Canals cleared of all traces of Berlin blue while in fresh water for 10 minutes. 7 :30 P.M. Placed in a solution containing India ink, 100 cc. ; water, 500 cc. for one minute. 7:31 P.M. Removed from India ink solution (1 minute). All lateral line canals of head and side were black. 7:31 P.M. Fish placed in clear water. 8:40 P.M. Canals of head and trunk now appeared completely cleared of India ink, the lateral line canals having absorbed and ex- pelled three different colored solutions in the space of two hours. Similar results were noted in another fish (5 cm. in length; whitish silvery color) which had been placed, six weeks previously, in a solu- tion of Berlin blue, 5 grams ; water. 2500 cc. With all canals deeply stained blue this fish was changed directly to a solution of India ink 20 cc. ; water, 200 cc. After one minute in India ink the nasal parts of the supraorbital canals and the anterior regions of the hyomandibular canals on both sides were black and readily distinguishable from the adjacent blue. Returned to the same Berlin blue solution in which it had been swimming for six weeks, the India ink in the above-mentioned canals could no longer be recognized at the expiration of 20 minutes ; all the canals of the head and trunk were again stained a bright blue. When goldfishes were kept in an environment of colored fluid for 318 GEORGE MILTON SMITH longer periods of time, such as one week to two months, the absorbed coloring matter in the lateral canal system varied from a condition of complete filling of the canals to one where only certain branches were incompletely filled. Occasionally no coloring appeared in any of the canals. In other words, the mucous secretion of the canals may clear away previously absorbed coloring substance and keep the canals partly or completely clear in spite of the fact that the fish is living in a colored solution. In the following experiment a goldfish was allowed to remain for one month in a solution of Berlin blue. When placed in fresh water at the expiration of that time, clearing of the canal system seemed unusually long (3 hours and 20 minutes.) Tested immediately after- wards for elimination of India ink, this substance was also slowly ex- pelled (3 hours and 8 minutes). The time of intake did not seem to be affected. Experiment: 11/30/29. Goldfish, whitish silver in color; 4 cm. in length was placed in a jar containing Berlin blue, 5 grams; water, 2500 cc. The lateral line canals of the head and body were stained a vivid blue in 30 seconds. Examined from time to time during the first three weeks, the fish showed variations in distribution of blue in different branches. Examined daily for the last seven days of a thirty- day period, all lateral line canals of head and trunk were intensely stained with blue. 12/30/29. After a month immersed in Berlin blue solution, with all the canals deeply stained blue, the fish was placed in clear water. In 3 hours and 20 minutes, all the canals of the head and body were clear of blue color. Changing the environment now to one of India ink (20 cc. ; water, 200 cc., the canals became quickly and completely stained black in 30 seconds. Returned to clear water, the canals were freed of India ink in 3 hours and 8 minutes. Returned finally to the original Berlin blue solution where the fish had lived previous to the present experiment for a period of one month, the canals took up an intense blue stain in 30 seconds. In goldfishes kept in a solution containing vermilion, the intake of red-pigmented particles was more leisurely performed, appearing in small patches in the course of the first twenty- four hours. Two fishes which were examined from day to day, during a period of two months, showed various branches irregularly filled with vermilion mixed with mucus contained in the canals. Not infrequently the ab- sorbed vermilion was bilateral in distribution and symmetrically ar- ranged in the different canals of the head and trunk. This symmetrical INTAKE AND EXPULSION OF COLORED FLUIDS 319 distribution of vermilion in the lateral canal system of a fish kept in a solution of vermilion, 10 grams ; water, 3000 cc. for two months is indicated in the accompanying figures (7-11) based on daily ob- servation for 5 days when a symmetrical pattern of intake happened to be present. DISCUSSION AND SUMMARY It is not essential for the present purpose to state in detail the his- torical data of the lateral line canals and organs. It may not be amiss, I 8. a 10. FIGS. 7, 8, 9, 10 AND 11. Diagrammatic representation of canal system of goldfish kept in a solution of vermilion, 10 grains ; water, 3000 cc., for a period of two months. Absorbed pigment, although usually irregularly distributed in lateral line canals, appeared symmetrically arranged in this instance during a period of five consecutive days. The canals dotted in black contained absorbed vermilion. 320 GEORGE MILTON SMITH however, to recall that the presence of lateral canals in fishes, as cited by Fuchs (1895) was known and described by at least three anatomists of the seventeenth century, — Nicolas Stenonis (1664), Lorenzini (1678), Rivinus (1687). The lateral line canals were generally regarded as mucous canals or Schleimkanale until the time of Leydig (1850-51), whose careful histological studies of the contained end organs led him to the con- clusion that the lateral organs were sensory organs. Since that time a vast amount of data has accumulated as the result of the work of many investigators, and reviews on the subject appear in connection with the important works of Ayers (1892), Fuchs (1895), Allis (1904) and Johnson (1917). From the functional standpoint, Lee (1898) has stated that there has been no concensus of opinion as to the exact function or mode of action of the lateral line sensory organs. His own conclusions were that the lateral lines have a sensory function which is closely connected with the motor organs and is analagous to the function of the ear, and hence they may 'be regarded as organs of equilibrium. Schulze (1870) had suggested earlier that this sense perception was possibly an appreciation of mass movement of the water or of movement of the body through the water; whereas Fuchs (1895), from carefully conducted researches, was led to the conclusion that the lateral line sensory organs gave sensory impressions of changes in hydrostatic pressure. Hofer ( 1908) believes from his studies that the lateral line organs are stimulated alone by weak currents of water. Parker (1918), in the course of researches conducted on the auditory apparatus, finds that the lateral lines respond to water vibrations which are slower than those which affect the auditory mechanism. Recent views of the lateral line sense organs place their function, according to Herrick (1927) intermediate between tactile and auditory organs. Their nerve supply, he states, is from the lateralis roots of the seventh and tenth cranial nerves. He points out the intimate associa- tion with the eighth nerve supplying the internal ear, and the termination of these nerves in the acoustico-lateral area of the medulla. According to Herrick (1927), the structure of the end organs of the lateral line system and those of the human ear are of the same type. From the experiments carried out on the goldfish cited in the present communication, it would seem that there is in the lateral line canals of the goldfish, demonstrable by the experimental use of colored fluid, a mechanism of intake and expulsion of fluids. The intake is rapid and seems to vary from a few seconds to a few minutes. The elim- ination from the canals is slower and more deliberate, taking from fif- teen minutes to one hour or more. Colored fluids in passing through INTAKE AND EXPULSION OF COLORED FLUIDS 321 the pores of the lateral canals mix rapidly with mucus existing in the canals, the mucus acting possibly as a diluent. The discharge of col- oring substance from the canals is effected by an outward discharge of mucus through the pores of the canals. The mixture of colored material and mucus appears in the form of delicate colored shreds or plugs as they are expelled. These colored mucous shreds quickly wash away in surrounding water. Therefore, experiments, such as these described, where lateral line canals take up and expel different coloring substances in suspension when the fish is changed to solutions of different color, suggest that the lateral canals of the goldfish function, in part, at least, as sensory testing mechanisms for chemical or physical changes in environment ; and that the ready flow of mucus from the canals furnishes an efficient means of eliminating fluids that have been tested by the end organs of the canal system. LITERATURE CITED ALLIS, E. P., 1904. The Latero-Sensory Canals and Related Bones in Fishes. Intcrnat. Monat. Anat. u. Phys., 21: 401. AYERS, H., 1892. Vertebrate Cephalogenesis. II. A Contribution to the Morph- ology of the Vertebrate Ear, with a Reconsideration of its Functions. Jour. Morph., 6: 1. FUCHS, S., 1895. Ueber die Function der unter der Haut liegenden Canalsysteme bei den Selachiern. Plugcr's Arch., 59: 454. HERRICK, C. J., 1927. An Introduction to Neurology. (See pages 124 and 233.) HOFER, BRUNO, 1908. Studien iiber die Hautsinnesorgane der Fische. Berichte aus der Kgl. Bayenschen Biologischen Versuchsstation in Miinchen, Vol. 1, p. 115. JOHNSON, S. E., 1917. Structure and Development of the Sense Organs of the Lateral Canal System of Selachians (Mustelus canis and Squalus acan- thias). Jour. Com par. New., 28: 1. LEE, F. S., 1898. The Functions of the Ear and the Lateral Line in -Fishes. Am. Jour. Physiol., 1: 128. LEYDIG, F., 1850. Ueber die Schleimkanale der Knochenfische. Arch. f. Anat. Physiol. u. Wis. Medicin.., p. 171. LEYDIG, F., 1851. Ueber die Nervenknopfe in den Schleimkanalen von Lepi- doleprus, Umbrina und Corvina. Arch. f. Anat. Physiol. u. Wis. Medicin. Mcd., p. 235. LORENZINI, S., 1678. Observazioni intorno alle Torpedini fatte da Stephano Lorenzini Fiorentioni e dedicate al serenissimo Ferdinando III Principe di Toscanio Firenze. Quoted by Fuchs, S. PARKER, G. H., 1904. The Function of the Lateral Line Organs in Fishes. Bull, of Bur. Fisheries, 24: 183. RIVINUS, 1687. Observatio anatomic circa poros in piscium cute notandos. acta erudit. Lipsiae. (Quoted by Fuchs, S.) SCHULZE, F. E., 1870. Ueber die Sinnesorgane der Seitenlinie bei Fischen und Amphibien. Arch. f. mikr. anat., 6: 62. STENONIS, NICHOLAS, 1664. De musculis et glandulis observationem specimen cum epistolis duabus anatomicis. Amstelodami. p. 54. (Quoted by Fuchs, S.) RAT VAS DEFERENS CYTOLOGY AS A TESTIS HORMONE INDICATOR AND THE PREVENTION OF CASTRATION CHANGES BY TESTIS EXTRACT INJECTIONS1 SUP VATNA HULL ZOOLOGICAL LABORATORY, THE UNIVERSITY OF CHICAGO I. INTRODUCTION The cytological and histological changes in the prostate glands and the seminal vesicles of the rat following castration have been worked out by Moore, Price and Gallagher (1930) and Moore, Hughes and Gallagher (1930) respectively, and it was found that there are some dependable criteria, by which one can tell whether the sex hormone is present or absent. It is desirable to know what other organs may be affected and if the changes will be consistent enough to serve as a sex hormone indicator. This paper will deal with the study of the vas deferens of the white rat in its normal state and after different periods of castration, and the effects of subcutaneous injections of extracts from the testicle upon the castrate condition. This study was suggested to me by Prof. Carl R. Moore as another unit in the program of sex studies now being carried on in the De- partments of Zoology and of Physiological Chemistry and Pharmacol- ogy. I am grateful to him for advice and assistance given to me throughout the course of the work. I will show in this paper that the structure of the vas deferens is controlled by the internal secretion of the testes and furthermore that this control can be maintained in the castrated animals by means of subcutaneous injections of the ex- tracts of bull testes. A preliminary account of the findings has already appeared (Moore, Vatna and Gallagher, 1930). The numbered prep- arations of bull testis extract were supplied in strengths unknown to us until after assay. They were prepared by Mr. T. F. Gallagher under the direction of Professor F. C. Koch in the Department of Physi- ological Chemistry and Pharmacology, to both of whom is expressed a debt of gratitude. The earlier papers from these laboratories (McGee ; McGee, Juhn and Domm ; Moore and McGee ; Moore and Gallagher ; Moore, Price, Hughes, Gallagher ; Gallagher and Koch ; Moore, Gal- 1 This investigation has been aided by a grant from the committee on research in problems of sex of the National Research Council ; grant administered by Prof. F. R. Lillie. 322 RAT VAS DEFERENS CYTOLOGY 323 laghcr and Koch) have presented the biological test methods previously employed, and the methods of hormone extraction, and the reader is referred to them for details. Other laboratories have recently reported positive results from at- tempted hormone extraction from the testis of various mammals and the urine of men (Martins and Rocha e Silva. 1929; Loewe and Voss, 1929: Funk, Harrow and Lejwa, 1929, 1930). II. MATERIAL AND METHOD White rats were used in this experiment. The stud)' involves the examination of the vas deferens from about thirty normal animals of varying ages, thirty-five castrated, and fifty castrated injected animals. Castration was performed through a mid- ventral abdominal incision. In some cases the body of the epididymis was cut through, leaving the tail of the epididymis attached to the vas deferens. With others the entire epididymis was removed with the testis. The proximal, or urethral end of the vas deferens presents a struc- ture that shows more marked effects from castration than does the distal, or epididymal end, hence the proximal two-thirds of this re- productive tube has usually been the part that has received the greatest attention. The tissues were fixed for histological study in Benin's fluid and Zenker formol mixture. Bouin's fluid was found to be the better of the two, and therefore was used throughout the work. The sections were cut at 4p. thickness and were stained in such mixtures as Delafield's haematoxylin with eosin as a counter stain, or iron haematoxylin, or Mallory's triple stain. Mann's osmo-sublimate fixative was also used to demonstrate the Golgi apparatus. The technique employed was that of Ludford's (1925, 1926) modification of the Mann-Kopsch method. Briefly, the vas was cut into small pieces of about three mm. in length and fixed in a freshly prepared mixture of an equal volume of one per cent osmic acid in distilled water and a saturated solution of mercuric chloride in normal salt solution, for about twenty hours. The tissue was then washed in two changes of distilled water for about thirty minutes, and placed in two per cent osmic acid solution in quantities sufficient to cover it, after which it was placed in the dark at room temperature for about seven days. At this time the osmic acid solution was discarded, the tissue washed once in distilled water, and transferred in distilled water, to an oven at about 35° C. for four days. The tissue was next washed in running tap water over night, and then put through the ordinary his- tological procedures, such as dehydration, clearing, imbedding, and sec- 324 SUP VATNA tioning. The sections were bleached in a solution of hydrogen peroxide in 95 per cent alcohol. TIL THE STRUCTURE OF THE NORMAL VAS DEFERENS The vas deferens of the rat is more or less spindle-shaped in ex- ternal form. Between the urethral end and the middle of the vas, is an elongated swollen region, from the distal end of which the tube tapers toward the epididymis and from the proximal end toward the urethra. In the normal, the vas is always full of spermatozoa. This can be detected with the naked eye because of the milky white streak which is present in the middle throughout its length. The swollen region is especially distended by spermatozoa. The normal vas deferens has been studied both from animals sacri- ficed for the purpose and from animals after unilateral castration of varying periods. The latter type has been used in order to see whether the spermatozoan content in any way modified the structure of the epithelial lining. In the mammals there is no question now as to the ability of one testis to keep up the normal state of the accessory repro- ductive organs. The vas deferens from the latter group is preferred for the sake of comparison, although there is no essential difference between the normal histology of the vas from the two sources mentioned, except when the spermatozoa have collected in an unusually large quan- tity. Then the height of the epithelium may be slightly lowered due to the distention of the lumen in general, but the arrangement of the nuclei of the epithelium is not at all disturbed. The cilia may be some- what distorted from normal shape. However, in all cases examined, their appearance is decidedly not that of a castrate type. The vas deferens of most mammals, as generally known, is not cil- iated ; some species, however, are well furnished with cilia. The mouse and the rat belong to the latter group. The word " cilia " in connection with the vas deferens, Benoit ( 1926) thought should be " stereocils " or " poils," on account of their non-vibratile nature. The short term " cilia " will be used in this paper to mean " cilia-like " structures. The histology of the vas deferens is a very simple one. The tube consists of three easily distinguishable layers, the outside muscular layer, the mucous, and the epithelial or inner layer. The outer coat covered by peritoneum consists of longitudinal and circular muscle layers, and makes up approximately four-fifths of the thickness of the walls of the tube. Internal to the muscular layer is the so-called mucous layer com- posed primarily of connective tissue-like cells and blood vessels. This layer is sensitive to operative manipulation which is in no way related RAT VAS DEFERENS CYTOLOGY 325 to hormone control. From an unoperated animal, it is narrow and the cells are more or less tightly packed together, whereas the vas from a unilaterally castrated animal has a much broader mucous layer and the cells are rather scattered. The internal epithelial layer bordering the small lumen is definitely separated from the mucous layer by a very thin cord of about one or two cells in thickness. This cord of cells forms the outline of the basal part of the epithelium, and will be re- ferred to in this paper as the " basement-cell layer." The epithelial layer is composed of tall columnar cells, resting upon a distinct basement-cell layer, and the free end of the cell is covered by a heavy mass of cilia-like structures projecting into the lumen. The nuclei of the cells are generally oval in shape and variable in chromatin constituents. They vary slightly in position in the vasa of different animals, but in any one animal they occupy the same relative position in all of the cells. Thus the nuclei are seen to form a definite layer paralleling the basement-cell layer (see Figs. 1 and 5). At many places in a section one observes a few nuclei that seem to be differentiating from the basement cells, with others present above the nuclear layer apparently migrating toward the lumen. In the lumen itself, one often finds a group of epithelial cells in various stages of degeneration. These findings suggest a series of changes in the normal vas deferens, wherein cells are added to the epithelium from the basement-cell layer, and at the same time others having functioned actively for a certain time, are thrown off into the lumen, where degeneration occurs. Between the nucleus and the ciliated border of each cell, the cyto- plasm is of a condensed homogeneous granular character, whereas that basal to the nucleus is much less dense and is fibrillar in character. The difference between the distal and proximal ends of the epithelial cell is very marked. The finely granulated material in the distal portion is believed to be made up of secretory products (Myers-\Yard, 1897: Benoit, 1920). Benoit (1926) by the use of a special technique found certain definite lipoid bodies which he called " parasomes " in the epi- thelial cells of the vas deferens of the mouse and rat. These " para- somes " were believed to be the product of protoplasmic differentiation. They first appear when the animals are about fifteen days old, and in the adult they are found scattered throughout the cell. He suggests that the " parasomes " normally undergo some sort of dissolution and contribute to the formation of a liquid product of secretion. The in- vestigation reported here has not involved a study of these " para- somes." There is always a small amount of secretion present in the lumen of the vas in normal animals. This secretion forms a finely granular homogeneous mass and stains with eosin. 326 SUP VATNA The vas deferens prepared by the Mann-Kopsch technique reveals definite, well-formed Golgi bodies in the epithelial cells. The Golgi bodies are located approximately midway between the nucleus and the lumen end of the cell and are of the reticular type. Their charac- teristic shape is shown most clearly in slightly under-impregnated sec- tions, in which case the threads making up the reticulum will be black- ened only on the outside, thus giving a double-lined appearance. The size of the Golgi bodies in the normal is about that of the nucleus, though they may be somewhat larger in some cases. IV. CHANGES IN THE VAS DEFERENS FOLLOWING CASTRATION In order to determine whether the vas deferens was affected by castration. I have studied preparations from animals in a closely graded series from three days up to seven months after testis removal. The tissue prepared from animals sacrificed at 3, 5, 6, 7, and 9 days after castration is essentially normal. The gross size, relative thickness of the layers, the character of the epithelium and the condition of the cilia do not differ markedly from the normals. The Golgi bodies, however, begin to show some differences for they become smaller in comparison to the size of the nuclei, and. more striking, the reticulum breaks up to form a group of crooked rods or coarse granules. At 10 and 15 days after castration, the gross size as well as the histological structure of the vas of some animals shows a decided change, characteristic of a longer time castrate. The vas deferens be- comes smaller and the epithelium may be typical of a 20-day castrate. However, other animals castrated for this period may retain essentially the normal condition in the vas. The Golgi bodies after ten to fifteen days of castration have under- gone a marked fragmentation. The portion of the cells where the Golgi bodies are normally found will be seen to be full of scattered osmiophilic granules. These granules may clump together, but the structure does not suggest a normal Golgi apparatus. Twenty clays after testis removal the vas deferens characteristically shows the effects of castration. This period is of special importance inasmuch as many of the effects of testis extract injection have been studied for this period of time after operation. The size of the vas is now noticeably smaller, due to the degenera- tion of the muscular layer, which normally makes up almost the whole thickness of the tube. The morphological structure of the mucous layer has no constant bearing upon castration. The most apparent changes occur in the epithelial layer. The ab- RAT VAS DEFERKNS CYTOLOGY 327 solute height of the epithelium from the basement-cell layer to the luminal border is slightly reduced. The cell walls are no longer clearly visible, and the nuclei instead of forming a well-defined layer paralleling the basement-cell layer are now more closely aggregated in an irregular distribution giving the appearance of pseudostratification. The epi- thelium now appears as a syncytium. The nuclei show little, if any, reduction in size, but because of the reduction in the amount of cytoplasm in the cells, they now lie close to the basement-cell layer. The cytoplasm between the nucleus and the lumen end of the cell is likewise greatly reduced. The ciliary border of the epithelium also differs greatly from the normal. The cilia are in most cases completely absent from the vas deferens of 20-day castrate animals (see Fig. 6). In a few others they may still be present but greatly reduced both in number and length and present often an interwoven, irregularly twisted condition. The secretion found in the lumen does not seem to be changed in quality, but is much reduced in quantity following castration. However, even after long-time castration, there is always a small amount of secre- tion present. Benoit (1926) reports from his study on mice and rats that the parasomes, the bodies responsible for the formation of secretory products, disappear completely after thirty days of castration. From our own study on the rat, we have been unable to confirm the statement regarding the absolute cessation of secretion. The Golgi bodies too are decidedly different from the normal at this period of castration. Their gross size, relative to the si/e of the nucleus, is very much reduced. The former reticular arrangement has usually changed to a granular one, and these granules sometimes form an irregular cap over the end of the nucleus. The typical condition of the twenty-day castrate animal given above holds for the majority of animals castrated for this period, but oc- casionally slightly different conditions may be encountered. A few apparently more resistant animals have suffered less from castration than others and appear almost normal, except for a lower epithelium and a slight crowding and displacement of the nuclei. The typical degenerate condition of the vas deferens at twenty days after testis removal represents, with some exceptions, essentially the condition that is to be found in later castrates. The series which I have studied includes animals castrated for periods of 21, 25. 30, 33, 40, 50, 60, 80, 110, 150, and 210 clays. As the age of castration in- creases there is little, if any, increase in the amount of involution. Fig. 7 shows the condition of the vas deferens in an animal castrated for two hundred and ten days and in comparison with the normal (see 328 SUP VATNA Fig. 5), clearly shows the absence of a ciliary border of the epithelium, the lowered height of this layer, the apparent stratification of the nuclei, the involuted mucous layer and the reduced muscular layer. Fig. 2 in comparison with Fig. 1 demonstrates clearly the difference between a five-month castrated vas and the normal. It is apparent, therefore, that castration leads to a marked degenera- tion of the vas deferens. Since this influence is to be attributed to the endocrine influence of the testis rather than to the gametogenetic influence, we have in this degeneration a means of testing the effective- ness of preparations of testicular extracts. If testis removal is fol- lowed by the injection of the testis extracts and the vas deferens re- mains in a normal condition, it will be apparent that the extracts ex- ercise an influence similar to that of the internal secretion of the testis. My observations on this point are described in the following section. V. THE EFFECTS OF TESTIS EXTRACT INJECTIONS In the preceding section, definite changes have been described for the various parts of the vas deferens. These are : Decrease in gross size, involution of the muscular layer, slight lowering of the epithelium, the syncytial character of the cells, pseudostratified appearance of the nuclei, loss of the cilia, and reduction in size of the Golgi bodies, with accompanying fragmentation. Early work from these laboratories supplies proof that the active principle of the internal secretion of the testes is contained in suitably prepared lipoid extracts of the glands of the bull. In the course of this study, many samples of the extracts have been used for injection on over fifty castrated males. Some of these were less potent than others, depending on the preparation methods and the dilution of the samples. The results, therefore, are of a wide range. The typical positive cases to be described were chosen from animals having re- ceived appropriate strength of the hormone solutions. Since twenty days was found to be the period at which the degen- erative changes of the epithelium reach their height, it was selected as minimal length of time for testing the hormone extracts. Animals have been injected daily immediately after castration in order to see whether the effects of testis removal could be indefinitely postponed. In addition to this procedure, other animals have been castrated and permitted to develop the castration condition with subsequent injection to test the capability of the extracts to restore the degenerate to a normal condition. This latter procedure has been followed in the case of animals castrated as adults as well as those castrated before puberty. RAT VAS DEFERENS CYTOLOGY 329 p 0 Cross-sections of rat vas deferens. Photomicrographs of Bouin-haematoxylin preparations. About 50 X before reduction. (All photomicrographs were made by Mr. Kenji Toda.) 1. From a normal animal. 2. From a five month castrate. 3. From a 110-day prepubertal castrate. 4. From a 110-day prepubertal castrate, given forty daily injections of bull testis extract. 330 SUP VATNA 1. The Maintenance Experiment In this series, the animals were given twenty daily injections, or more in some cases, immediately after castration to maintain the normal condition. • The histological study of such injected castrates shows a normal structure of the vas. The epithelium is simple columnar and abundantly supplied with cilia, and the nuclei have the simple regular arrangement, typical of the normal. The Golgi bodies are approximately normal. 2. The Repair Experiment a. Prepubertal castrates Two series of prepubertally castrated animals have been utilised for injection. The first group of four animals was castrated at four- teen days after birth and the second group of five animals was cas- trated at forty days of age. The second one is more instructive, hence it will be described in detail as to the procedures. Five animals of the same litter were castrated at forty days after birth, and at one hundred days after castration, four animals were injected with the testis extracts No. 8922, — one-half cc. being injected daily. When the in- jections had been given for ten days, one of the four injected animals was killed, and at the same time the uninjected control was also killed. At twenty days after the injections, one of the three was killed. The next one was killed after having received thirty daily injections, while the last one was killed at forty days. The results of the study of the experimental series are as follows : The uninjected control showed every sign of a castrated condition (see Fig. 3), with the typical loss in gross size, changes in nuclear ar- rangement, lowering of the epithelium, etc. The vas deferens of the 10-day injected animal resembles the castrate type except that it shows an increase in the height of the epithelium with a partial disappearance of the pseudostratified effect. The secre- tion in the lumen and in the distal ends of the epithelial cells is greater in amount. In the 20-day injected animal the vas is nearer normal in that it shows a strikingly high epithelium, with a fair amount of cilia. The 30-day vas is indistinguishable from that of a normal, as far as the structure of the epithelium is concerned. The size of the vas as a whole is considerably larger than its castrate control but not as large as the normal. The vas deferens from the 40-day injected prepubertal castrate is RAT VAS DEFERENS CYTOLOGY 331 normal both in structure and size. The diameter of the entire vas is now double that of the control (Fig. 4). This study shows that the prepubertal castrated vas deferens re- sponds definitely to the introduced testis extract as do the adult castrates and returns to the normal condition in forty days despite its undeveloped state for a period of about one hundred and ten days. Cross-sections of rat vas deferens. Photomicrographs of Bouin-haematoxylin preparations. About 650 X before reduction. 5. Portion of Fig. 1. (Normal animal.) 6. From a 20-day castrate. 7. From animal No. 96 — tissue removed seven months after castration. 8. From same animal (No. 96), \\hich hadi received thirty daily injections of bull testis extract after the removal of the tissue shown in Fig. 7. b. Adult Castrates A number of adult animals were castrated and allowed to remain for various periods of time before injections were begun. The intro- duction of testis extracts has always served to return the vas deferens to the normal condition provided the concentration of the lipoid ex- tract was sufficiently great. 22 332 SUP VATNA The results of injecting the extract into long time castrates will he illustrated by reference to one animal (No. 96). This animal was cas- trated and seven months later was operated upon for removal of one vas deferens to serve as the control, and its condition is shown in Fig. 7. The animal was then subjected to testis extract injection daily for a period of thirty days; one-half cc. was injected subcutaneously each day. It was killed and the opposite vas deferens removed to show the effects of the injection. A cross section of the vas after injection is shown in Fig. 8, and should be compared with its mate removed before injections were begun (in Fig. 7). It can be seen clearly that whereas the seven month castrated vas deferens is in a highly degenerate state, its partner has been returned to the normal condition by means of the injections. A second animal treated similarly, but injected for a period of only twenty days, showed that the vas deferens had returned to an almost normal condition within this period. When castration has been of shorter duration, injections have been followed by similar return to the normal condition. VI. DISCUSSION In this study we have demonstrated that the vas deferens is also under the control of the sex hormone for its normal maintenance, as was shown to be the case for the prostates and the seminal vesicles by Moore, Price and Gallagher (1930) and Moore, Hughes and Gallagher (1930) respectively. If the hormone-producing glands — the testes— are removed, certain definite degenerative changes set in, and these changes are maximal by about twenty days after testis removal. The vas reacts more slowly to castration than do the seminal vesicles and prostates of the rats and therefore has not provided as delicate a method for hormone assay, nor one as easily read as the light area of the prostates or the secretion granules of the seminal vesicles. Al- though the changes following castration do not appear as rapidly in the vas, they are as definite as those that appear in the other accessory re- productive glands that have been studied. The vas responds positively to potent injections of testis extract, therefore it provides a supple- mentary test for the presence of the male hormone. In other sections of this paper, data have been presented showing that by injections (1) vasa of castrated animals have been maintained at the normal level, (2) vasa that had been allowed to regress for seven months after castration have been built up to normal, and (3) vasa of prepubertally castrated animals have been allowed to regress for one hundred and ten days and have been built up to a normal functioning- state in forty days. RAT VAS DEFERENS CYTOLOGY One experiment was described in detail in which a rat was castrated and after seven months one vas was removed and the other remained to lie removed after thirty days of injections. The former was a typical castrate, and the latter showed a condition normal in every respect. From these data, there can be no doubt that the active principle of the testis has been supplied by testis extract injections. With varying potencies of hormone, the results of injections varied from negative effects to complete replacements of the vas to the normal state. The epithelium itself is more sensitive and responds more readily to hormone injection than does the muscular layer and consequently the vas may return to an approximately normal condition while the. gross size is below that of the normal. This same condition obtained in the prostate and the seminal vesicles. Since, by testis extract injection, the vas can be maintained in a normal state as is proved by histological and cytological study, it pro- vides us with another male hormone indicator method to add to those already developed — the spermatozoon motility test, the electric ejacula- tion test, the seminal vesicle test, the prostate cytology test, and the capon comb growth test. VII. SUMMARY AND CONCLUSIONS 1. The vas deferens can be used as a male hormone indicator because it is under the control of the internal secretion of the testis. 2. After castration, definite regressive changes take place within twenty days in all animals. 3. These changes involve : a. Reduction in gross size through regression of the muscular layer of the vas. b. Diminution of the amount of secretion in the lumen. c. Reduction in epithelial height. d. Loss of the cilia covering the epithelium. c. Crowding together of the cells and obliteration of the cell walls. /. Stratification of the nuclei. g. Great reduction in the amount of cytoplasm in the cells. //. Changes in the Golgi bodies involving loss in gross size and frag- mentation of the Golgi material into rods or granules instead of the typical reticulum of the normal. 4. All these changes can be prevented from developing in the cas- trated animal by daily injections of suitably potent male hormone pre- pared from the lipoid fraction of fresh bull testes and dissolved in olive oil. • "N. / I (u-i L I * • * i /' 334 SUP VATNA 5. If the changes have been allowed to develop, the vas can be built up to normal by daily injections of testis extracts. 6. In animals castrated before puberty and allowed to regress for one hundred and ten days the vas can be built up to a normal func- tioning state by injections ; a process which involves bringing the un- differentiated duct to a normal adult state. 7. Injections of pure olive oil fail to prevent castration changes, therefore the potent factor lies in the hormone itself. LITERATURE CITED BENOIT, J., 1920. Sur 1'existence de phenomenes secretaires dans le canal de- ferent. Compt. rend. Soc. de BioL, 83: 1640. BENOIT, J., 1926. Recherches anatomiques, cytologiques et histophysiologiques sur les voies excretrices du testicule, chez les mammiferes. Arch, d'anat.. d'liist. ct d'cmbryol., 5: 176. FUNK, C, AND HARROW, B., 1929. The Male Hormone. Proc. Soc. Expcr. Biol. and Mcd., 26: 569. FUNK, C. B., HARROW, B., AND LEJWA, A., 1929. The Male Hormone II. Proc. Soc. Exper. Biol. and Mcd.. 26: 569. FUNK, C, HARROW, B., AND LEJWA, A., 1930. The Male Hormone. Am. Jour. Physio!., 92: 440. GALLAGHER, T. F., AND KOCH, F. C., 1929. The Testicular Hormone. Jour. Biol. Chem., 84: 495. LOEWE, S., AND Voss, H. E., 1929. Gewinnung, Eigenschaften und Testierung eines miinnlichen Sexualhormons. Sits. Akad. Wiss. U'icn. Math. Naturw. KL. Oct. 24, 1929. LUDFORD, R. J., 1925. Some Modifications of the Osmic Acid Methods in Cyto- logical Technique. Jour. Roy. Mic. Soc., Part 1, p. 31. LUDFORD, R. J., 1926. Further Modifications of the Osmic Acid Methods in Cytological Technique. Jour. Roy. Mic. Soc., 46: 107. McGEE, L. C., 1927. The Effect of the Injection of a Lipoid Fraction of Bull Testicle in Capons. Proc. Inst. Med. Chicago, 6: 242. McGEE, L. C., JUHN, MARY, AND DOMM, L. V., 1928. The Development of Secondary Sex Characters in Capons by Injections of Extracts of Bull Testes. Am. Jour. PhysioL, 87: 406. MARTINS, T., AND ROCHA E SILVA, A., 1929. The Seminal Vesicles of the Cas- trated Mouse, Test for the Testicular Hormones. Suf>pL d. Mem. Inst. Oswaldo Cruz, 9: 196. Rio de Janeiro. MOORE, C. R., AND McGEE, L. C, 1928. On the Effects of Injecting Lipoid Ex- tracts of Bull Testes into Castrated Guinea Pigs. Am. Jour. PhysioL, 87: 436. MOORE, C. R., AND GALLAGHER, T. F., 1929. On the Prevention of Castration Effects in Mammals by Testis Extract Injections. Am. Jour. Physiol., 89: 388. MOORE, C. R., HUGHES, WINIFRED, AND GALLAGHER, T. F., 1930. Rat Seminal Vesicle Cytology as a Testis Hormone Indicator and the Prevention of Castration Effects by Testis Extract Injections. Am. Jour. Anat., 45: 109. MOORE, C. R., AND GALLAGHER, T. F., 1930. Seminal- Vesicle and Prostate Func- tion as a Testis-Hormone Indicator; the Electric Ejaculation Test. Am. Jour. Anat., 45: 39. • RAT VAS DEFERENS CYTOLOGY MOORE, C. R., PRICE, DOROTHY, AND GALLAGHER, T. F., 1930. Rat-Prostate Cytology as a Testis-Hormone Indicator and the Prevention of Castra- tion Changes by Testis-Extract Injections. Am. Jour. Anal., 45: 71. MOORE, C. R., GALLAGHER, T. F., AND KOCH, F. C., 1929. The Effects of Ex- tracts of Testis in Correcting the Castrated Condition in the Fowl and in the Mammal. Endocrinology, 13: 367. MOORE, C. R., VATNA, S., AND GALLAGHER, T. F., 1930. Rat Vas Deferens Cytology as a Testis Hormone Indicator and the Prevention of Castration Changes by Testis Extract Injections. Anat. Rcc. (In press.) MYERS-WARD, C. F., 1897. Preliminary Note on the Structure and Function of the Epididymis and Vas Deferens in the Higher Mammalia. Jour. Anat. London, 32: 135. ON DISTOMUM VIBEX LINTON, WITH SPECIAL REFER- ENCE TO ITS SYSTEMATIC POSITION H. W. STUNKARD AND R. F. NIGRELLI BIOLOGICAL LABORATORY, UNIVERSITY COLLEGE, NEW YORK UNIVERSITY Distoinniii vibcx was described by Linton (1900, 1901, 1905), from the pharynx and intestine of the smooth puffer, Spheroides maculatus. For many years this species has been studied as the representative of digenetic trematodes by the classes in Invertebrate Zoology at the Marine Biological Laboratory of Woods Hole. Since the early and brief re- ports of Linton, little or no research has been done on the parasite. The purpose of this study is, therefore, to supplement the earlier de- scriptions of its morphology and to allocate the species in the system of classification of the digenetic trematodes. LINTONIUM NEW GENUS Distoinniii Diesing 1850 is the equivalent of Distoina Retzius 17SJ. a name proposed as a substitute for Fasciola Linnaeus 1758 — and con- sequently a synonym. Looss (1899) showed that Distomum is not a generic but a group name, and with the subdivision and disappearance of the previously accepted genus Distoinniii, the proper generic name and systematic position of D. ribc.v has remained an open question. Since Distomum is not a valid generic name, and since the species can not be assigned to any existing genus, we propose the new genus Lintonium to contain it. The distribution of Lintonium i-Vr V tK^^I ^^W« " «WS*c5ftS»? >s tf-.. FIG. 1. Photographic reproduction of part of Phite III from Fol's paper (1879) on Astcrias i/lucialis, the drawings of which were from the living egg. In Fig. 1, a. b, and c are three successive phases of the same zoospcrm, zc. An extension of the entrance cone is at Sa. The phase in which the zoosperm entered is omitted here. In Fig. 2, a, />, c (d omitted) c, f, g, (It omitted) and i are seven views of the same objects; in Fig. 2, b, c, a /oosperm, ,;c, is approach ing. In c and / a second zoosperm, .;", is approaching. In iy*, ,^:.\f< FIG. 3. Progressive changes in the form of an exudation cone. of the cone before the spermatozoon has migrated halfway through the jelly, Fig. 2, B. Its complete elevation over the egg occurs within 5 to 20 seconds later. The conversion of the entrance cone into the exudation cone (Fol's cone d'exudation) takes place after the spermatozoon has passed into the egg. Ever-changing, flame-like processes develop on the cone, Fig. 2, M, N, which finally withdraw and the cone disappears, frequently leaving behind minute globules, Fig. 2, O—Q, which become dispersed in the space between the fertilization membrane and the egg. A varia- tion of the exudation cone is shown in Fig. 3. In over-inseminated eggs several spermatozoa may become attached, Fig. 4, A, each to the tip of a filament extending from the egg. Al- FIG. 4, A. Two spermatozoa migrating together into an over-inseminated egg. B. One lost its attachment and was discarded, while the other successfully entered the egg. 352 ROBERT CHAMBERS though these spermatozoa begin to move through the jelly, there is a tendency for only the most advanced one to reach and penetrate the egg. The others, before reaching the egg, tend at one time or another to lose connection with their filaments. Such released spermatozoa, after a spasmodic twitch or two, remain permanently motionless. Fig. 3, B, in the jelly. The filaments which have lost their spermatozoa are quickly withdrawn and, together with their cones, soon sink into the egg. 'The filaments, extending from a cone to a spermatozoon, are usually at right angles to the egg's surface. That this is not always the case is shown in Fig. 5, where two convergent filaments are shown. This argues against the pre-existence of definite radial canals in the egg- jelly through which the spermatozoa might be supposed to move. The shape of the head of the spermatozoon, as already commented upon by Fol, occasionally changes considerably as the head moves through the jelly. The change seems to be due mainly to a bulging of the neck-piece on one or both sides of the head. Fig. 6, A, B (cf. Fig. " 40" 60" FIG. 5 FIG. 6 FIG. 5. Two spermatozoa attached to insemination filaments which are convergent and not radial as usual. FIG. 6. A. Sketches to show variations in shape of the heads of spermatozoa migrating through the egg-jelly. B. Changes in shape of the head of one sper- matozoon at intervals of 20, 40, and 60 seconds. 5). In Fig. 6, B are three sketches of a single spermatozoon, at in- tervals of 20, 40 and 60 seconds after insemination. The impression that the head of the spermatozoon is bent to one side may be due to the distorted shape of the neck-piece. Occasionally, a spermatozoon ap- pears to be carried through the jelly with the base of its tail at right angles to the attachment of the insemination filament, while the rest of the tail is curved so as to trail behind. SPERM ENTRY IN THE STARFISH EGG 353 Figures 7—10 represent variations. Fig. 7 shows a sperm-head which was unusual in performing active, wriggling movements for fully one minute after having penetrated the egg while the tail hung motionless outside. During these movements the sperm-head left the usual hy- aline pathway and could be seen jostling and pushing aside the cyto- plasmic granules encountered. Fig. 8 shows a spermatozoon whose head, after passing through ^•% •>- IS I 1 ^v;y^ ^x 2& ^ ^ SC to FIG. 7. Four successive steps in the progress of an unusually active sperm- head after it had penetrated an egg. ri.-, .+•*••-(., \-^.._ ••*•*> j: - •' . i"- - . •* '4 -.<,':, ., - -V-i - T -yU.% ; (J ' B -• -•' ': O B 1'4S" C 4' 15" FIG. 8 FIG. 8. A spermatozoon which on entering an egg left its neck-piece outside the fertilization membrane. the fertilization membrane, broke away from its neck-piece which was left outside with the tail. Figs. 9 and 10 show the reactions of late arriving spermatozoa. Fig. 9 shows a spermatozoon which succeeded in passing through an already lifted fertilization membrane. During the process the cone changed shape and flattened out, while the fertilization membrane be- came appreciably indented. In Fig. 10 the spermatozoon reached the cone, A-C, but failed to enter. The fertilization membrane wrinkled and the cone formed accessory elevations, D—F, but, when the cone finally withdrew from the membrane, the spermatozoon was left out- 354 ROBERT CHAMBERS side. The head of the spermatozoon then sprang back for a short dis- tance where it remained motionless and attached to the membrane by a slender thread, G, nine minutes after insemination. mmi ^spf; ',>^« *m^, £$$&$$! Iplp jjlJ®3im "!:^-?'M>''; ^'•'?-:+' jj^'^iJI B |'2S' 2'ifS" 0 E JIS" 3' 25" FIG. 9. Delayed entry of a spermatozoon through a fertilization membrane formed by the penetration into the egg of another spermatozoon not shown in the figure. 3. The Origin of the Insemination Filament The insemination filament is so fine that it is practically invisible except when the cone at one end and the sperm-head at the other end are brought simultaneously into focus. Considerable practice is re- quired to detect the sperm at the moment when it is beginning to mi- grate into the jelly. In the outer border of the jelly among several spermatozoa whose heads are moving to and fro while their tails lash about, one's attention becomes attracted to a sperm-head which has ceased its side-to-side movements and, instead, is moving steadily and in a straight line into the depths of the jelly. By looking along the direction of its movement, a cone on the egg's surface becomes apparent and, between the cone and the sperm, is to be seen the delicate, tenuous insemination filament. In fresh maturing eggs I have never been able to see the cone without also seeing the advancing sperm and the filament connecting the two. The formation of the filament is apparently too rapid. In immature eggs the cone is relatively much larger and as already described (Chambers, 1923) I have several times observed a tapering extension grow out from it until contact is made with a sperm, whereupon the extending portion retracts and draws the sperm in with it. In mature eggs which have been standing in sea-water for 2 to 4 hours there is frequently a greater response to multiple cone formation than in fresh, maturing eggs and consequently the chances are better to catch the initial stages. Eggs, 3 hours old, were placed in a shallow hanging drop in a moist chamber and, after being brought under ob- servation, a suspension of sperm was blown into one side of the field by means of a micro-pipette. The spermatozoa quickly spread in the interstices between the eggs and several became attached to the SPERM ENTRY IN THE STARFISH EGG 355 outer border of the jelly of the egg in view. Within 10 seconds a number of minute, conical, blister-like elevations developed on the egg's surface opposite the sperm. A delicate membrane appeared as if it were being lifted off the egg's surface by the rising cones. A few of the hyaline cones increased in size and, during the several succeeding seconds, there was no sign of any connection between them and the sperm lying on the periphery of the jelly. One cone increased ap- 1 I I I C ' !' 1'20" b'1 3'' 40" iiaaf!"-v~f"-v2ji.'5-.v5.' FIG. 10. Attempted penetration of a delayed spermatozoon which was finally discarded. preciably in size and suddenly, within an instant, a distinct line could be seen connecting its tip with the head of a spermatozoon. The other spermatozoa remained on the surface of the jelly while the spermatozoon in question began to migrate inward. While this was occurring, the rounded surface of the cone tapered more and more and the ever- shortening filament became appreciably thicker. A curious phenomenon which may be of significance is the fact that, in the majority of cases, the insemination filament always connects with a spermatozoon. Because of this one is almost inclined to believe in 356 ROBERT CHAMBERS a specific attraction such as Fol suggested. I may cite, for example, a case in which about 30-50 spermatozoa were blown on the surface of an egg. Most of the spermatozoa immediately became attached to a restricted region on the outer border of the jelly. One, however, wandered off a short distance and suddenly a cone appeared with a _.V.7 ::-v...// ,- • , ,. , - , i y, • - ^•..\>'1' !"]"><•' •' '- ».-^^^]r^^Lj^.\-'.'^. V/;-^ '&.** V;v FIG. 11 FIG. 12 FIG. 11. Polyspermy in an egg 5 hours old. The egg nucleus and two polar bodies show prominently in the middle of the figure. Sperm at x, although more advanced, entered later than sperm at y. FIG. 12. Polyspermy in an immature egg. tenuous filament extending to the spermatozoon diagonally through the jelly. The filament then retracted with the spermatozoon on its tip and insemination resulted. 4. Insemination of Immature and of Aged Eggs Eggs aged by standing in sea- water lose their protective reaction against polyspermy. Fig. 11 represents an egg which was inseminated after it had been standing in sea-water for five hours, which is over four hours longer than is usual for normal fertilization. Within one minute numerous cones formed on the egg. The figure shows the egg with six attached spermatozoa, all of which were taken in. Owing to the rapidity of the procedure and the variations in the angles of direction which the filaments take, it was impossible to ascertain whether or not the cones in the figure which show no filaments did in reality possess filaments with spermatozoa attached to them. There is often a lack of uniformity in the sequence of the sperm entries. In Fig. 1 1 the spermatozoon at x was in advance of its neighbor at y. In spite of this, spermatozoon y entered before x. One egg, two hours after maturation, formed two cones with in- SPERM ENTRY IN THE STARFISH K0 M B b b 11 2 b 00 r T— t (N O CN 6 cs b tN C b b c :8 •lu b o 10 b o\ OC OO T to N O < vO 0 E >, CJ >"S x ^ ^ ^ w -M "^ O~ b b b S "5 •:'~v o ro IO OJ OJ E K- JD ^f-l 03 03 E >, 03 u 03 C U ~ ^_ ^ ^ ^ O ^"a" *o ^ b QJ O cu B-g b-S 0.3 b b b b b b •a T3 Cfl Q^ 0) t "»* THE MARINE BIOLOGICAL LABORATORY Editorial Board GARY N. CALKINS, Columbia University FRANK R. LlLLIE, University of Chicago E. G. CONKLIN, Princeton University CARL R. MOORE, University of Chicago E. N. HARVEY, Princeton University GEORGE T. MOORE, Missouri Botanical Garden SELIG HECHT, Columbia University T. H. MORGAN, California Institute of Technology M. H. JACOBS, University of Pennsylvania G. H. PARKER, Harvard University H. S. JENNINGS, Johns Hopkins University W. M. WHEELER, Harvard University E. E. JUST, Howard University EDMUND B. WILSON, Columbia University ALFRED C. REDFIELD, Harvard University Managing Editor APRIL, 1930 Printed and Issued by LANCASTER PRESS, Inc. PRINCE & LEMON STS. LANCASTER, PA. THE BIOLOGICAL BULLETIN is issued six times a year. Single numbers, $1.75. Subscription per volume (3 numbers), $4.50. Subscriptions and other matter should be addressed to the Biological Bulletin, Prince and Lemon Streets, Lancaster, Pa. Agent for Great Britain: Wheldon & Wesley, Limited, 2, 3 and 4 Arthur Street, New Oxford Street, London, W.C. 2. Communications relative to manuscripts should be sent to the Managing Editor, 240 Longwood Avenue, Boston, Mass. Entered October 10, 1902, at Lancaster, Pa., as second-class matter under Act of Congress of July 1 6, 1894. CONTENTS Page HOADLEY, LEIGH Some Effects of HgCl2 on Fertilized and Unfertilized Eggs of Arbacia punctulata 123 WHITAKER, DOUGLAS, and MORGAN, T. H. The Cleavage of Polar and Antipolar Halves of the Egg of Chaetopterus 145 REDFIELD, ALFRED C. The Absorption Spectra of Some Bloods and Solutions Con- taining Hemocyanin 150 CONKLIN, CECILE Anoplophrya marylandensis n. sp., a Ciliate from the Intes- tine of Earthworms of the Family Lumbric dae 176 DEMPSTER, W. T. The Growth of Larvae of Ambystoma maculatum under Nat- ural Conditions 182 SMITH, DIETRICH C. The Effects of Temperature Changes upon the Chromato- phores of Crustaceans 193 Ol Volume LVIII Number 3 THE BIOLOGICAL BULLETIN PUBLISHED BY THE MARINE BIOLOGICAL LABORATORY Editorial Board GARY N. CALKINS, Columbia University FRANK R. LlLLIE, University of Chicago E. G. CONKLIN, Princeton University CARL R. MOORE, University of Chicago E. N. HARVEY, Princeton University GEORGE T. MOORE, Missouri Botanical Garden SELIG HECHT, Columbia University T. H. MORGAN, California Institute of Technology M. H. JACOBS, University of Pennsylvania G. H. PARKER, Harvard University H. S. JENNINGS, Johns Hopkins University W. M. WHEELER, Harvard University E. E. JUST, Howard University EDMUND B. WILSON, Columbia University ALFRED C. REDFIELD, Harvard University Managing Editor JUNE, 1930 Printed and Issued by LANCASTER PRESS, Inc. PRINCE & LEMON STS. LANCASTER, PA. THE BIOLOGICAL BULLETIN is issued six times a year. Single numbers, $1.75. Subscription per volume (3 numbers), $4.50. Subscriptions and other matter should be addressed to the Biological Bulletin, Prince and Lemon Streets, Lancaster, Pa. Agent for Great Britain: Wheldon & Wesley, Limited, 2, 3 and 4 Arthur Street, New Oxford Street, London, W.C. 2. Communications relative to manuscripts should be sent to the Managing Editor, 240 Longwood Avenue, Boston, Mass. Entered October 10, 1902, at Lancaster, Pa., as second-class matter under Act of Congress of July 16, 1894. CONTENTS Page COE, WESLEY R. Unusual Types of Nephridia in Nemerteans 203 GRAY, I. E., and HALL, F. G. Blood Sugar and Activity in Fishes with Notes on the Action of Insulin 217 BLUM, HAROLD F. Studies of Photodynamic Action. I. Hemolysis by Previously Irradiated Fluorescein Dyes 224 REDFIELD, ALFRED C. The Equilibrium of Oxygen with the Hemocyanin of Limulus polyphemus determined by a Spectrophotometric Method . . 238 HOADLEY, LEIGH Polocyte Formation and the Cleavage of the Polar Body in Loligo and Chaetopterus 256 PICKFORD, GRACE EVELYN The Distribution of Pigment and other Morphological Con- comitants of the Metabolic Gradient in Oligochaets 265 SlVICKIS, P. B. Distribution of Setae in the Earthworm, Pheretima ben- guetensis Beddard 274 JAHN, THEODORE L. Studies on the Physiology of the Euglenoid Flagellates. II. The Autocatalytic Equation and the Question of an Auto- catalyst in Growth of Euglena 287 HARVEY, ETHEL BROWNE The Effect of Lack of Oxygen on the Sperm and Unfertilized Eggs of Arbacia punctulata, and on Fertilization 288 RAFFEL, DANIEL The Effect of Conjugation within a Clone of Paramecium aurelia 293 SMITH, GEORGE MILTON A Mechanism of Intake and Expulsion of Colored Fluids by the Lateral Line Canals as Seen Experimentally in the Goldfish (Carassius auratus) 313 VATNA, SUP Rat Vas Defer ens Cytology as a Testis Hormone Indicator and the Prevention of Castration Changes by Testis Extract Injectims ' 322 W., and NiGRELLl, R. F. OiDistomum vibex Linton, with Special Reference to its Systematic Position 336 CHAMBERS, ROBERT The Manner of Sperm Entry in the Starfish Egg 344 MBL WHOI LIBRARY « • • iii i II I f | |l || WH 17IA •/.