EDITOR'S NOTE Line 7 on p. 184 of the article by Merkel II. Jacobs should read : THE BIOLOGICAL BULLETIN PUBLISHED BY THE MARINE BIOLOGICAL LABORATORY Editorial Board GARY N. CALKINS, Columbia University E. G. CONKLIN, Princeton University FRANK R. LlLLJE, University of Chicago E. N. HARVEY, Princeton University CARL R. MOORE, University of Chicago SELIG HECHT, Columbia University GEORGE T. MOORE, Missouri Botanical Garden LEIGH HOADLEY, Harvard 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 VOLUME LXII FEBRUARY TO JUNE, 1932 Printed and Issued by LANCASTER PRESS, Inc. PRINCE & 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: \Yheldon & Wesley, Limited, 2, 3 and 4 Arthur Street, New Oxford Street, London, \V.C. 2. Communications relative to manuscripts should be sent to the Managing Editor, Marine Biological Laboratory, Woods Hole, Mass., between May 1 and November 1 and to the Institute of Biology, Divinity Avenue, Cambridge, Mass., during the remainder of the year. Entered October 10, 1902, at Lancaster, Pa., as second-class matter under Act of Congress of July 16, 1894. LANCASTER PRESS, !NC. LANCASTER, PA. CONTENTS No. 1. FEBRUARY, 1932 PAGF LUNDSTROM, H. M., AND P. BARD Hypophysial Control of Cutaneous Pigmentation in an Elas- mobranch Fish 1 LUTZ, B. R., AND L. C. WYMAN Reflex Cardiac Inhibition of Branchio-vascular Origin in the Elasmobranch, Squahis acanthias 10 WYMAN. L. C., AND B. R. LUTZ The Effect of Adrenalin on the Blood Pressure of the Elas- mobranch, Squalus acanthias 17 DILL, D. B., H. T. EDWARDS, AND M. FLORKIN Properties of the Blood of the Skate (Raia ocellata) 23 FREMONT-SMITH, F., AND M. E. DAILEY The Nature of the Reducing Substances in the Blood Serum of Limulus polyphemus and in the Serum, Cerebrospinal Fluid and Aqueous Humor of Certain Elasmobranchs 37 BARRON, E. S. G. Studies on Cell Metabolism. I. The oxygen consumption of Nereis eggs before and after fertilization 42 BARRON, E. S. G. The Effect of Anaerobiosis on the Eggs and Sperm of Sea Urchin, Starfish and Nereis and Fertilization under Anaerobic Conditions 46 ROMANOFF, ALEXIS L. Fat Metabolism of the Chick Embryo under Standard Condi- tions of Artificial Incubation 54 JACOBS, M. H., AND A. K. PARPART Is the Erythrocyte Permeable to Hydrogen Ions? 63 BERRILL, N. J. Ascidians of the Bermudas 77 JAI-IN, THEO. L., AND L. R. KUHN The Life-history of Epibclella melleni Maccallum 1927, a Mono- genetic Trematode Parasitic on Marine Fishes 89 iv CONTENTS ADOLPH, EDWARD F. The Vapor Tension Relations of Frogs 112 WKIKR, T. ELLIOT A Comparison of the Plasticl with the Golgi Zone 126 No. 2. APRIL, 1932 HARVEY, E. NEWTON Physical and Chemical Constants of the Egg of the Sea Urchin, Arbacia punctulata 141 HARVEY, ETHKL BROWNK The Development of Half and (Juartcr Eggs of Arbacia punc- tulata and of Strongly Centrifuged Whole Eggs 155 BENNITT, RUDOLF, AND AMANDA DICKSON MERRICK Migration of the Proximal Retinal Pigment in the Crayfish in Relation to Oxygen Deficiency 168 JACOBS, MERKEL H. Osmotic Properties of the Erythrocyte. III. The applica- bility of osmotic laws to the rate of hemolysis in hypotonic solutions of non-electrolytes 178 PARKER, GEORGE H., AND MARGARET VAN ALSTYNK Locomotor Organ. s of Kchinarachnius parma 195 FAURE-FREMIET, E. Stron iliidium Calkinsi, a New Thigmotactic Species 201 HAHNERT, WILLIAM F. Studies on the Chemical Needs of Amoeba proteus : a Culture Method 205 AREY, LESLIK I!. The Formation and Structure of the Glochidial Cyst 212 No. 3. JUNE, 1932 COLMAN, JOHN A Statistical Test of the Species Concept in Littorina 223 RAFFEL, DANH i Inherited Variation Arising during Vegetative Reproduction in Paramecium aurelia 244 SONNEBOKN, T. M., AND C. J. LYNCH Racial Differences in the Early Physiological Effects of Con- jugation in Paramecium aurelia 258 MOORK, W. G. The Effects ii!" X-rays on Fertility in Drosophila melanogaster Treated at Different Stages in Development 294 CONTENTS v BARNES, T. CUNLIFFE, AND HENRY I. KOI-IN The Effect of Temperature on the Leg Posture and Speed of Creeping in the Ant, Lasius 306 JACOBS, M. H., AND ARTHUR K. PARPART Osmotic Properties of the Erythrocyte. IV. Is the permea- bility of the erythrocyte to water decreased by narcotics? 313 STUNKARD, HORACE W., AND RAYMOND M. CABLE The Life History of Parorchis avitus (Linton), a Trematode from the Cloaca of the Gull 328 INDEX 339 Vol. LXII, No. 1 February, 1932 THE BIOLOGICAL BULLETIN PUBLISHED BY THE MARINE BIOLOGICAL LABORATORY HYPOPHYSIAL CONTROL OF CUTANEOUS PIGMENTA- TION IN AN ELASMOBRANCH FISH1 HELEN M. LUNDSTROM AND PHILIP BARD EDITOR'S NOTE: The equation on page 277 of the article by E. N. Harvey in the December, 1931, issue should be corrected to read as follows: 2.12 X IP'7 X .8 X 10-2 X 103 X 1.6 X 103 6.28 X 23 X 10- = °'19 P. E. Smith (1916), Allen (1917, 1920), Atwell (1919), Rehberg and Krogh (Krogh, 1922) and Hogben and his collaborators (Hogben, 1924). It is a curious fact that, so far as we have been able to ascer- tain, no experimental investigation of chromatic function in the elasmobranch group has ever been made, although these animals possess dermal chromatophores (melanophores and xanthophores) not unlike those seen in the frog (Daniel, 1922). That pituitary removal alone is followed by the development of pallor was shown in a series of hypophysectomies performed on 40 dogfish. The operation is easily carried out through a buccal ap- proach. 1 A report of results obtained in an investigation carried out in the Course in Physiology at the Marine Biological Laboratory, Woods Hole, during the summer of 1931. 2 1 Vol. LXII, No. 1 February, 1932 THE BIOLOGICAL BULLETIN PUBLISHED BY THE MARINE BIOLOGICAL LABORATORY HYPOPHYSIAL CONTROL OF CUTANEOUS PIGMENTA- TION IN AN ELASMOBRANCH FISH l HELEN M. LUNDSTROM AND PHILIP BARD (From the Marine Biological Laboratory, Woods Hole) In a study of the effects of ablation of various parts of the brain of the dogfish, Mustelis canis, it was noticed that certain operated individuals became, in the course of the first few post-operative hours, lighter in color than their fellows. The degree of paling was such as to render the affected animal distinctly lighter than any normal dogfish of this species that we haVe seen. Inspection of the cranial contents of the operated fish showed that whenever the albinous appearance was produced the pituitary body had been extirpated in the intra- cranial intervention. On the other hand, the hypophysis was invari- ably intact in those operated fish which retained their normal dark grayish color. Quite naturally we were led to suspect that the dogfish possesses an hypophysial control of its cutaneous pigmentation similar to that which has been so clearly established for the amphibia by P. E. Smith (1916), Allen (1917, 1920), Atwell (1919), Rehberg and Krogh (Krogh, 1922) and Hogben and his collaborators (Hogben, 1924). It is a curious fact that, so far as we have been able to ascer- tain, no experimental investigation of chromatic function in the elasmobranch group has ever been made, although these animals possess dermal chromatophores (melanophores and xanthophores) not unlike those seen in the frog (Daniel, 1922). That pituitary removal alone is followed by the development of pallor was shown in a series of hypophysectomies performed on 40 dogfish. The operation is easily carried out through a buccal ap- proach. 1 A report of results obtained in an investigation carried out in the Course in Physiology at the Marine Biological Laboratory, Woods Hole, during the summer of 1931. 2 1 2 H. M. LUNDSTROM AND PHILIP BARD METHOD The fish was tied down in the dorsal position on a small shark board and a stream of sea water introduced by means of a tube into the pharyngeal region to maintain respiration (under these circumstances normal breathing movements continue). After incision and retraction of the mucous membrane the base of the skull was lightly scraped along the midline at the level of the pupils of the eyes. This brings to view through the translucent cartilage the chiasma, hypothalarni, and posterior hypophysis. A medial opening was then cut in the skull between the chiasma and the posterior lobe. This exposes the elongated anterior lobe of the pituitary which lies in the depression between the hypothalarni and extends forward nearly to the chiasma. The width of the exposure was such as to lay bare the medial half of each ventral hypothalamic surface. So much may be done without the slightest hemorrhage. But it was found that extension of the skull defect backward so as to expose merely the cranial aspect of the large bifid neuro-intermediate or posterior lobe resulted in profuse bleeding. Therefore this structure was carefully scooped out by means of a small curette introduced through the limited opening. With a little care this can be done without provoking hemorrhage and with minimal injury to the base of the brain. The anterior lobe may be left intact or it may be picked up with forceps, separated from the overlying nervous tissue, and so extirpated with very little damage to the hypothalamus. No attempt was made to close the wound. Provided the base of the brain had not been severely traumatized, the operated fish swam around normally and showed a general behavior indistinguishable from that of unoperated specimens. They usually succumbed after three or four days with loss of righting reactions and dyspnea as premonitory symptoms. A few lived as long as a week, which is about the average length of life of unoperated specimens kept under laboratory conditions. It seems likely that death was often induced by cerebral injury, some degree of cerebral herniation being the usual finding at autopsy. \Ye have not been disturbed by this rather crude postoperative situation. It is apparently without significance for the specific problem in hand, namely, the experimental analysis of pituitary control of cutaneous pigmentation, for injury or herniation of the hypothalamic region without removal of the hypo- physis never gave rise to the slightest pallor. Several animals survived such a condition several days and retained the same shade as normal control animals. HYPOPHYSIAL CONTROL OF CUTANEOUS PIGMENTATION RESULTS Complete removal of the pituitary body (30 experiments) always resulted in pallor. Slight paling was first noticeable about thirty minutes after pituitary ablation and after three hours the change was generally very marked, but maximal pallor was not attained until about the twelfth postoperative hour. It was our practice to select pairs of fish having the same degree of pigmentation, hypophysectomize one and retain the other for comparison. Plate 1 is a photograph of two such fish taken twenty-four hours after complete hypophysectomy in one. It shows clearly the extreme pallor characteristic of the animal without hypophysis. That the pallor is due to "contraction" of the dermal melanophores is shown in Plates 3 and 4, which are microphotographs of homologous pieces of skin taken from the same fish just before and nineteen hours after hypophysectomy. As shown in Plate 3 the melanophores of the normal fish are in a state of con- siderable expansion, while Plate 4 indicates clearly that the pallor developing after pituitary removal is the result of the assumption by the melanophores of a more or less rounded contour. Were it not for the outlines of several placoid scales one might easily mistake these plates for microphotographs of the skin of a normal and a hypo- physectomized frog or axolotl. Removal of the anterior lobe alone was done in five dogfish and these survived the operation for several days without showing the slightest change in cutaneous pigmentation. Nor did leaving the anterior lobe intact impede in any way the development of the pallor ensuing upon removal of the posterior lobe. As already mentioned, severe traumatization of the hypothalamus was found to be without influence on pigmentation provided the posterior lobe of the pituitary was not ablated. In four fish the anterior lobe and hypothalamic protuberances were exposed in the usual way, transverse incisions made in these structures to a depth of one or two millimeters and much of the substance of the hypothalamus pulled out with forceps. Two of these animals survived thirty hours without at any time presenting the slightest change in hue. The other two, whose color also remained unchanged, were subjected to posterior pituitary removal after twenty-four hours whereupon there ensued a pallor which developed at the usual rate and to the usual extent. Further evidence for an hypophysial control of the cutaneous melanophores was sought in a study of the effects of pituitary prepa- rations when injected into pale hypophysectomized dogfish. Each of five pale specimens received subcutaneously a crushed dogfish posterior lobe suspended in one or two cubic centimeters of sea water. A 4 H. M. LUNDSTROM AND PHILIP BARD distinct general darkening appeared within the first three minutes and after an hour these fish were indistinguishable from the darkest normal animals. Then in the course of the next five or six hours they gradually returned to the albinous condition. Injections of similar suspensions of cerebellum, skeletal muscle, and pancreas from normal dogfish were entirely without effect. In three experiments it was found that administration of crushed anterior lobe resulted in a moderate dark- ening which was slower to develop and which disappeared more rapidly than the more intense darkening caused by an equal quantity of crushed posterior lobe. In one pair of pale fish injection into one of a crushed posterior lobe fragment of about half the size of the anterior lobe produced an immediate darkening which reached its maximum in one hour, while injection of the crushed anterior lobes of two fish into the other produced a moderate darkening which was slow to develop and had a very short duration. These results together with those of the ablation experiments indicate that the posterior lobe is the chief source of the melanophore-expanding substance or secretion. The activity of the anterior lobe material may well have been due to diffusion into it of active substance of posterior lobe origin. Shortage of fish prevented our exploring this question further. I -A tracts of several posterior lobes were made. Immediately after its removal from a living dogfish the lobe was crushed, ground up, stirred and shaken in sea water, and the resulting suspension filtered. After appropriate dilution with sea water the clear filtrate was tested by injection of varying amounts into thoroughly pale hypophysecto- ini/ed fish of the same size as the donor of the lobe. It was found that an amount of filtrate corresponding to one twenty-fifth of one posterior lobe was necessary to produce a noticeable darkening while it required from one-tenth to one-seventh of a filtrate to make such a fish as dark as normal controls. It is unlikely that such simple extracts as these represent the full potency of the glands. Hogben (192-1) states that the amount of melanophore stimulant in the pituitary of one frog is sufficient to d.irken the skin of fifty-six animals of the same species. The posterior pituitary preparations "pituitrin" (Parke, Davis & Co.) and "infundin" (Burroughs, \\Cllconie \ Co.) produced darkening in pale pituitaryless fish when given subcutaneously or intramuscularly. Plate 2 is a photograph of the pair of fish shown in Plate 1. It was taken one hour after the hypophysectoini/ed individual had received a subcutaneou> injection of 0.4 cc. of "obstetrical pituitrin," an amount equivalent to four international oxytocic units. It can be seen th.it this dose had darkened the animal almost to its preoperative HYPOPHYSIAL CONTROL OF CUTANEOUS PIGMENTATION 5 shade. In the case of several pale hypophysectomized fish, twenty to thirty inches in length, it was found that a subcutaneous dose of about 0.5 cc. (5 international units) of "obstetrical pituitrin" was required to produce in them a darkness equal to that of normal controls. It seemed interesting to test the action on elasmobranch melanophores of the two active principles separated from pituitary extracts by Kamm and his associates and now supplied by Parke, Davis and Co. under the names "pitressin" and "pitocin." It was found that 0.25 cc. (5 pressor units) of "pitressin" was capable of producing approximately the same effect as 0.5 cc. (5 international units) of "pituitrin." On the other hand 0.5 cc. (5 international units) of "pitocin" had no effect whatever on pale individuals of the same size. When twice this quantity of "pitocin" was injected there occurred a slight darkening which reached its maximum much more slowly and disappeared more rapidly than the reaction to "pitressin" or "pituitrin." Since "pitocin" contains one-half unit of pressor activity per cubic centimeter, it is probable that this mild positive result was due to the pressor substance present. These results are in general agreement with those of L. W. Rowe (1928), who reported that "vasopressin" ("pitressin") but not "oxytocin" ("pitocin") caused expansion of frog melanophores, a finding which is in contra- diction to the inference of Hogben and Winton (19226) that the uterine and melanophore stimulants of pituitary extracts are identical. We had hoped to secure evidence of the presence of melanophore- expanding material in the circulating blood, but shortage of dogfish during the latter part of the summer made this impossible. Indirect evidence that pituitary substance acts directly on the dermal melano- phores was, however, secured in our observation, often repeated, that the melanophores of small isolated pieces of skin from pale dogfish expanded markedly when such skin fragments were placed in dilute sea-water solutions of "pituitrin" or "pitressin" or in suspensions or extracts of dogfish posterior lobe. When such pieces were immersed in sea water alone the melanophores remained contracted. The expanded melanophores of bits of skin from normal fish gradually contract when placed in sea water. Addition of "pituitrin," etc., prevented this slow contraction. Equivalent solutions of "pitocin" had no action on the melanophores of skin fragments. DISCUSSION The results presented in the foregoing section give abundant proof that the normal dark grayish color of the dogfish is maintained through the agency of the hypophysis. Apparently this organ continuously 6 H. M. LUNDSTROM AND PHILIP BARD delivers to the blood stream a quantity of melanophore stimulant capable of maintaining the rather widely expanded state of the cutaneous melanophores to which the normal color of the fish is due. The experimental findings strongly suggest that this humoral agent has its chief, if not its sole, origin in the posterior lobe. This portion of the gland is more correctly termed the neuro-intermediate lobe. Apart from a rather small extension of neuroglia fibers from the border of the infundibular cavity it is wholly composed of a highly vascular mass of basophil cells which, on the basis of embryological as well as histological observations, obviously constitutes the homo- logue of the pars intermedia of higher forms (de Beer, 1926). The ventral lobe of the selachian hypophysis which is attached dorsally to the neuro-intermediate lobe and ventrally to the skull and which, according to de Beer (1926), is composed largely but not entirely of basophil cells, has not been taken into separate account in our experi- ments. In all of our posterior lobe extirpations it was doubtless removed. It would seem highly probable that the melanophore- expanding substance originates in the pars intermedia. There is excellent evidence that such is the case in the amphibia. \Ye know that there too anterior lobe removal is not followed by pallor (Hogben and \Vinton, 1923) and that while the melanophore stimulant is not lacking from extracts of pars anterior and pars nervosa of beef glands, there is more of it per gram of dried tissue in the pars intermedia than in the other two portions of the gland (Hogben and Winton, 1922a, 19226). Some further confirmation of this point comes from Atwell's report (1919) that albinous tadpoles darken as a result of melanophore expansion when placed in dilute extracts of beef pars intermedia. But by far the mcst conclusive results have been those obtained by B. M. Allen (1920) in his study of the effects of trans- plantation of different parts of the hypophysis of the adult frog (Rana pipiens) into tadpoles from which the pars buccalis had been removed (anlage of pars nervosa left intact) and which, as a result, exhibited pallor as well as failure to metamorphose and retardation of growth. The operated larvae which received anterior lobe trans- plants showed "not the slightest tendency to return to the origin. il hi, iik color except for a slight tendency at the beginning," although this treat incut caii-cd marked acceleration of growth. On the other hand, the albinous tadpoles regained their dark color when the inter- mediate lobe was engrafted into them. In a recent communication (1930) Allen reports experiments in confirmation of his earlier results and shows that transplants of pars nervosa do not cause pigmentary effects. Allen's work is a most significant contribution to the problem HYPOPHVSIAL CONTROL OF CUTANEOUS PIGMENTATION 7 of localization of pituitary function. Although it presents the only unequivocal evidence for the precise origin of the amphibian melano- phore stimulant, it is not referred to in Hogben's monograph (1924). Shortly after the appearance of Allen's paper, Swingle (1921) reported that intraperitoneal transplantations of pars intermedia from various frogs into bullfrog larva1 caused darkening of the skin. It can be concluded that in the frog the pigmentary hormone is produced only by the pars intermedia and our own findings lend support to the view that such is the case in the elasmobranch. It is well known that amphibia exhibit a rhythm of color response which depends on a balance of such environmental factors as humidity, temperature, oxygen supply, and illumination (Hogben, 1924). In the majority of amphibia light causes pallor while its absence promotes darkening. We have found that dogfish taken from a moderately illuminated tank paled markedly when placed in a brightly illuminated tank, while other fish originally of the same shade and from the same tank darkened perceptibly when placed in a tank from which nearly all light was excluded. These changes require several hours. They cannot, of course, be referred to changes in humidity and we were at some pains to be sure that during these tests the temperature and aeration of the tanks were equal. Segments of illuminated fish which were covered by bands of adhesive plaster did not remain dark. We conclude, therefore, that light causes pallor by a general rather than a local action, and since pale hypophysectomized fish remained pale when kept in darkened tanks we can assume that these pigmentary changes evoked by changes in illumination are the results of variations in pituitary activity. Hogben (1924) is strongly of the opinion that regulation of color response in the amphibia by melanophore contraction and expansion is correlated with a fluctuating pituitary secretion. He believes this to be an adequate explanation of all the pigmentary phenomena seen in these animals and he is inclined completely to discount all assump- tions of a direct nervous control of amphibian melanophores. He suggests that such chromatic effects as have been observed following nerve section or stimulation or drug administration have been due to vasomotor changes. Since it is clearly established that adrenalin injected intravenously causes pallor in the frog we made some effort to determine whether injection of this substance has any influence on the pigmentation of the dogfish. In a few preliminary experiments we gained the impression that large subcutaneous injections of adrenalin caused slight paling. Subsequently, at a time when the supply of fish had greatly diminished, we tested the effect of injections made H. M. LUNDSTROM AND PHILIP BARD directly into the heart chambers. In collaboration with Dr. Rene Gayet, who kindly offered to share a number of fish with us, it was found that 1 cc. of a 1 : 1000 solution of adrenalin chloride evoked a paling which usually became noticeable after from five to ten minutes, sometimes not until after fifteen minutes, progressed to a maximum in about an hour, and persisted for approximately two hours. In one case the pallor was maximal; in the others it was not. Direct micro- scopical examination showed that in these cases the melanophores had actually undergone contraction. But the inadequate number of experiments performed makes it hazardous to venture an inference. Yet the comparative sluggishness of the responses provoked by these enormous doses which obviously produced a state of shock does not strongly suggest that the dogfish possesses an adrenal or sympathetic control. It may well have been that vasoconstriction of cutaneous vessels led to an asphyxiation and consequent contraction of the melanophores. Because of scarcity of material we were obliged to postpone to a later occasion an adequate investigation of the interesting question of a possible extra-pituitary control of elasmobranch chro- matophores. SUMMARY 1. Complete hypophysectomy in the dogfish, Mnstel-is can-is, results in pallor of the skin. This same result follows removal of the posterior (neuro-intermediate) lobe alone. Extirpation of the anterior lobe does not result in paling. Severe traumatization of the hypothalamus without destruction of the posterior lobe of the hypophysis has no influence on cutaneous pigmentation. 2. The pallor following hypophysectomy is due to "contraction" of the cutaneous melanophores which are normally in a state of considerable "expansion." 3. Suspensions or extracts of dogfish posterior lobe and the com- mercial posterior pituitary preparations, "pituitrin," "infundin," and " pitressin " produce melanophore expansion and consequent darkening when injected subcutaneously or intramuscularly into pale hypophy- sectomi/ed dogfish. Suspensions of dogfish cerebellum, skeletal muscle, and pancreas have no effect upon pigmentation. The oxytocic principle of the preparation, "pitocin," does not produce melanophore expansion in hypophysectomi/ed fish. Administration of suspensions of dogfish anterior lobe cause a relatively weak darkening of pale fish. The melanophores of isolated pieces of dogfish skin react to these various substances in the same way as do the melanophores of fish receiving them by injection. 4. Bright illumination of normal dogfish for several hours causes a PLATE 1. Photograph of a pair of dogfish, originally of the same shade, taken twenty-four hours after removal of the hypophysis in the animal on the right. Both animals had been kept in the same tank and the operated indi- vidual showed no abnormalities of behavior. I'I.ATK 2. IMidio-iMpli of ilic smie pairof dogfish taken one hour after the hypophysectomized individual (on the right ) li.ul n-rriM'd ,i sulirui.iiH'ous injection of 0.4 cc of i>itnitrin" (Parke, Davis «S; Co.). PLATE 3. Microphotograph of a mount of skin removed from a noimal dogfish showing expanded condition of the melanophores. I 'i. ATI-: 4. Microphotograph of an homologous piece of skin taken from the same fisli nineteen hours after hypophysectomy. HYPOPHYSIAL CONTROL OF CUTANEOUS PIGMENTATION 9 certain degree of pallor. Covered segments of the skin pale to the same extent as exposed regions. Absence of light leads to darkening of normal fish, but darkness does not modify the pallor of hypophy- sectornized individuals. It is concluded that the posterior lobe of the hypophysis is re- sponsible for the considerable degree of melanophore expansion characteristic of the skin of normal dogfish. Although the possibility of an extra-hypophysial control of cutaneous pigmentation has not been excluded, it is apparent that the pituitary gland plays a dominant role in the pigmentary alterations exhibited by the normal dogfish. BIBLIOGRAPHY ALLEN, B. M., 1917. Biol. Bull., 32: 117. ALLEN, B. M., 1920. Science, 52: 274. 'ALLEN, B. M., 1930. Proc. Soc. Exper. Biol. and Med., 27: 504. ATWELL, W. J., 1919. Science, 49: 48. DANIEL, J. F., 1922. The Elasmobranch Fishes. Berkeley. DE BEER, G. R., 1926. The Comparative Anatomy, Histology, and Development of the Pituitary Body. Edinburgh. HOGBEN, L. T., 1924. The Pigmentary Effector System. Edinburgh. HOGBEN, L. T., AND F. R. WiNTON, 1922o. Proc. Roy. Soc., London, Series B, 93: 318. HOGBEN, L. T., AND F. R. WINTON, 19226. Biochem. Jour., 16: 619. HOGBEN, L. T., AND F. R. WINTON, 1923. Proc. Roy. Soc., London, Series B, 95: 15. •KROGH, A., 1922. The Anatomy and Physiology of Capillaries. New Haven. ROWE, L. W., 1928. Endocrinol., 12: 663. •SMITH, P. E., 1916. Anal. Rec., 11: 57. SWINGLE, W. W., 1921. Jour. Exper. Zool, 34: 119. REFLEX CARDIAC INHIBITION OF BRANCHIO-VASCULAR ORIGIN IN THE ELASMOBRANCH, SQUALUS ACANTHI AS BREXTOX R. LUTZ AXD LELAXD C. WVMAX (From the Mount Desert Island Biological Laboratory and the Physiological Laboratory of Boston' University School of Medicine) Reflex cardiac and respiratory inhibition has been observed in elasmobranchs upon mechanical or electrical stimulation of various regions, external and internal, including the gill region (Lutz, 1930a). Reflex cardiac inhibition in mammals may also be obtained upon stimulating various regions. The distribution of the latter sensory areas is, however, much more restricted than in the elasmobranchs. Receptors which are especially important in regulating the circulation appear to be located in the vascular organs themselves. The depressor mechanism with receptors located in the aorta is well known (Eyster and Hooker, 1908: Anrep and Segall, 1926), and recently the carotid sinus has been shown to be an important zone, reflexly controlling heart rate and vasomotor tonus (Heymans, 1929). The physiological stimulus is apparently an alteration of pressure within this vessel. The present paper is concerned with an attempt to ascertain whether similar alterations of pressure within the gill vessels of elasmobranchs are effective cardio-inhibitory stimuli. MATERIAL AND Mi i HOD Specimens of the dogfish, St/nalns acanthias, 600 to 1500 grams in weight, taken during the month of August from Frenchman Bay, Maine, were used. The spinal cord was pithed posteriorly from the level of the sixth vertebra. The fish was secured ventral side up in a shallow tank of sea water. Although respiration continued in an apparently normal way, perfusion of the gills with sea water through the mouth was carried out in most cases, in order to insure an adequate supply of oxygen. The ventral aorta was exposed and ligated between it^ first and second branches. The former give rise to the first and second afferent branchial arteries which supply the hemibranch and the first holobranch. Therefore the second, third and fourth holo- br.iiH -hs were functional. Changes of pressure in the first branches and i heir derivatives were effected through a cannula inserted in the \rniral aorta, anterior to the ligature, and connected with a burette 10 CARDIAC INHIBITION FROM GILL VESSELS 11 tilled with a physiological solution (urea-saline, Lutz, 1930/7). The heart beat was recorded on a smoked drum by means of a heart lever or a mercury manometer recording ventral aortic pressure through a cannula inserted posteriorly to the ligature. RESULTS Cardiac inhibition was obtained when the first ventral aortic branches, and thus the gill vessels arising from them, were suddenly subjected to increases of internal pressure applied by opening the connection between the cannula and the burette for varying periods (Fig. 1, A). The amount of fluid entering during the application of pressure varied from 0.2 cc. to 5 cc., depending upon the height at which the burette was set and the length of time it was held open. FIG. 1. Reflex cardiac inhibition following sudden increase of pressure within the gill vessels at P. Heart beat recorded by lever. In this and in subsequent figures the time record shows five-second intervals. Cord pithed posteriorly from the level of the sixth vertebra. Urea-saline solution in pressure burette except in B. A. Increase of pressure in the first branches of the ventral aorta. B. Same, with dogfish blood in the pressure burette. C. Same, with gill capillaries occluded with cornstarch so that no fluid entered. D. Increase of pressure in the third, fourth and fifth branchial arteries. Inhibition was invariably obtained when the pressure applied (50 to 125 cm. of urea-saline) was obviously much higher than the dorsal aortic blood pressure. Simultaneous measurements of the ventral and dorsal aortic blood pressures showed that the pressure differences between the two aortae averaged 16.1 mm. Hg for systolic pressure and 4.3 mm. Hg for diastolic pressure, and that the pulse wave is transmitted through the gill capillaries to the dorsal aorta (Fig. 2, A). It is apparent, therefore, that the gill capillaries are relatively wride, and that the pressure in the afferent aortic system beyond the ligature is essentially the same as that in the dorsal aorta. Dorsal aortic systolic pressure ranged from 11 to 28 mm. Hg (8 12 B. R. LUTZ AND L. C. WYMAX measurements), and it was noted that fluid would not enter the ventral aorta from the burette below 25 cm. urea-saline (18.6 mrn. Hg\ A sudden increase of pressure averaging 10.7 mm. Hg above the average -i"lic pressure existing in the dorsal aorta was found to constitute an adequate stimulus for cardiac inhibition (Fig. 2, B), and in one instance a difference as low as 3 mm. Hg produced the response (Fig. 2, C). Although urea-saline is considered to be a physiological solution, a control experiment was performed using heparinized fresh dogfish blood in the burette and its connections. A similar inhibitory response to increased pressure in the gill vessels was obtained (Fig. 1, B). I;u;. 2. Ilcartlic.it recorded by Hg manometer. Figures at left show pressure levels in mm. 11^. A. Simultaneous records of ventral (upper) and dorsal (lower) . ">rtic blood pressure. H. Ventral aortic blood pressure record. Threshold increase of pressure in tlie first branches of the \cntral aorta at T (33 cm. urea-saline, 24.6 nun. I Ig). C. Same, threshold 25 cm. urea-saline (IS. 6 mm. 1 Ig). D. Ventral aortic blood pressure record. Spontaneous ejection reflexes at R. Cardiac inhibition was obtained when the burette was opened and immediately closed, allowing only 0.2 cc. of fluid to enter. Further- more, holding the burette open, until 5 cc. of fluid had entered the ventral aorta, did not increase or continue the initial inhibition during the period of flow. It follows that the effective stimulus is the initial increase of pressure per se, and not flow of fluid through the gill vessels. This was substantiated by blocking the flow through the gill capillaries so that the application of pressure alone served to stimulate. At- tempts to do this by tying the efferent branchial supply from the hemibranch and first holobranch were unsuccessful because of the numerous anastomoses between these vessels and the neighboring efferent system. An effective block on the afferent side was obtained CARDIAC INHIBITION FROM GILL VESSELS 13 by occluding the gill capillaries with cornstarch, injected in suspension in urea-saline solution, through the ventral aorta. This suspension was withdrawn from the larger afferent vessels with a syringe and the vessels were refilled with urea-saline, thus giving a vascular pocket on the afferent side against which pressure could be applied without passage of fluid (Fig. 1, C). That the receptors for the cardio-inhibitory response to increase of pressure within the gill vessels are not limited to the first two branchial arteries was shown by an experiment in which the ventral aorta was ligated between its first and second branches, and pressure applied to the third, fourth and fifth branchial arteries through a cannula inserted through the conus arteriosus (Fig. 1, D). FIG. 3. Sudden increase of pressure within the first branches of the ventral aorta at P. Heart beat recorded by lever. A, before and B, after cutting the vagus supply to the heart on both sides. C. Vagus supply to heart cut on right side and increase of pressure in right gill vessels (crossed reflex). D, increase of pressure in right gill vessels, and E, in left gill vessels after cutting cranial nerves from fifth through tenth on right side (afferent supply from gills). The cardiac inhibition following increase of pressure within the gill vessels is a reflex response inasmuch as it disappeared when the brain was destroyed by pithing. When the ramus posttrematicus of the fourth branchial division of the vagus (5th branchial nerve) and the ramus visceralis were cut on both sides the reflex was also abolished, thus showing that the efferent side of the reflex is in the vagus supply to the heart (Fig. 3, A, B). When these nerves were cut on the right side and the first right branch of the ventral aorta was clamped, thus leaving only the left gill vessels open to stimulation, the usual inhibition was obtained. When the clamp was shifted to the first left branch of the ventral aorta, the response was also obtained upon stimulating the right gill vessels (Fig. 3, C). Evidently the pathways for this reflex are both unilateral and crossed in the central nervous system. The threshold of pressure for the crossed reflex was about 10 cm. of urea-saline higher than that for the reflex obtained from stimulation of both branches. 14 D. R. LUTZ AX I) L. C WVMAX The afferent innervation of the hemibranch and the first two gills was interrupted on both sides by cutting the first three branchial nerves (ninth and the first and second branchial divisions of the vagus). Following this no cardiac response to increase of pressure in both first branches of the ventral aorta was obtained. The integrity of the remaining cardie-inhibitory reflex mechanism was demonstrated by obtaining reflex inhibition upon pinching with forceps the skin between the fourth and fifth gill slits, the snout, the base of the pectoral fin, or upon cutting open the abdomen or scratching the wall of the pericardial chamber with a needle. The afferent side of the reflex following increase of pressure within the anterior gill vessels is thus located. Inasmuch as cutting the first three branchial nerves involves opening the anterior cardinal sinus with considerable hemorrhage, the afferent nerve supply from the gills was interrupted on one side by transecting the cranial nerves from the fifth through the tenth close to the brain with a single incision through the skin. Following this, the usual cardio-inhibitory reflex was not obtained from the gill vessels on the operated side. Mechanical stimulation of the gill region on this side also failed to elicit a reflex, but both increase of internal vascular pressure and mechanical stimulation of the gills on the intact side evoked the cardio-inhibitory response (Fig. 3, D, E). DISCUSSION The blood pressure in the elasmobranch has been found to be of the low pressure type. Thus Schoenlein (1895) found in Torpedo a branchial systolic pressure of 16 to 18 mm. Hg and in Scyllium one of 30 to 33 mm. Hg; Hyde (1908) found the mean pressure in a branch of the aorta of the skate to be 20 mm. Hg; Lyon (1926) found in sand sharks an average branchial pressure of 32 mm. Hg and an average dorsal or systemic pressure of 23.3 mm. Hg. In Sqiiahts acanthias \\e found the average pressure in the ventral aorta to be 28.2 mm. Hg and th.it in the dorsal aorta to be 15.4 mm. Hg. The average pulse pressure in the ventral aorta, as recorded by the mercury manometer, \\.is 1 3.3 mm. I k \\ Inch is 47.2 per cent of the average systolic pic-- ure. The ventricular beat as a factor in maintaining blood pressure is obviously important. This is in accordance \\ith the fact that no \.i-omotor innervation has been demonstrated in these fishes. The heart reflexes, therefore, must be of importance as regulatory factors lor the \.iscular needs of the body. The cardio-inhibitory reflex, as -hown above, can be elicited by an increase in gill blood pressure \\ell within physiological limits, and may temporarily decrease the CARDIAC INHIBITION FROM GILL VESSELS 15 diastolic blood pressure to a significant degree. In one case this decrease was 4 mm. Hg, which was 50 per cent of the diastolic pressure. Such a mechanism might come into play during an ejection reflex, when the sudden constriction of the branchial muscles forces water from the pharynx. The accompanying external pressure on the gill vessels would force blood from them in both directions. The blood pressure in the ventral aorta would rise, inasmuch as the blood would meet the valves in the conus arteriosus, evoking a sudden need for cardiac inhibition to prevent dangerous consequences, such as injury to the thin-walled afferent system. The increase in internal pressure would evoke the reflex, momentarily preventing additional pressure increase due to the ventricular action. As a matter of fact, our records show an increase of ventral aortic blood pressure well above threshold value followed by cardiac inhibition during spontaneous ejection reflexes (Fig. 2, D). This mechanism, therefore, being of physiological significance, may be compared with the carotid sinus mechanism in mammals. Cardiac inhibition during an ejection reflex may also be evoked by the mechanical stimulation of the gill region, inasmuch as the receptors for the heart reflexes have a wide distribution in the elasmobranch. It is conceivable that in the course of evolution the widespread sensitive areas of the ancestral form, possibly typified by the elasmo- branch, were concentrated or delimited until the condition seen in the mammal was reached. The carotid arteries of the mammal are derivatives of the primitive branchial system. The cardio-inhibitory reflex following increase of pressure within the gill vessels in elasmo- branchs may exemplify, therefore, the evolutionary forerunner of the carotid sinus mechanism of mammals as it existed in whatever may have been the ancestral form. This is one of many instances in which it is apparent that physiological as well as morphological factors should be considered in evolutionary reasoning. SUMMARY 1. Cardiac inhibition follows sudden increase of pressure within the gill blood vessels of Squalus acanthias. An average increase of 10.7 mm. Hg above the average systolic pressure existing in the dorsal aorta constitutes an adequate stimulus for the inhibitory response. Respiratory reflexes were frequently seen associated with the cardio-inhibitory reflex induced by changes in blood pressure. Inasmuch as the carotid-sinus mecha- nism in mammals has recently been shown to be concerned in the regulation of respiration, further work on the peripheral control of respiration through receptors within the gill vessels of elasmobranchs is in progress. 16 B. R. LUTZ AND L. C. WYMAN 2. The cardio-inhibitory response is a reflex, with afferent pathways located in the branchial nerves and efferent pathways in the vagus supply to the heart. The reflex is both unilateral and crossed. 3. The average ventral aortic systolic blood pressure in Squalus acanthias is found to be 28.2 mm. Hg, the average dorsal aortic systolic pressure 15.4 mm. Hg, and the average ventral aortic pulse pressure 13.3 mm. Hg. The inhibitory reflex to increased ventral aortic pressure modifies the diastolic blood pressure to a significant degree. 4. The adaptive nature of the reflex is pointed out and its phylo- genetic significance is discussed. BIBLIOGRAPHY ANREP, G. V., AND II. V N.< .ALL, 1926. Jour. PhysioL, 61: 215. EYSTKR, J. A. 1C., AND 1). R. HOOKKR, 1(>08. Am. 'jour. PhysioL, 21: 373. HEYMANS, C., 1929. Le sinus carotidicn. Louvain and Paris. HYDE, I. II., 1908. Am. Jour. PhysioL, 23: 201. LUTZ, B. R., 1930o. Biol. Bull., 59: 170. LUTZ, B. R., 1930/;. Am. Jour. PhysioL, 94: 135. LYON, E. P., 1926. Jour. Gen. PhysioL, 8: 279. SCHOKNLEIN, K., 1895. Zeitschr. f. Biol., 32: 511. THE EFFECT OF ADRENALIN ON THE BLOOD PRESSURE OF THE ELASMOBRANCH, SQUALUS ACANTHI AS LELAND C. WYMAN AND BRENTON R. LUTZ (From the Mount Desert Island Biological Laboratory and the Physiological Laboratory of Boston University School of Medicine) In elasmobranch fishes the distribution and relative preponderance of the components of the autonomic nervous system and the effects of adrenalin on the heart and digestive tract differ considerably from the same factors in mammals, as pointed out by Lutz (1930a, 1930&, 1931). The effect of adrenalin (1 : 50,000 to 1 : 25,000) on the heart was found to be inhibitory and was interpreted as the response of an unbalanced parasympathetic mechanism in an organ lacking a sympa- thetic accelerator innervation (Lutz, 1930a). The well-developed chromaphil system seen in elasmobranchs (Lutz and Wyman, 1927) leads one to believe that it must be of functional significance, and a possible inhibitory emergency theory has been suggested to account for its existence (Lutz, 1930o). The lack of accelerator nerves to the heart and the failure to demonstrate a vasomotor innervation to the blood vessels makes the abundance of chromaphil tissue in these fishes of especial interest to the physiologist. The well-known effect of adrenalin on the blood pressure of mammals, an effect which is relatively brief and readily duplicated by successive doses, appears to differ from its effect on the blood pressure of cold-blooded animals. Bieter and Scott (1929) reported a rise of blood pressure in the frog following a dose of 0.2 cc. of epinephrine hydrochloride, 1 : 10,000, which persisted for at least one and one-quarter hours after the injection. MacKay (1931) has found that adrenalin causes a rise of ventral aortic blood pressure in skates (Raia), which is of much longer duration than the pressor effect in mammals. Herein are reported the results of a preliminary investigation of the effects of adrenalin on the vascular system of the elasmobranch, Squalus acanthias. In view of the system of branchial capillaries located between the ventral and dorsal aortse, simultaneous records of the blood pressures in both systems were considered essential. MATERIAL AND METHOD Specimens of the dogfish, Squahis acanthias, 730 to 1400 grams in weight, taken from Frenchman Bay, Maine, were used. The spinal cord was pithed posteriorly from the level of the sixth vertebra. 17 18 L. C. WYMAN AND B. R. LUTZ The fish was secured in a tank of sea water and the gills were perfused through the mouth as described in a previous paper (Lutz and \Yyman, 1931). Simultaneous ventral and dorsal aortic blood pressures were recorded on a smoked drum by means of two mercury manometers and two cannulas, one inserted in one of the first branches of the ventral aorta thus leaving the second, third and fourth holobranchs on that side and all the gills on the other side for respiration, and the other inserted in the coeliac artery close to the dorsal aorta. The cannulas and manometer connections were filled with a physiological solution (urea-saline, Lutz, 1930a) containing a little heparin to prevent clotting. Adrenalin chloride (Parke, Davis and Co.), diluted in urea- TABLE I Effect of Initial Doses of Adrenalin Exp. no. Dose of adrenalin Increase of pressure Ventral aorta Increase of pressure Dorsal aorta Chance of heart rate Diast. Pulsi Syst. Diast. 30 31 35 36 27 29 33 34 37 1 cc 1000 .... per cent 71.5 29.1 70.0 55.6 39.2 60.0 32.1 33.3 per cent 40.0 30.8 46.1 13.3 38.5 33.3 20.0 14.3 42.8 per cent 150.0 27.8 114.1 233.0 40.0 100.0 63.6 60.0 23.1 per cent 75.0 53.8 91.0 111.8 66.5 50.0 45.5 91.6 per cent 72.8 80.0 133.3 50.0 36.3 111.0 beats per mill. -6 +4 0 -30 0 -4 0 -4 +4 | , , ]()()() ' cc 1000 .... ' cc 1000 > cc 10 000 2 cr 10 000 2 cc 50 000 2 cc 1 100 000 2 cc 1 500 000 . . . 33.3 saline solution, was injected into the portal vein. MacKay (1931) has shown that in the skate the effects of adrenalin administered in this way are the same as those obtained by injection into a large systemic vein. Control injections of urea-saline solution and adequate controls for manipulation during injections were carried out. RESULTS Intravenous injections of adrenalin, in doses of one or (wo cc. of 1 : 1000 to 1 : 500, 000, produced rises of ventral and dorsal aortic blood pressure, both systolic and diastolic, persisting for at least 30 minutes (Table I; Fig. 1, A). No attempt was made to determine the maximum duration of the prcssor effect, but MacKay (.1931) has found it to be' from one to two and one-half hours in the skate. There was also a marked increase of pulse pressure. The effect on the heart rate was not constant, but there was a general tendency ADRENALIN AND BLOOD PRESSURE OF SQUALUS 19 20 L. C. WYMAN AND B. R. LUTZ toward a decrease, especially following the stronger doses (1 : 1000, 1 : 10,000). Subsequent doses of adrenalin following doses stronger than 1 : 500,000 produced small temporary increases of systolic pressure with little or no effect on the diastolic pressure, and consistent small increases of pulse pressure (Table II; Fig. 1, C). The rise of systolic TABLE II Effect of Siibsequent Doses of Adrenalin (For initial doses see Table I) Ex p. no. Dose of adrenalin Time II.l first dose Increase of pressure Ventral aorta Increase of pressure Dorsal aorta Change of heart rate Syst. Diast. Pulse Syst. Diast. 27 29 33 34 37 1 cc. 1 1000 win. 33 41 51 7.5 10 3 5 per cent 1.8 2.7 11.8 2.6 5.7 3.1 3.1 8.8 20.0 16.1 12.5 per cent 11.1 0 4.8 0 16.1 0 0 6.2 11.1 12.5 12.5 per cent 4.3 8.3 23.0 5.9 — 5.9 6.2 6.2 11.1 33.3 20.0 12.5 per cent 17.1 3.8 4.5 4.4 15.0 6.2 6.7 12.5 33.3 37.4 41.1 per cent 16.7 0 0 23.5 6.7 0 13.3 12.5 33.3 23.1 beats per min. -2 0 _2 -6 0 + 2 -4 -4 0 0 -6 2 cc. 1 10,000 2 cc. 1 10,000 2 cc. 1 10,000 2 cc. 1 50 000 2 cc. 1 100 000 2 cc. 1 100 000 8.5 11.5 9.5 22.5 28.5 2 cc. 1 1000 2 cc. 1 500,000 2 cc. 1 500 000 2 cc. 1 1000 pressure was, in most cases, actually due to the increase of pulse pressure. The effect of subsequent doses on the heart rate, however, was similar to that of initial doses. Control injections of 2 cc. of urea-saline solution gave results which were essentially similar to those following subsequent doses of adrenalin with respect to blood pressure and pulse pressure. The heart rate, however, was unaltered (Table III; Fig. 1, D). It is probable, therefore, that such subsequent doses of adrenalin have little or no effect on the blood pressure. After a dose <>l two cc. of adrenalin, 1 : 500,000, which gave a less prolonged pressoi dlcct than stronger initial doses, subsequent doses of the same size were effective, producing significant increases of blood pressures and of pulse pressure which again were of shorter duration than the pressor effects of stronger doses (Table II, Exp. 37). This suggests that the ineffectiveness of adrenalin given during the pro- longed pressor effect following stronger doses is due to the existence of the maximum pressor action of which the vessels are capable. Such ADRENALIN AND BLOOD PRESSURE OF SQUALUS 21 an explanation obviates the necessity of assuming that a pharmaco- logical tolerance to adrenalin has been acquired. Excluding cardiac effects, a change in ventral aortic pressure following the administration of adrenalin might be due to alterations in either the gill vessels or in the systemic vessels. The increase of blood pressure following initial doses of adrenalin was often associated with slowing or with no change in heart rate. The percentage increase of diastolic pressure in the dorsal aorta was consistently greater than that in the ventral aorta (Table I). These facts suggest a vaso- constrictor action of adrenalin, peripheral to the gill capillaries. Neither recording the venous outflow from the excised spiral valve (the blood vessels of which were perfused with adrenalin solutions) nor microscopic observation of the minute vessels of the tail, during TABLE III Effect of Control Injections of Urea-Saline Solution Increase of pressure Increase of pressure Ventral aorta Dorsal aorta Change Exp. no. Dose of heart rate Syst. Diast. Pulse Syst. Diast. per rent per cent per cent per cent per cent beats per min. 31 2 CC. -3.1 0 -5.2 0 0 0 33 2 cc. 16.7 5.0 40.0 17.6 12.5 0 33 2 cc. 9.6 5.0 18.2 10.5 5.5 0 34 2 cc. 4.2 -12.5 37.5 9.1 — 0 37 2 cc. 11.5 0 23.1 9.1 0 0 37 2 cc. 3.1 0 6.2 5.8 0 0 the injection or direct application of adrenalin, gave evidence of vasoconstriction. Further work is necessary to locate the region of action of adrenalin in producing the pressor effect in Squahis acanthias. There is considerable doubt concerning the existence of a sympa- thetic vasoconstrictor innervation in elasmobranchs. The long sus- tained pressor action of small amounts of adrenalin and the accelerator and augmentor action of minute doses on the heart of the skate (Huntsman, 1931) suggest that a possible function of the well-devel- oped chromaffin system is to take the place of sympathetic nervous factors in vascular regulation. However, the discovery by MacKay (1931) of pressor changes in the ventral aorta following sensory stimulation in spinal or anesthetized skates, which were too brief to be due to reflex discharge of adrenin, invites further search for vaso- constrictor nerves. L. C. \VVMAX AND B. R. LUTZ SUMMARY 1. Intravenous injection of adrenalin, in doses as low as 2 cc. of 1 : 500,000, produced long-sustained pressor effects in Squalus acan- thias, together with marked increase of pulse pressure and a tendency toward decrease of heart rate. 2. Subsequent doses of adrenalin following doses stronger than 1 : 500,000 were ineffective. This is interpreted as being due to already existing maximum pressor action. Doses subsequent to a dose of 2 cc. of 1 : 500,000 were effective. 3. The pressor effect is interpreted as being due to extra-cardiac factors, peripheral to the gill capillaries, but the region of action of the adrenalin was not located. BIBLIOGRAPHY BIETER, R. X., AND K. II. SCOTT, 1929. A ni. Jour. Pliysiol., 91: 265. HUNTSMAN, M. E., 1931. Conlr. Can. Biol. and Fish., X.S., 7: 33. I.i i/, B. K., 1930o. Am. Jour. Phvsiol., 94: 135. I i i/, I'-. R., 19306. Bwl. Hull., 59: 211. I.i i/, B. R., 1931. Biol. Bull., 61: 93. I i \7., B. R., AND L. C. \\YMAN, 1927. Jour. F.xper. Zool, 47: 295. LUTX, B. R., AND L. C. \YYMAX, 1931. Biol. Bull., 61: 217. M.U-KAY, M. E.f 1931. Contr. Can. Biol. and Fish., X.S., 7: 19. PROPERTIES OF THE BLOOD OF THE SKATE (RAIA OSCILLATA) D. B. DILL, H. T. EDWARDS AND M. FLORKIN » (From the Fatigue Laboratory, Morgan Hall, Harvard University, and the Marine Biological Laboratory, Woods Hole, Mass.) The system developed in the skate for transport of oxygen and carbon dioxide is similar in some respects to that of man. Oxygen combines reversibly with hemoglobin and probably passes in and out of the blood by diffusion. Carbon dioxide, transported as bi- carbonate, is prevented from greatly changing the blood reaction by the buffer function of proteins. If one seeks a more detailed description of the skate's respiratory system, it is found to differ from that of man in many respects. Thus a labile body temperature introduces a degree of freedom in the skate's blood which is absent in the blood of normal man. Other differences depend upon the physical state of the environment. In the one case the blood is separated by a membrane from a moving liquid. From this liquid, oxygen in solution diffuses into the blood stream; into it, carbon dioxide passes directly. In the other case the lung, acting as a buffer between the blood and a variable external environment, maintains air nearly constant in temperature and composition in the ultimate areas where gas exchange takes place. Carbon dioxide passes from the blood, not into a virtual vacuum, as from the gills of the skate, but into a gas phase where the partial pressure of carbon dioxide fluctuates within narrow limits about a mean value of 40 mm. So much can be said by induction; it remains to be seen how well the properties of the skate's blood are adapted to the requirements. The experimental methods used in this investigation have been described by Dill and Edwards (1931) and need not be discussed here. Two difficulties arose, both related to the character of the erythrocytes. These are very resistant to rupture and in determining oxygen content low values may be obtained because of incomplete hemolysis. When the quantity of saponin in the ferricyanide reagent was tripled, hemolysis was complete and oxygen could be determined satisfactorily. The other difficulty was a consequence of the high metabolic rate of these nucleated cells. After blood is equilibrated and sealed in sampling tubes a rapid decrease in its oxygen content occurs. Usually 1 Fellow of the C. R. B. Educational Foundation. 23 24 DILL, EDWARDS AND FLORKIN oxygen was determined immediately but occasionally delay was unavoidable. In order to correct for changes in oxygen content, the rate of oxygen consumption was determined at several temperatures. Since these observations on metabolic rate of the blood may be of some interest in themselves they have been tabulated in Table I. In this experiment the blood had been saturated at a temperature of 10.5° with an oxygen partial pressure of 200 millimeters. Assuming the same solubility as in human blood of the same water content, dissolved oxygen amounted to 0.82 volume per cent and combined oxygen, 4.74 volumes per cent. With this information and the data of Table I and of Fig. 1 on the metabolic rates at 0, 20 and 40°, one can calculate the oxygen consumption of any specimen of blood TAHLK I Oxygen Consumption by Skate's Blood Inithl oxygen content = 5.56 volumes per cent Time Oxygen Content min. rot. per cent 0° C. 20° C. 40° C. 21 4.70 96 1-sx: 115 1.64 140 3.70 375 1-37 1000 4.29 within a wide range of conditions. It appears from Fig. 2 that the metabolic rate is not a linear function of the reciprocal of the absolute temperature. The observations are too few in number, however, to define this curve precisely. Observations have been made on the oxygen dissociation curves of three specimens of blood (two of which were composited from several skates) at four temperatures. A preliminary experiment was carried out in the usual way with variable oxygen pressures and with carbon dioxide partial pressures ranging from 10 to 100 millimeters. Since the blood has about one-half the buffer value of human blood it was supposed that this range of carbon dioxide pressures would have a greater effect on the position of the oxygen curves than in human blood. On the contrary, the effect was too small to be evaluated. A second experiment (Blood B; temperature 10.4°; Table II) was carried out with a greater range in carbon dioxide pressures, — from 0.5 to 140 millimeters. With this extreme range there appeared to be a small decrease in affinity for oxygen with increasing acidity. PROPERTIES OF SKATE'S BLOOD 25 TABLE II Eqiiililirinin Values for Oxygen Absorption pCO2 pOs Total O» HbOj mm. II a mm. Hg vol. per cent per cent of capacity Blood A. HbO2 capacity 5.68 vol. per cent 0.8 12.5 0.22 4 Temperature 37.5° 1.0 48.2 1.54 25 0.8 128 3.97 64 1.3 248 5.29 82 Temperature 25° 0.8 14.6 0.41 6 0.7 31.3 1.79 30 0.7 84.6 4.53 75 1.0 100 4.55 75 Temperature 0.2° 0.5 2.4 0.60 10 0.6 10.8 2.80 48 0.5 13.8 3.65 62 0.7 24.3 4.94 85 0.7 48.6 5.68 96 1.5 150 6.43 100 Blood B. HbO2 capacity 0.7 3.6 0.34 6 5.54 vol. per cent 0.7 5.3 0.32 6 Temperature 10.4° 0.5 8.1 1.00 18 1.5 16 2.20 39 0.5 26.5 3.62 64 0.6 40.1 4.56 79 0.8 58.8 5.15 89 1.1 76.8 5.54 95 2.3 190 6.30 100 10.7 25.8 2.68 47 140 23.2 2.29 40 141 33.0 3.07 53 137 40.7 3.41 59 Blood C. HbO2 capacity 4.18 vol. per cent 2.2 108 2.60 56 Temperature 37.5° 2.7 254 4.13 84 Temperature 25° 0.6 23.2 1.24 28 0.5 48.9 2.50 56 0.5 83.1 3.19 70 0.6 108 4.04 89 1.0 165 4.49 95 1.6 212 4.75 98 Temperature 10.4° 0.9 18.8 2.49 58 0.5 22.3 2.42 56 0.6 38.9 3.14 71 7.2 167 4.95 100 26 DILL, EDWARDS AND FLORKIN This effect is represented quantitatively in Fig. 3. Ordinarily it is convenient to express the relation between position of the oxygen dissociation curve and pH by such an equation as log = fl(pH) + b, where A'so represents the partial pressure of oxygen when the blood is 50 per cent saturated with oxygen, and the term a, the slope, is a measure of the rate of change of affinity of the blood for oxygen with rate of change in pH. This equation cannot be applied to 45 4.0 -03.5 o _o CO c> (J o o O. o o '40° -0.5 .0032 .0033 .0034 .0035 .0036 .00(37 1/1 100 150 200 250 TIME m Minutes 300 350 400 450 FIG. 1. Rate of oxygen consumption of normal skate's blood at temperatures of 0, 20, and 40°. 1 i<;. 2. (Inset.) Logarithm of the metabolic rate of skate's blood (oxygen used in cc. per hr. per 100 cc. blood) as a function of the reciprocal of tin .tl»olute trmpcr.it lire. our data on skate's blood since we do not have direct determinations of pH nor knowledge of the value of pK' for scrum or for cells. For Min- purposes, however, it is enough to use calculated values for [(BHCO8y(H2CO8)6]. This quantity differs from pH by a PROPERTIES OF SKATE'S BLOOD 27 constant (or nearly constant) amount and hence the slope a will be essentially the same as though pH values were used. The curves for blood of the crocodile and of man shown in this figure are from Dill and Edwards (1931). The curves for blood of the crocodile and for man are much steeper than for that of skate's blood. The contrast is greatest in the acid range; in fact writhin the range which corresponds to pCO2 values from 10 to 140 mm. there is almost no change in position with change in reaction. However, the skate does not normally function within these limits but, as will be shown below, at a carbon dioxide pressure 1.7 i.e °^i «; o 1-D 10 II <^ <""i fi 1 /I £ !l.3 Q. o 1.1 1.0 0. . 3. «i \ \ \CRO( \ ^ODILE V \ ^5r . SKATE ^\ \ \ \ \ "^ v X X X X X .MAN ^ ^\ X X X \ \ \ \ X X s X X \ \ \ \ \ 5 2 0.4 0.6 0.8 1.0 1.2 14 1.6 1. Log [feHCO^b/tHaCO^b] when HbOa^SO^o Position of the oxygen dissociation curves (log p()^ when HI FIG. per cent) in relation to log (BHCO3)/(H2CO3). = 50 of one or two millimeters. In this range there is a distinct acid effect on oxygen affinity; here the slope a is roughly one-half as great as in man and one-fourth as great as in the crocodile. The blood of the skate is like that of Urechis, studied by Redfield and Florkin (1931), in the acid range but quite different when the reaction is more alkaline. They have found that the affinity of Urechis blood for oxygen is unaffected by change in hydrogen ion concentration when pCO2 is varied from 0.5 to 92 millimeters. It would be interesting to speculate on the possible significance of these relationships in connection with structure of the hemoglobin molecule. However, we must remember 2s DILL, EDWARDS AND FLORKIX that the environment of hemoglobin is very different in these cases. It will be recalled that an abnormal value for a similar relationship was found in man in diabetic coma (Dill and others, 1929). Further di>< -ussion of this question had better wait, therefore, until it is possible to prepare these hemoglobins in the pure state and study them under strictly comparable conditions. The data given in Table II have been used to construct the oxygen dissociation curves of Fig. 4. These have been drawn as members of the same family of curves, and aside from a few bad results the fit accords with this assumption. The effect of temperature on the affinity of blood for oxygen has been shown graphically for human blood by Brown and Hill (1923). Use has been made of their data and that of Fig. 4 in Fig. 5. This comparison indicates that q, the heat of reaction of 1 gram molecule of hemoglobin with ;/ gram mole- 60 80 100 120 140 160 p02 in rnm.Hcj ISO 200 220 Z40 260 1;1G. 4. Oxygen ('d, =-- 1 ±0.5 mm. rules of oxygen, is the same for the blood of the skate and of man. The significance of n remains somewhat obscure, but it is useful in characteri/ing the slope of the oxygen dissociation curve when log p().j = I log Hb'HbO-j. It has the value of 2.2 in man and in skate's blood it is slightly smaller, i'iz., 2.0. If we accept the definition ..I ;/ ui\en by Hrown and Hill, it follows that the value for (J, the heat evoKfd \\lirn 1 grain molecule of oxygen combines at constant \olume with hemoglobin, is about the same as in human blood. ( MM- miM have a description of the carbonic arid dissociation curve of blood in order to understand the conditions under which carbon dioxid- is excreted through the gills. It is known from the work of Collip (1920), confirmed by others (Kokubo, 1927 and Smith, \(>2{>t. that the carbonic acid content of selachian blood is no more than 10 to 12 volumes per cent. Accordingly the carbonic acid 1'UOI'KRTIES OF SKATE'S BLOOD 29 dissociation curves have been studied over a low range of partial pressures. The curves for oxygenated blood are shown for tempera- tures of 10.4° (body temperature2) 25° and 37.5° in Fig. 6. The experimental procedure was simplified by the fact that, as in human blood, the relation of log (pC()2) to log (Total CO2) is linear or nearly so. Hence it was only necessary to determine a few points and fit the best straight line. The smoothed values were then transformed to the more familiar system of coordinates used in Fig. 6. The alkaline reserve of blood, as suggested by Van Slyke and Cullen (1917), is most accurately defined by the bicarbonate content of arterial blood. When this is impractical, their method of equi- librating plasma of venous blood with alveolar air gives approximately correct values for human blood. The application of this method to fish blood, as by Collip (1920) and Kokubo (1927) does not define the alkaline reserve of fish blood but merely the carbon dioxide .0031 .0032 .0033 .0034 .0035 .0036 .0037 1 XT- FIG. 5. Position of oxygen dissociation curves (log pC>2 when HbOa = 50 per cent) as a function of the reciprocal of the absolute temperature. The data for man are from Brown and Hill (1922-3) and for Urechis, from Redfield and Florkin (1931). In each case the oxygen is expressed in terms of partial pressures at the actual temperatures involved. The partial pressures of carbon dioxide were approximately 40 mm. for man, 1 for the skate and 12 for Urechis. content at a partial pressure of carbon dioxide which is possibly twenty times greater than that of blood in vivo. It is probable that the values given by Collip for carbon dioxide content of blood equi- librated with atmospheric air measure the alkaline reserve more accurately than when alveolar air was used. Having attained a description of oxygenated blood there remained to determine the effect of oxygenation on the position of these curves. There are several a priori reasons for supposing that this effect is too 2 Body temperature was observed by rectum. It was maintained approximately constant by circulating sea water. Blood drawn at 10.4° and equilibrated at 25° and at 37.5° does not necessarily reflect the properties of blood drawn from animals acclimated to these higher temperatures. Possibly the blood would be altered in respect to available base and in other respects by change in body temperature. 30 DILL, EDWARDS AND FLORKIX small to be measurable in skate's blood. These are: (a) the hemo- globin is one-fourth its concentration in normal human blood and the effect of oxygenation on the carbonic acid dissociation curves of blood will be reduced accordingly; (6) the alkaline reserve is lower and the distance between the curves will on this account be low, as may be Total C02 Vol.% ^^ 20 16 12 10 8 6 7 7 0 2 4 6 8 10 12 pC02 in mm. Hg FIG. 6. Carbonic acid dissociation curves of skate's blood, body temperature 10.4°; equilibration temperatures, 10.4, ?5 and 37.5°. The broken line corresponds to human blood at 37.5° in terminal chronic nephritis. seen from the curves for human blood in diabetic coma (Dill and others, 1()29); (c) the effect of acid on the oxygen dissociation curves is much less in skate's blood than in human blood (see Fig. 3); accordingly PROPERTIES OF SKATE'S BLOOD 31 the effect of oxygen on the base bound by hemoglobin should be correspondingly less. For all these reasons taken together it would appear that oxygenation should have little effect on the carbonic acid dissociation curves and in fact several experiments, including one on concentrated blood, revealed no significant difference between the curves of oxygenated and of reduced blood. It appears, then, that in the skate carbon dioxide is transported principally by virtue of buffering properties of blood proteins and we shall now direct our attention to that subject. It will be convenient to consider first the buffer value of separated plasma. The results of experiments in which the carbonic acid dissociation curves were used to calculate buffer value of plasma specimens are shown graphically CO 24 dJ Q_ *= ^ ^^ ^ ^ ^ SUBJECT o MAN PROTEIN UREA NaCI -J= nd(BHCPJ PROTEIN d(pH) 65 0.5 6 0.103 a HAN 65 16 17 0.108 * SKATE Z4 15 17 0.193 o" * : 5 KATE 31 17 16 0.209 o CQ 8 4 °G «^ -^ "^5 4 6.6 6.8 7.0 7.2 7.4 7.6 7.8 8. pH FIG. 7. Buffer value of skate's plasma and human plasma compared. Enough salt and urea was added to one specimen of human plasma to simulate skate's plasma in respect to those constituents. Concentrations of protein, urea and salt are in grams per liter. in Fig. 7. Van Slyke's measure of buffer value is the rate of change of base bound with rate of change in pH : dB These curves define the buffering capacity of plasma, but since this is due principally to protein, it has been divided by the protein concen- tration in each case. The calculations inserted in Fig. 7 indicate an unusually high value for the buffer value of plasma protein of the skate. To determine the extent to which this is due to the high 32 DILL, EDWARDS AND FLORKIX concentrations of urea and of sodium chloride in skate's plasma, a human plasma was prepared with concentrations of these substances typical of the skate. The effect of this modification proved to be very small and it follows that the proteins of the skate's plasma have, per gram, about twice the intrinsic buffering value of human plasma proteins. It will be noted that the protein concentration in the skate's plasma is from one-third to one-half as great as in human plasma. In effect, then, the buffer value of plasma per unit volume is almost equal in the two species. It is possible to compare the buffer value of skate's whole blood with that of human blood by reference to the chart of Henderson and associates (1930, Fig. 3). The observations given in Table III are TABLE III Buffer \'tilite of Oxygenated Skate's Blood at 37.5° • COn Capacity AC0260_M Blood r 111)0, < '.iji.irity at pCOj =40 mm. Observed Calculated from human blood * Ratio Normal . . mi'.i}. per I. 2.0 ml;.q. per I. 8.4 mi'.q. per I. 3.3 mEq. per I. 3.1 1.1 ( '(ilirrllt rated 4.7 7.0 3.15 3.2 1.0 * These values are calculated from the empirical chart of Henderson and associ- ates (1930, Fig. 3). based on one specimen of normal blood from the skate and one concen- trated specimen obtained by centrifuging normal blood and removing about one-half the plasma. The comparison indicates that skate's blood has about the same buffer value as human blood of the same oxygen- combining capacity. Since the ratio of cell proteins to oxygen- combining capacity is about one-half greater in the skate than in man, it appears that the buffer value of cell proteins per unit weight is much less in the skate than in man. This information regarding the physicochemical properties of ^k. ite's blood constitutes a suitable basis for study of changes in the respiratory cycle. The additional observations required are the oxygen and carbon dioxide contents of arterial and venous blood. \Ye succeeded in obtaining for this purpose blood from the dorsal aorta and from the conus arteriosus while sea water at body tempera- ture (9 to 10° C.) was being circulated over the gills. By application ol the data thus obtained to the carbonic acid and oxygen dissociation PROI'KRTIES OF SKATE'S BLOOD curves it is possible to calculate the partial pressure of carbon dioxide and of oxygen in the blood in vivo. This and other calculations are shown in Table IV where a comparison also is made between respira- tory changes in the skate and in man in terminal chronic nephritis (Henderson and others, 1927), a state which approximates in many respects to that of the normal skate. The data given in Table IV have several points of interest. The TABLE IV Comparison of the Skate with Man in Terminal Nephritis Skate * Man f Body temperature, ° C 10.4 37.5 Cell count, million per mm.3 0.2 1.0 Red cell volume, cc. per 100 cc 20.0 14.7 Oxygen capacity, vol. per cent 6.00 5.60 Free oxygen, arterial blood, vol. per cent 0.32" 0.28 Combined oxygen, arterial blood, vol. per cent 5.58 5.32 Combined oxygen, arterial blood, per cent of capacity 93 95 Free oxygen, venous blood, vol. per cent 0.07 0.06 Combined oxygen, venous blood, vol. per cent 1.91 1.62 Combined oxygen, venous blood, per cent of capacity 32 29 CO2 content, arterial blood, vol. per cent 7.70 8.61 COi content, venous blood, vol. per cent 10.84 11.75 CO2 transport, vol. per cent 3.14 3.14 O2 transport, vol. per cent 3.92 3.92 pCO2, arterial blood, mm. Hg 1.3 15 pCO2, venous blood, mm. Hg 2.6 23 pO2, arterial blood, mm. Hg 70 110 pO2, venous blood, mm. Hg 14 27 pHs, arterial 7.82 J 7.00 pHs, venous 7.67 J 6.95 ApHs 0.15 0.05 * While these data given for the skate are based on a specimen of arterial blood from one skate and of venous blood from another, observations on other individuals have verified the approximate accuracy of the figures given in the table. f These values have been obtained directly or by calculation from the study of terminal nephritis by Henderson and associates (1928). The values for carbon diox- ide and oxygen transport have been arbitrarily made equal to the observed values in the skate. Associated changes in dependent variables have been read from the alignment chart for blood in nephritis. | These values for plls are calculated on the assumption that pK's = 6.24 under the experimental conditions. The absolute values for pHs may be incorrect, but the value for ApHs will not be affected. striking difference as compared with man is in the carbon dioxide partial pressure. Man in terminal nephritis has a greatly increased rate of pulmonary ventilation but the partial pressure of carbon dioxide cannot be kept below 15 mm.; the skate, on the contrary, keeps the carbon dioxide pressure below 2 mm. in arterial blood. The steep character of the carbonic acid dissociation curve in this 34 DILL. EDWARDS AND FLORKIN range makes possible the transport of 3 volumes per cent of carbon dioxide with a change in its partial pressure of only 1.3 millimeters. Another point of interest is the large change in pH of skate's blood. This is no doubt related to the greater effect in human blood of oxygenation on base bound by hemoglobin. These facts are of particular interest in connection with the equilibrium between blood and air in respect to oxygen and carbon dioxide. The oxygen dissociation curve of skate's blood at 10.4° is approximately the same as man's at 37.5° and the oxygen tension in sea water is approximately the same as in air. The fact that arterial blood of the skate has about the same percentage saturation with oxygen as that of man indicates that the adequacy of oxygen transfer is approximately the same in the two species. It has been shown by Bock and Field (1924) that in man the carbonic acid pressure is about the same in alveolar air as in arterial blood, most of the differences in partial pressure being within ± 1 millimeter. It is now possible to say that the partial pressure of carbon dioxide in the arterial blood of the skate exceeds that in water passing over the gills by no more than 1 or 2 millimeters. Collip has suggested that it is possible that a steep pressure gradient exists "between the dissolved carbon dioxide in the blood on the one side, and in the sea water on the ether." Such may be true of some bony fishes but it is not true of the skate.3 His argument is as follows: "As the hydrogen ion concentration of sea water is in most instances lower than that obtaining in the blood of marine forms and as the bicarbonate content of the latter is much higher than that of the former it is evident that the amount of the dissolved carbon dioxide in the blood or body fluids of marine forms must be considerably greater than that occurring in sea water. The tension of carbon dioxide in the blood of marine forms must also be proportionately higher than that in sea water." This argument is sound provided one assumes that sea water is in equilibrium with atmospheric air in respect to free carbon dioxide. It may be misleading, however, because of the use of the word "con- siderably." Let us make a specific comparison of skate's blood and sea water: BHCOj. pCOj. vol. per cent mm. Ilg Sea water 5 0.2 Arterial blood of the skate 8 1.4 3 In advance of its publication, we have had the privilege of reading the paper by Root (Biol. Bull., 61: p. 427), on the respiratory function of the blood of marine fishes. His single observation on arterial blood of Prionotus carolinus shows sub- M.mtially the same pressure gradient of carbon dioxide fiom blood in the gills to water as we have found in the skate. PROPERTIES OF SKATE'S BLOOD Assuming the same pK', the blood of the skate will be more acid by 0.64 pH units. It is true that the ratio of carbon dioxide pressures is 7 to 1 but pressure gradient depends not on the ratio but on the difference in pressure. This difference, 1.2 mm., is small, — of the same order as in man. SUMMA R Y In the acid range, carbon dioxide pressure has almost no effect on affinity of skate's whole blood for oxygen. In the physiological range the effect is appreciable but still only one-half as great as in man. No difference was discerned between the carbon dioxide dissociation curves of oxygenated and of reduced blood. This was partly due to the facts that the hemoglobin concentration is one-fourth as great as in man and that the carbonic acid-combining capacity (when pCO2 = 40 mm.) is less than in man. The effect of temperature on the oxygen dissociation curves is identical with that found by Brown and Hill (1923) for human blood but somewhat different from that found by Redfield and Florkin (1931) for Urechis blood. The buffer value of plasma proteins is about twice as great, per unit weight, as that of human plasma proteins. Since the concen- tration of protein in skate's plasma is one-third to one-half as great as in human plasma, it follows that the buffer value of plasma of the two species is about the same. Buffer value of whole blood is nearly equal to that of human blood of the same oxygen-combining capacity. Transfer of gases between the blood and the external medium takes place under conditions which are quite different from those in the lungs of man. Nevertheless arterial blood is about equally saturated with oxygen in the two species. The absolute values for carbon dioxide pressure in man and the skate are very different because the blood of the skate is exposed to a virtual vacuum in respect to carbon dioxide. The pressure head of carbon dioxide from blood to the external medium, however, is of the same order of magnitude, about 1 mm. in each species. The supposition that there is a steep pressure gradient in respect to carbon dioxide in such a marine species as the skate is incorrect. BIBLIOGRAPHY BOCK, A. B., AND H. FIELD, JR., 1924. Jour. Biol. Chem., 62: 269. BROWN, W. E. L., AND A. V. HILL, 1923. Proc. Roy. Soc. B, 94: 297. COLLIP, J. B., 1920. Jour. Biol. Chem., 44: 329. DILL, D. B., A. V. BOCK, J. S. LAWRENCE, J. H. TALBOTT AND L. J. HENDERSON, 1929. Jour. Biol. Chem., 81: 551. DILL, D. B., AND H. T. EDWARDS, 1931. Jour. Biol. Chem., 90: 515. 36 DILL, EDWARDS AND FLORKIX HENDERSON, L. L, A. \". BOCK, D. B. DILL AND H. T. EDWARDS, 1930. Jour. Biol. Chem., 87: 181. HENDERSON, L. J., A. \'. BOCK, D. B. DILL, L. M. HURXTHAL AND C. VAN CAULAERT, 1('.!7. ' Jour. Biol. Ghent., 75: 305. KOKTHO, S, 1()27. Science Reports, Tohoku Imper. Univ., 2: 325. REDFIELD, A. C., AND M. FLORKIN, 1931. Biol. Bull., 61: 185. SMITH, H. \V., 1929. Jour. Biol. Chem.. 81: 107. VAN SLYKK, D. D., AND G. E. CULLEX, 1917. Jour. Biol. Ghent., 30: 289. THE NATURE OF THE REDUCING SUBSTANCES IN THE BLOOD SERUM OF LIMULUS POLYPHEMUS AND IN THE SERUM, CEREBROSPINAL FLUID AND AQUEOUS HUMOR OF CERTAIN ELASMOBRANCHS FRANK FREMONT-SMITH AND MARY ELIZABETH DAILEY (From the Department of Neuro pathology, Harvard Medical School, Boston, Mass., arid the Marine Biological Laboratory, Woods Hole, Mass.) From the data in the literature it seems to be generally agreed that a variable part of the material determined as blood sugar is not glucose. Different investigators have attempted to determine glucose alone by using different reagents that react only with glucose, or by removing interfering substances (West, Scharles and Peterson, 1929) (Somogyi and Kramer, 1928). Folin and Svedberg (1926), because of discrepancies in the residual reducing substances after fermentation as determined by the Folin and Wu (1920) and Folin (1926) methods, say that the fermentable sugar is not all glucose. Both Benedict (1928) and Somogyi and Kramer (1928) conclude that there is no measurable amount of fermentable sugar other than glucose present in human blood. We are presenting in this paper evidence to show that the reducing substances in the blood of Limulus polyphemus and in the blood, cerebrospinal fluid and aqueous humor of certain elasmobranchs are fermentable by yeast, and not appreciably affected by hydrolysis. We have determined the total reducing substances, reported as mg. glucose per 100 cc., by the method of Folin and Wu (1920) in the serum of Limulus polyphemus in twelve instances; in the serum of elasmobranchs in seven instances, in their cerebrospinal fluid in five, and aqueous humor in two instances. Yeast fermentation was done by the method described by Benedict (1928) in all these cases. The yeast blank which was subtracted from the residual after fermentation ranged from 4 to 6 mg. per 100 cc. Hydrolysis was done according to the method of Folin and Berglund (1922) in the twelve specimens of Limulus polyphemus. We have already reported data on the relative composition of sea water and the blood of Limulus polyphemus (Dailey, Fremont-Smith and Carroll, 1931). The data reported here were obtained from the same samples of blood. The blood was taken by inserting a needle 37 FRANK FREMONT-SMITH AND M. E. DAILEY into the body cavity at the attachment of the caudal spine. In the elasmobranchs we endeavored to obtain samples allowing as little asphyxia as possible. The aqueous humor was obtained by puncture of the anterior chamber of the eye. Cerebrospinal fluid was obtained by puncture through the anterior fontanelle while the gills were still immersed in running sea water.1 To collect blood the fish was then removed from water, the abdominal wall rapidly slit, the viscera pushed aside, and blood aspirated from the inferior vena cava. In animal 7, Table II, blood was obtained also from the abdominal aorta. The blood was collected under oil, centrifuged, and a "protein-free filtrate" made at once. TABLE I Reducing Substances in the Serum of Linnilus Polyphemus Animal Date Total Reducing Substances nig. glucose per 100 cc. Residual after Fermentation mi;, glucose per 100 cc. After Hydrolysis mg. glucose per WO cc. 1 8/12/29 10 0 10 2 8/12/29 11 3 13 3 8/13/29 15 2 15 4 8/13/29 11 3 15 5 8/13/29 12 1 11 6 8/13/29 14 3 18 7 8/26/29 14 0 14 8 8/26/29 21 1 21 9 8/2d "' 16 0 21 10 8/29/29 13 2 13 11 8/29/29 22 0 22 12 8/29/29 15 1 19 The data on Linnilus polyphemus are presented in Table I ; those on the elasmobranchs in Table II. In the sera of Linnilus polyphemus the residual reducing substances after fermentation varied from 0 to 3 mg. per 100 cc. and averaged 1.5. In the fluids from elasmobranchs there was in one instance a residual of 5 mg. per 100 cc., but the average Mas only 2 mg. per 100 cubic centimeters. The greatest change in reducing substances of the blood of Linnilus polyphemus with hydrolysis \\as 5 mg. glucose per 100 cc., while in eight of the t \\el\e cases there was a change of 2 mg. per 100 cc. or less. These changes are too slight to be of significance. Considering the limits of error of the methods used it may be said that the total reducing substances in serum, cerebrospinal fluid or aqueous humor examined ' linn i ' uiic question as to whether this fluid, obtained from the pcrimcningeal ^p"' is niilv analogous to the mammalian cerebrospinal fluid. (Smith, II. \V., L929. Jour. /•>'/«/. Chem., 81: 407.) REDUCING SUBSTANCES IN BLOOD SERUM OF LIMULUS 39 s 2 w s w ^ I >^<» a 8 • fc 8 6.0 s • u I AQUEOUS HUMOR Residual After Fer- mentation mg. glucose per 100 cc. O "-< 3 o H Reducing Substances mg. glucose per 100 cc. o lO CEREBROSPINAL FLUID Residual After Fer- mentation mg. glucose per 100 cc. *— 1 o « * o "rt O Reducing Substances mg. glucose per 100 cc. o 1 — 1 CM CM 3 q 5 SERUM Residual After Fer- mentation mg. glucose per 100 cc. CM t-l « - CM CM rt O cS "rt O H Reducing Substances mg. glucose per 100 cc. 0 «-• ON ""> T— 1 1— 1 ~ I - o i - l - ^ fo" *I 00 to f^> -t-< o cd V a Q 0s ON CM CM IO f— OO OO ON ^ i x" I - r- 1 . — i X ON CM oT CN c- X ON -i c- ON CM ON CM Animal 8. Carcharinus obscurus Carcharinus obscurus tn CO 10 Mustelis canis. . . Mustelis canis. . . Raia ocellata .... Carcharias littoral Carcharias littoral Carcharias littoral ••-H CM 1*3 -t- ir, NO «^ 40 FRANK FREMONT-SMITH AND M. K. DAILEY are fermentable by yeast and not changed by mild hydrochloric acid hydrolysis. Tin- striking difference in the level of total reducing substances in the blood of Limulus polyphemus and of the elasmobranchs studied is interest ing in relation to the relative activity of the two groups. This relationship was discussed by Gray and Hall (1930). From their investigation we would expect to find low blood sugars in the sluggish and inactive Limulus and higher sugar values in the more active elasmobranchs. Our findings are consistent with theirs, as seen in Tables I and II. It is well-known that asphyxiation produces an increase in the amount of sugar in the blood of fishes (Denis, 1922; Menten, 1927; and Scott, 1921), which may in our cases have accentu- ated the higher blood sugar level of the elasmobranchs. The fact that a fermentable reducing substance, not affected by hydrol\>is, and therefore probably glucose, is consistently found in the blood of Limulus polyphemus and these ekismobranchs is interesting because there is little information as to the concentration or nature of the reducing substances in the blood of lower animals, and because of the ancient lineage of both species. \Ye found the level of glucose in the sera of Linutliis polyphemus to be relatively constant, varying only from 10 to 22 mg. per 100 cc. in the twelve animals studied. There are some data in the literature on the blood sugar of elasmo- branchs. Denis (1922) found the sugar of Mnstelis canis to vary from 80 to 181 mg. per 100 cc. ; Gray and Hall (1930) found 65 and 87 mg. per 100 cc. in two dogfish. These figures are within the limits of our findings, but we do not know the relationship of the blood sugar level, which varied from 36 to 190 mg. per 100 cc., to the degree of asphyxiation in the animals studied. In Table II are reported five instances of parallel determinations of reducing substances in serum and cerebrospinal fluid. In four of these cases total and fermentable reducing substances were distinctly higher in venous serum than in cerebrospinal fluid. In case 7 the venous blood \\a^ lower, but the arterial higher. These studies give 11^ no indication as to the relationship between reducing substances in the serum and cerebrospinal fluid at equilibrium as asphyxial elevations of the blood sugar had undoubtedly occurred after the cerebrospinal fluid was removed. CONCLUSION The reducing substance found in the blood serum of Limulus Polyphemus, and in the blood serum, cerebrospinal fluid and aqueous humor of elasmobranchs is fermentable by yeast and not appreciably REDUCING SUBSTANCES IN BLOOD SERUM OF LIMULUS 41 affected by hydrochloric acid hydrolysis. It is, therefore, probably glucose. BIBLIOGRAPHY WEST, E. S., F. H. SCIIARLES, AND V. L. PETERSON, 1929. Jour. Biol. Chem., 82: 137. SOMOGYI, M., AND H. V. KRAMER, 1928. Jour. Biol. Chem., 80: 733. FOLIN, O., AND A. SVEDBERG, 1926. Jour. Biol. Chem., 70: 405. FOLIN, O., AND H. Wu. 1920. Jour. Biol. Chem., 41: 367. FOLIN, O., 1926. Jour. Biol. Chem., 67: 357. BENEDICT, S. R., 1928. Jour. Biol. Chem., 76: 457. FOLIN, O., AND H. BERGLUND, 1922. Jour. Biol. Chem., 51: 213. DAILEY, M. E., F. FREMONT-SMITH, AND M. P. CARROLL, 1931. Jour. Biol. Chem., 93: 17. GRAY, I. E., AND F. G. HALL, 1930. Biol. Bull., 58: 217. DENIS, W., 1922. Jour. Biol. Chem., 54: 693. MENTEN, M. L., 1927. Jour. Biol. Chem., 72: 249. SCOTT, E. L., 1921. Am. Jour. Physiol, 55: 349. STUDIES ON CF.I.I. METABOLISM I. THE OXYGEN ( MX-IMI-TION OF NEREIS EGGS BEFORE AND AFTER FERTILIZATION E. S. GUZMAN" BAUUOX (From the Marine Biological Lahoralory, \\'o»ds //<>/<•, Massachusetts and the Lasker Foundation for Medical Research and the /><;/•/ wtvj/ of Medicine, The I'lihrrsity of Chicago') The sudden and enormous increase in the oxygen consumption of Arbacia eggs soon after fertilization, which was observed by Warburg (1908) and Shearer (1922a) and confirmed by measurements of the heat production (Shearer, 19226, Rogers and Cole, 1925), had led to the assumption that the fertilization phenomena require a great expenditure of energy. Shearer correlates this increase to cortical changes in the egg and to the formation of the fertilization membrane. Loeb attributed the increase to the fact that the unfertilized eggs of the sea urchin are usually in the resting stage. To prove the validity of his hypothesis, he measured in collaboration with \Yasteneys (1912) the oxygen consumption of fertilized and unfertilized starfish eggs, which as a rule are immature when taken out of the ovary, but as soon as they are placed in sea water may become mature. No noti< v- able increase in the rate of oxidation of fertilized starfish eggs was observed by these investigators, since they postulated "those oxida- tions which lead to nuclear division were already going on in the e. at the time the spermatozoa entered." As I.illie and Just (1924) point out, with the formation of the hyaline plasma layer, the egg of Arlxn in may be said to be well started on its first cleavage cycle. This would mean th.it "the measured oxygen consumption belongs rather to the physiology of cell division than to the more specific events of fertilization." Lillie .uid Just, 1924.) It w.is therefore of particular interest to the physiology of fertili- /.ition, to measure the respiratory changes of eggs in which this process can be observed from its initial stages. The eggs of the annelid Nereis are particularly suited to this purpose. As the excellent studies of Wilson (1892) and K. R. l.illic (1912) have established, the ovum of Nereis eggs is inhibited at the end of its period of growth. The germinal vesicle undergoes none of the preparatory stages of maturation unless the egg be fertilized. The spermatozoon remains more or less quiescent within the egg during the completion of the 42 OXYGEN CONSUMPTION NEREIS EGGS 43 maturation division and the internal events of fertilization are resumed after the formation of the second polar body. EXPERIMENTAL The experiments were performed in Woods Hole during the sum- mers of 1929 and 1930. Those of 1929 were done with the collabora- tion of Dr. A. Tyler. All the experiments were performed with freshly-caught Nereis. Large flat-bottomed Warburg vessels with Barcroft micro-manometers were used for the measurement of the oxygen consumption. The temperature of the water bath was kept at 25° C. The eggs were placed in the main side of the vessels. In some cases the sperm were placed in the side arm and fertilization was performed by pouring the sperm into the main side of the vessel, after the oxygen consumption of the unfertilized eggs had been TABLE I Oxygen consumption of Nereis eggs, before and after fertilization. The oxygen consumption after fertilization was followed until the two-cell stage had been reached. Oxygen consumption in cu. mm. per hour Per cent division Before fertilization After fertilization 44.5 45.2 70 37.0 35.1 65 51.3 48.2 69 53.1 52.4 60 48.1 49.2 63 10.8 9.1 68 6.8 6.5 70 14.6 15.0 70 13.5 13.9 69 measured. In other experiments, the sperm were added directly by opening the vessel. The results obtained by both methods were identical. A total of forty-two experiments were performed. The oxygen consumption after fertilization was measured until the eggs had reached the two-cell stage, i.e., for about 80 minutes. The maximum percentage of cell division observed was 70 per cent and the minimum 45 per cent. In Table I are given figures of some of these experiments, at hourly periods. Figure 1 shows the oxygen consump- tion at different intervals after fertilization. In most cases (about 80 per cent) there was an average increase of 25 per cent in the oxygen consumption during the first eighteen minutes after fertilization. According to Lillie's observations on the fertilization events of Nereis, 44 E. S. GUZMAN BARRON this corresponds to the time when the fertilization cone has flattened out, but the spermatazoon is still external to the membrane. But if one compares the total oxygen consumption of fertilized eggs (from the moment fertilization takes place, until the two-cell stage has been reached) with the oxygen consumed by the same eggs when unfertilized in equal length of time, it is observed that there is no difference. The subsequent decrease in the oxygen consumption is possibly due to a diminution in the rate of diffusion of the COo produced by the cells on account of the jell which surrounds them. The amount of jell secreted by the ovum increases from the time of fertilization. 20 40 60 80 100 120 Time in Minutes FIG. 1. The oxygen consumption of Nerd* e^s before and after fertili/.ition. as compared to the oxygen consumption of sea urchin eggs before and after fert ili/.i- tion. Arbacia, the lower line. Xcreis, the upper line. Arrows mark the time of fertilization. I )ISCUSSION The striking increase in the oxygen consumption observed soon after Arbacia eggs have been fertilized has exercised such influence in the minds of most biologists, that one always finds it stated that the fertilization phenomenon is in gencnil accompanied by an enormous increase in the oxidatixr processes of the fertilized cell. The case of fertilized starfish eggs, \vhere such an increase is not observed, has not received corresponding attention. If I.oeb's explanation for this lack of increased oxidation in fertilized starfish eggs is correct, one would expect in fertilized Nereis eggs about the same increase in the oxygen consumption as found in fertilized sea urchin eggs, since the initial process of fertilization starts here only when the spermatozoon enters the ovum. As has been shown in the experimental part, such is not OXYGEN CONSUMPTION NEREIS EGGS 45 the case. All one can say is that during the first eighteen minutes (which corresponds to the formation of the fertilization cone in the egg protoplasm while the spermatozoon is still found external to the substance of the entrance cone) there is a small increase of about 25 per cent, which is negligible compared to that found in fertilized sea urchin eggs where the oxygen consumption goes up to eight times that before fertilization. Fertilization in Nereis eggs proceeds without a decided increase in the cellular oxidative processes just as in starfish eggs. If expenditure of extra energy is needed for the performance of this activity, it is possibly supplied through a hydrolytic process, probably glycolysis. In the breakdown of glycogen to lactic acid an amount of energy is liberated, which according to Burk's calculations (1929) is from one and a half to two times greater than the heat of reaction. The great increase in the oxygen consumption of Nereis eggs after the addition of a reversible dye seems to support the view that these eggs possess a carbohydrate fermentation process, since the catalytic effect of dyes is mainly due to oxidation of carbohydrates which have already been rendered easily oxidizable by the action of enzymes (Barron, 1929). l CONCLUSION The oxygen consumption of Nereis eggs before and after fertilization to the two-cell stage has been measured. 1. The total oxygen consumption, from the initiation of fertilization to the two-cell stage, is no greater than the oxygen consumption for a similar period of time before fertilization. 2. When the oxygen consumption is measured at shorter intervals there is a small increase of about 25 per cent during the first eighteen minutes. 3. The present observations, together with those on the starfish eggs reported by Loeb, suggest that the enormous increase in oxygen consumption of fertilized sea urchin eggs may be considered an exception. BIBLIOGRAPHY WARBURG, O., 1908. Zeilschr. f. Physiol. Chem., 57: 1. SHEARER, C., 1922a. Proc. Roy. Soc. London, B, 93: 213. SHEARER, C, 19226. Ibid., 93: 410. ROGERS, C. Y., AND K. S. COLE, 1925. Biol. Bull., 49: 338. LOEB, J., AND H. WASTENEYS, 1912. Arch.f. Entw. Organism., 35: 555. LILLIE, F. R., AND E. E. JUST, 1924. Cowdry's General Cytology, University of Chicago Press. WILSON, E. B., 1892. Jour. Morph., 6: 361. LILLIE, F. R., 1912. Jour. Exper. Zoo/., 12: 413. BURK, D., 1929. Proc. Roy. Soc. London, B, 104: 153. BARRON, E. S. GUZMAN, 1929. Jour. Biol. Chem., 81: 445. 1 These experiments were performed with cresyl blue as catalyst. The increase in the oxygen consumption after dye addition is about 300 per cent. T1IK EFFECT OF ANAEROBIOSIS ON THE EGGS AND SPERM OF SEA URCHIN, STARFISH AND NKRKIS AND FERTILIZATION UNDJ-.k .\\.\KKOBIC CONDITIONS E. S. GUZMAN BARROX (From the Marine Biological Laboratory, Woods Hole, and the Lasker Foundation for Medical Research and the Department of Medicine, The I 'nii-ersity of Chicago) The effect of anai'-robiosis upon fertilization has been studied by various authors using the sea urchin as material for experimentation. All these studied have been confined to the effect of lack of oxygen after the sperm and the egg had been in contact, missing therefore the initial process of fertilization. Loeb and Lewis (1902) exposed the eggs of Arbacia soon after fertilization to lack of oxygen and stated that the process was inhibited. Mrs. Harvey (1927) subjected the eggs of the same species to a stream of hydrogen in a "modified Engleman's chamber" at different times after fertilization. In these experiments anaerobiosis was not reached according to the author until twenty minutes later. It is also possible that the sea water became hypertonic because of the small amount of fluid used in the chamber. More recently, Mrs. Harvey studied the effect of lack of oxygen on the eggs and sperm of sea urchin and on fertilization (Harvey, 1930). In order to study the effect of anai'-robiosis upon the fertilization phenomena, it is essential in the first place to select eggs and sperm "I species which can resist the lack of oxygen without injury, at least during the time before they are brought in contact; in the second place, to use a technique where the absence of oxygen is controlled and where there is no danger of the sea water becoming hypertonic. It is the purpose of this paper to describe such a technique and to relaic the effects of anai'-robiosis upon the initial phase of fertilization of the following species: sea urchins, starfish and Nereis. These experiment were performed in Woods Hole during the summer of 1929. METHOD FOK SM DYING -NIK EFFECT OF ANAEROBIOSIS ON FERTILIZATION Three large specimen bottles (150 cc. capacity), tightly stoppered with rubber stoppers and connected to each other through glass tubing, 46 EFFECT OF ANAEROBIOSIS ON EGGS AND SPERM 47 were used for every experiment. Bottle A contained water in order to saturate the gas. The gas inlet was on bottle A. Bottle B con- tained the eggs suspended in 50 cc. of sea water. These eggs were previously received in large crystallizing dishes, washed twice with sea water and passed through cheesecloth. Bottle B contained a spoon where the sperm was kept when the fertilization was performed with sperm in anaerobiosis or a burette containing boiled sea water, when the fertilization was performed with fresh sperm. In the latter case there also was a platinum electrode and a saturated. KC1 bridge connected to a saturated calomel electrode. Bottle C contained a solution of safranin (approximate E'o - 0.290 v.) in Sorensen's phosphate buffer pH 8..0 and some platinum asbestos. Test jor A bsence of Oxygen It is essential in an experiment where the effect of anaerobiosis is followed, to be reasonably sure that no detectable amounts of oxygen are present in the vessels where the experiments are performed. In the early days reduced methylene blue had been used for such a purpose, but this is an unreliable test. The absence of luminescence in Cypridina luciferin-lnceferase or in luminous bacteria has been advocated by Harvey (1926). When reversible systems showing a change of colour or light from the oxidized to the reduced state are used as tests for the absence of oxygen, it is necessary to have in mind that the rate of oxidation of the system is about proportional to the E'o of the system. Reduced indophenol will remain reduced for some minutes when the vessel containing the leuco dye is exposed to the air. Reduced methylene blue will oxidize more quickly but will remain reduced for a time in the presence of small oxygen concentra- tions. The more negative the E'o of the dye is, the more easily it will be oxidized by oxygen when in the reduced state (Barren, 1931). On these grounds the writer has used and recommends reduced safranin T as an excellent test for detecting the presence of minute traces of oxygen. When fertilization was performed with fresh sperm the E. M. F. of the sea water obtained through a bright platinum electrode against a saturated calomel electrode was also used to detect the presence of oxygen. It is known that this electrode is extremely sensitive to oxygen. Safranin was reduced with hydrogen and platinized asbestos. After complete reduction had been obtained, nitrogen (purified according to Michaelis and Flexner, 1928) was bubbled throughout the experiment. Experiments in which there was the slightest coloration of safranin were discarded. 48 E. S. GUZMAN BARRON The Effect of Anaerobiosis on the Eggs and Sperm of Sea Urchin and Fertilization under Anaerobic Conditions I.oeb (1915) stated that sea urchin eggs could be better preserved in absence of oxygen, and related experiments where eggs kept for twenty-four hours under hydrogen could be fertilized as well as fresh eggs. It is possible that in these experiments, the eggs were not exactly under anaerobic conditions but under low oxygen pressures. Loeb's experiment has been repeated, where purified nitrogen was passed through the vessel for a period of twenty-four hours. The eggs were then transferred to a flat-bottomed dish and fertilized with fresh sperm. There were many cytolyzed cells, and on fertilization the eggs gave only 15 to 20 per cent of cell division. Similar experi- ments performed with fertilized eggs soon after fertilization (the vessels were kept on ice while deaeration was completed) gave identical results, i.e., although there was 100 per cent membrane formation, only 20 per cent of the eggs reached the two-cell stage. Many of these cells gave abnormal types of division. When the eggs are kept in anaerobic conditions for periods of one hour to six hours they can be normally fertilized with fresh sperm (Table I). From experiments on the effect of KCN on the life of sea urchin sperm, Cohn (1918), confirming similar observations of Drzewina and Bohn (1912), concludes that lack of oxygen preserves the life of these celjs. \Yhen the experiments are performed in true anaerobiosis such is not the case. Sea urchin sperm are very sensitive to lack of oxygen. Concentrated sperm suspensions were kept in anaerobiosis for periods of one to five hours. It was observed that they could live, remain motile and fertilize fresh eggs only after one to two hours of anaero- biosis. Three hours later the sperm had lost their vitality and ferti- lized fresh eggs no more (Table I). I rom what has been said about the extreme sensibility of sperm to lack of oxygen it is obvious that fertilization could not be possible when eggs and sperm were kept in anaerobic conditions. Such is the case. Fertilization was performed by introducing the spoon which contained the sperm suspension into the fluid containing the eggs while the vessel was shaken. The nitrogen flow was continued for one hour more. The vessel was opened, and immediately formol was ad(lct size and quality of eggshell l\i>maiioft', 1929). The incubation was carried on in a special electric laboratory incubator i K< nn.uiotT, 1929,/) under the conditions already described koinanott, 19SIM with the temperature 38.0 ± 0.2° C., and the relative Inimidiiy C>0.0 ± 1.0 per cent. 54 FAT METABOLISM OF CHICK EMBRYO 55 TABLE I Iodine Value, Sapouificaliaii Xitinhcr and Refractive Index of Fat (ether extract} of Yolk of Fowl's Egg Author Percentage of Fat in \\Vt Weight Iodine Value Saponi- ficHtion Number Re- fractive Index Parke (1866) per cent 22.82 (n)D Ainthor and Zink (1897) 75.4 189.4 1.4670 Kitt (1897) 1900 71.1 190.2 Peninngton (1909) 32.39 63.6 182.2 1.4611 Serono and Pallazi (1911) ?0 17 82.31 198.85 Thomson and Sorlcv (1924) 31.10 74.73 183.79 , Romanoff (present rpt.) 30.71 73.89 190.69 1.4690 At intervals of 24 hours all eggs were candled for dead embryos, and at least four eggs with normally developed embryos were removed for analysis. Previous experience in our laboratory showed that at certain stages of embryonic development it is almost impossible to separate the yolk from albumen or allantoic fluid, and to determine it quanti- tatively with appreciable accuracy. Therefore the boiling of eggs,— to complete coagulation of yolk and albumen, — had been employed. This method proved to be very quick and satisfactory, not only in determining the total values of yolk and albumen, but also in some chemical analysis, such as fat (ether extract) determination. TABLE II Distribution of Fat (ether extract) in a Fresh and Hatched Egg .rarts 01 tne c,gg Amount Percentage Fresh egg: Yolk grams 5.8275±1.0971 per i fit! 99.50 Albumen 0.0046±0.0013 0.08 Shell 0.0033 ±0.0025 0.06 Shell membranes 0.0213±0.0167 0.36 Total . 5.8567 100.00 Hatched egg: Chick 1.8887 ±0.1224 48.57 Yolk sac 1.931S±0.0762 49.67 Shell with waste matter 0.0682±0.0126 1.76 Total 3.8884 100.00 Difference (combusted fat) ... 1.9683 33.61 Fat (ether extract) 56 ALEXIS L. ROMANOFF TABLE III Changes in the Fat (ether extract) Content of the Chick Embryo Incubation Wright of Kmbryo Fat (ether extract) Content Wet Weight Dry \Vi-ight Percentage (if Dry Matter Comi> Samples Average Fat Content Percentage of Wet Weight Percentage of Dry Weight days grams grams per cent embryos grams per cent per cent 8 1.350 0.099 7.33 18 0.0133 0.99 13.13 9 2.083 0.147 7.06 12 0.0233 1.12 15.85 10 2.700 0.238 8.81 6 0.0310 1.15 13.03 11 3.850 0.305 7.92 6 0.0539 1.40 17.67 12 5.499 0.513 9.33 5 0.1015 1.85 1'>.79 13 7.213 0.703 9.75 ' 4 0.1610 2.22 22.90 14 9.670 1.159 11.99 2 0.2575 2.66 22.22 15 11.327 1.544 13.63 2 0.3475 3.07 22.51 16 15.360 2.582 16.81 2 0.5012 3.26 19.41 17 18.095 3.396 18.77 2 0.7997 4.42 23.55 18 22.310 4.220 19.92 2 0.9823 4.40 23.28 19 29.891 5.812 19.44 1 1.3024 4.36 22.41 20 31.298 5.999 19.17 1 1.4053 4.49 23.43 TABLE IV /;/ the Fat (ether extract) Content of the Yolk Sac ge of Incubation As i-rage ligg Weight Weight of Yolk Sac Fat (ether extract) Content Wet Weight Dry Weight Percentage of Dry Matter Average Content Percentage of Wet Weight Percn ..I 1 ' Weight days grams grams grams per i fill grams per cent per i en! 0 61.6 18.44 9.654 52.35 5.0633 30.71 58.66 1 (,2.0 19.19 10.110 52.68 6.4837 33.79 64.17 2 62.S 19.79 10.113 51.10 6.0309 30.47 5'). 54 3 02.5 20.10 9.540 17.49 6.0338 30.02 63.21 4 62.5 21.7" 10.390 47.68 (..5621 30.12 63.K. 5 61.3 20.22 8.7S1 13.43 5. 35 I1' 26.48 60.98 6 61.8 21.51 8.872 41.24 5.1276 25.23 61.18 7 62.3 21.81 9.354 I !.88 5.8198 ^6.68 02.22 8 02.5 20.58 8.011 11.S4 5.4874 26.66 63.73 9 60.3 19.84 8.726 13.98 5.6041 28.25 64.22 10 62.0 19.08 8.767 45.95 5.2271 27.39 5'). 62 1 1 61.3 18.80 8.3S2 11.59 5.3328 28.37 63.62 12 61.5 1X.27 s.017 47.17 5.4158 29.64 62.85 13 62.0 18.31 9.097 49.68 5.4335 29. os 59.72 11 59.3 16.64 9.539 57.35 5.5897 33.5') 58.5') 15 10.34 9.676 59.25 5.4803 33.55 56.64 16 60.0 1 1.54 9.006 61.94 4.5241 31.11 50.23 17 61.0 14.li ' 9.132 65.14 4.4922 32.01 49.19 18 13.87 8.255 59.52 3.0893 22.27 37.42 19 59.5 12.03 6.986 58.57 2.4567 20.42 35.17 20 00.5 10.53 6.015 57.12 2.1434 20.35 35.64 FAT MKTAMOUSM OK CHICK KMBRYO 57 In these experiments two eggs were boiled for 20 minutes, the coagulated yolk was weighed, then dried to constant weight in a Freas electric vacuum oven at 55° C. and vacuum at about 63.5 cm., ground and subjected for 48 hours to extraction of fat with anhydrous ether by a Soxhlet extraction apparatus. From two or more unboiled eggs the embryos were separated for determination of dry matter and fat content by methods similar to those described above. The extracted fat from both the embryo and the yolk sac was tested for saponification number (Koettstorfer, 1879), iodine value (\Vijs, 1898), L and refractive index, by a Zeiss butter refractometer. 1.5 h 1.0 X cu 0.5 0.0 o o Lmbiyo Egg YoU J 6.5 6.0 5.5 5.0 X r+- 3.0 a 2.5 -I 2.0 Days 0 2 H 6 8 10 12 \H 16 18 20 Incubcition Age FIG. 1. Changes in the fat (ether extract) content of the embryo and egg yolk during incubation. RESULTS AND DISCUSSIONS It has been found that the fat content of a fresh fowl's egg is directly proportional to its size. For that reason the amounts of fat obtained from an egg by previous workers can hardly be compared, 1 Both saponification number and iodine value were determined after the methods given in Official and Tentative Methods of Analysis of the Association of Official Agri- cultural Chemists, third edition, 1930, on pp. 321-322. 6 58 ALEXIS L. ROMANOFF unless the data are expressed in percentages of dry or wet weight of the substance. Similar inconsistency is observed in the data on iodine value, saponification number and refractive index of the fat from an egg yolk (Table I). These dissimilarities might be accounted for by the lack of uniformity in material and methods, such as: the method of extraction, care of the extract, age of eggs and possibly seasonal quality of eggs. It is evident from the data of Table II that the egg yolk is the main source of fat to the developing embryo. In a fresh egg there is Pel cent LJ 24 22 20 18 16 12 01 Q) in. thy Emlnyo />' / Peiceniaqe oj NUtlet ii\ & 9 10 It 12 13, tf \5 IG 17 18 19 20 Incubcdicm Age FIG. 2. I Vrri-Mt.iKes of fat (ctlu-r extract) in the- wet and dry weights of the embryo at various stages of incubation. (The curves of thft above values are not identical due to changes in the dry mat NT content of the embryo, which is shown by dotted line.) about W.5 per cent of yolk fat and only 0.5 per cent of fat in the rc-niaindcr of the egg, including albumen and shell with shell mem- branes. At the time of hatching the yolk sac still has the largest portion of reserve fat for the nutrition of the chick after hatching. Next in amount is the fat absorbed by the embryo. Then a small amount of fat is retained in the waste matter and in the shell with FAT METABOLISM OF CHICK EMBRYO 59 shell membranes. The combusted fat represents 33.61 per cent, or perhaps a larger amount if synthesis of fat had taken place during the incubation. The daily fat content of the developing embryo and egg yolk (Tables III and IV) gives a general idea of the changes which go on within an egg during incubation (Fig. 1). The fat of the embryo increases noticeably only after about two weeks. The fat of the yolk at first decreases slowly; then it rapidly falls from about the sixteenth day up to hatching time. The curves on the percentages of fat in the wet and dry embryo (Fig. 2) are not identical. Change in the moisture content of the embryo throughout the incubation period is the main influencing factor. Therefore a true percentage value should be taken from dry rather than wet weight of the embryo. Pel cent 65 60 55 x. o 50 o QJ 25 20 Pe7cen.ta.ge of Matter in. YoU Fat in. Wet Yolx Pays 024 ] Embryo Fat Yolk Fat days 0 — 1.4685 9 1.4879 — 10 [1.4903] — 11 1.4877 — 12 1.4841 — 13 1.4844 1.4687 11 1.4816 1.4695 15 1.4758 1.4693 16 1.4763 1.4695 17 1.4741 1.4675 18 1.4735 1.4691 19 1.4727 1.4696 20 1.4719 1 .4690 growth of the embryo; and the drop due to the rapid and sole utili/ation of fat for growth and for energy supply to the embryo. The combusted fat accompanying embryonic development, is the expression of the energy expended. It has been observed that the greatest part of the egg fat is used toward the end of hatching time. The curve of the combusted fat uoiild be very similar to the rurye of the Drouth of the embryo ( Romanoff, 1930). The e\tra< ie<| fat from egg yolk throughout the incubation period -hows hardly any change in the saponification and iodine numbers .ind refractive index, giving on an average 73.89, 190.69, and 1.4690 ii --pet-lively. On the other hand, there was found a pronounced change in the iodine number and refractive index of the fat from the embryo (Table Y, I;ig. 4). The iodine number of the fat from the embryo at 16 days of incubation was 78.84; on the following days it FAT MKTABOLISM OF CHICK EMBRYO 61 14900 860 3 82° i 14800 »— « £ 780 _> "5 760 ^T 7fO 720 14700 G80 660 o o Embryo D^s 10 12 1* 16 18 20 Incubation. Age FIG. 4. Refractive index of the fat from the embryo and egg yolk during incubation. was 82.03, 83.69, 84.03, 88.20, reaching 90.34 at the time of hatching. The refractive index was found to be high at the beginning of obser- vation, nine days of embryonic age, and was rapidly decreasing towards hatching. The author wishes to express his thanks to H. A. Faber for as- sistance in routine analytical work covering this paper. SUMMARY 1. Fat content of fresh eggs is a subject of great variation, due primarily to disparity in size of individual eggs. 2. Iodine value, saponification number and refractive index of the fat from fresh eggs are rather constant, but only under a given experi- mental condition. 3. The main source of fat in an egg is the egg yolk, which furnishes the fat for nutrition and energy-supply of the developing embryo. 62 ALEXIS L. ROMANOFF 4. The amounts of fat in the growing embryo and egg yolk give a comprehensive idea about the changes in fat which go on within an egg during incubation. 5. The relative increase of fat in the embryo and the decrease of it in the yolk can be well demonstrated by the curves plotted from the data on the percentages of fat in dry weight. 6. The curve of combusted fat is quite similar to the curve of the growth of embryo. 7. Iodine value, saponification number and refractive index of the fat from the yolk sac of the developing egg are almost constant through the incubation period. 8. Iodine value and refractive index of the fat from the developing embryo are increasing and decreasing respectively during the latter part of incubation. REFERENCES AMTHOR, C., AND J. ZINK, 1897. Beitrag zur Chemie der Thierfette. Zeitschr. f. analyt. Chem., 36: 1. EAVES, E. C., 1910. The Transformation in the Fats in the Hen's Egg during Development. Jour. Physiol., 40: 451. IDZUMI, S., 1924. Biochemische und serologische Untcrsuchungen von bcbriiteten Hiihnereiern. Mill. med. Fak. Univ. Tokyo, 32: 197. KITT, M., 1897. Zur Kenntniss des Eieroles. Chemiker-Zeitung, 21: 303. KoETTSTOKi'KK, J., 1879. Neue Methode zur Untersuchung der Butter auf fremde Fette. Zeitschr. f. analyt. Chem., 18: 199. MURRAY, II. A., 1926. Physiological Ontogeny. A. Chicken Embryos. VII. The Concentration of the Organic Constituents and the Calorific Value as Functions of Age. Jour. Gen. Physiol., 9: 405. NEEDHAM, J., 1925. The Metabolism of the Developing Egg. Physiol. Rev., 5: 1. NEEDHAM, J., 1927. The Energy-sources in Ontogenesis. IV. The Relation bet \vrrn Absorption and Combustion of Protein ami Fat in the Developing Avian Egg. Brit. Jour. Exper. Biol., 4: 258. PARKH, J. L., 1866. Ueber die chemische Constitution des Eidotters. Hoppe- Seyler's Med. -Ghent. l'ntcr>,,h., 2: 209. PEXIXXGTON, M. E., 1909. A Chemical and Bacteriological Study of Fresh Eggs. Jour. Biol. Chem., 7: 109. ROMAXOFF, A. L., 1929. Study of the Physical Properties of the Hen's Eggshell in Relation to the Function of Shell-secretory Glands. Biol. Bui!., 56: 351. ROMAXOI i . A. I ., 1929a. Laboratory Incubator for the Biological Study of Chick Embryo. Science, 69: l'>7. ROMAXOFF, A. L., 1930. I'.ioi In mistry and Biophysics of the Developing Hen's Egg. I. Influence of Humidity. Cornell Univ. Agr. Expt. Sta. Memoir, 132: 1. SAKURAGI, 1917. Der Kohlehydratstoffwechsel in den bcbriiteten Hiihnereiern; Entstehung von Zucker aus Fett im Tierkorper. Mill. med. Gesellsch. w, 31: 1. SEROXO, C., AXD A. PALOZZI, 1911. Sui lipoidi contenuto ncl tuorlo d'uovo. Arch. farm, sperm., 11: 553. Iiiii\i-Mi\, R. T., AXU J. SORLEY, 1924. The Composition and Decomposition of Eggs. Analyst, 49: 327. \Yijs, J. J. A., 1898. Berichte d. Deutschen chem. Gesellschaft, 31: 750. IS THE ERYTHROCYTE PERMEABLE TO HYDROGEN IONS? M. H. JACOBS AND ARTHUR K. PARPART (From the Department of Physiology, University oj Pennsylvania, and the Marine Biological Laboratory, Woods Hole, Massachusetts) I The erythrocyte is generally believed to differ from most other cells in the ease with which its internal reaction is influenced by that of its surroundings. Even in the absence of so-called "penetrating" acids and bases such as COo, fatty acids, ammonia, etc., which alone are effective with ordinary cells, it is easy to bring about in it striking internal pH changes. The mere hydrogen ion concentration (strictly speaking, the hydrogen ion activity) of the surrounding medium, regardless of how it is produced, seems automatically to determine in the erythrocyte an internal hydrogen ion concentration that can be predicted by the principles of the Donnan equilibrium (Warburg, 1922; Van Slyke, Wu and McLean, 1923). This interrelation of external and internal pH is part of an important mechanism for preserving an approximately constant blood reaction under all ordinary physiological conditions. While the establishment of a Donnan equilibrium between the erythrocyte and its surroundings is almost universally considered to be due to a passage of ions across the cell membrane, there has been some doubt as to the particular ions involved in the case of pH changes. Partly, no doubt, owing to our customary methods of measuring and denning the reactions of aqueous solutions, and partly to the fact that physiologists, in general, have had to deal more frequently with the penetration of cells by acids than by alkalies, it has been customary in the past to postulate a ready permeability of the erythrocyte to hydrogen ions. But it has been pointed out by Van Slyke, Wu and McLean (1923) and others that exactly the same end results would be obtained if the permeability were to hydroxyl rather than to hydrogen ions, since, as is well known, the relation existing between these two ions in aqueous solutions is such that the activity of the one is related in a fixed manner to that of the other. In cases where our interest is primarily in equilibria, it is a matter of indifference whether the cell is permeable to the hydrogen or to the hydroxyl ion, or to both; indeed, by the study of equilibrium 63 64 M. H. JACOBS AND A. K. PARPART states alone it is impossible to reach any decision as to the mechanism by which these states have been reached. Since most of the work heretofore done on the erythrocyte has been concerned with equilibria, it is not surprising that the question of the relative penetrating powers 1 hour The probable reason for the accelerating effect of Na-SO4 on the color change will be discussed below. That for the inhibition of hemolysis is not certainly known, but at all events such experiments furnish a clear demonstration of the possibility of intracellular color changes. Though liberation of hemoglobin from the cells is not necessary in order that the reaction may occur, it must not be thought that the entrance of acid is not influenced by injury to the cells. In observing the behavior of a given suspension, it may be noted that the color change does not proceed gradually and fairly regularly, as it does in the case of hemolyzed blood at sufficiently high pH values, but that for a time the change is extremely slow or entirely invisible, and then suddenly becomes much more rapid, as if a barrier of some sort had been broken down by injury to the cell. Shortly after this point, hemolysis occurs. As contrasted with the slow rate at which the color change is produced by HC1, that in the presence of so-called "penetrating" acids, such as acetic or butyric acids at the same or even considerably less acid pH values, is strikingly rapid. In this respect the behavior of the erythrocyte is similar to that of ordinary cells containing natural or artificially introduced indicators. In the experiments here reported the erythrocytes were obtained from defibrinated ox blood. The cells were not "washed," since it \\ .is believed to be less serious to introduce into the solutions employed the slight traces of blood proteins unavoidably present in dilutions of 1 : 250 or 1 : 500 than perhaps to change the fundamental properties of the erythrocytes in the manner described by Kerr (1929) by previ- ously removing these proteins. The pH determinations were in all cases made with the quinhydrone electrode at the conclusion of each set of experiments. Though the addition of the blood somewhat reduced the acidity of the original solutions, the latter were present in such excess that the pH changes in most cases did not amount to more than a few tenths, or at the highest concentrations of acid, a few hundredths of a pi I unit. The color changes were determined in test tubes by eye, the time given being that at which a distinct color difference could be detected l>< 'tween the experimental tube and an appropriate control. Though it was found that a certain degree of refinement could be introduced I iv the iiM' of a colorimeter, the advantages of this instrument for the present purposes were not sufficient to compensate for the considerably PERMEABILITY OF THE ERYTHROCYTE 67 greater time required with it to carry out a series of experiments. The determinations of the time of hemolysis, as has already been mentioned, were made by the method of Jacobs (19306), of necessity on a separate, but as nearly as possible identical, suspension of the same blood in the same solution. The few cases in which hemolysis appeared slightly to precede the color change are probably to be accounted for by the fact that the figures were obtained not from a single experiment but from two parallel experiments. Ill The method described in the preceding section was first employed by us in an attempt to throw further light upon an earlier observation (Jacobs, 1930a) that the rate of acid hemolysis is greatly influenced by the salt concentration of the external medium. Illustrations of TABLE I Effect on time of color change and of hemolysis of adding different amounts of NaCl to 0.02 M HCl in 0.3 M saccharose. All times are in seconds. Concentration of NaCl Color change Beginning of hemolysis 75 per cent hemolysis — 116 130 153 0.0002 118 127 147 0.0004 108 128 135 0.0008 104 • — - — 0.0016 no — — 0.0031 116 — — 0.0063 111 123 137 0.0125 102 111 125 0.025 87 95 119 0.05 71 80 92 0.1 60 71 81 0.2 45 72 80 this effect will be found in Table I and in Tables II and III which, taken together, show the same thing. The more rapid rate of acid hemolysis in the presence of NaCl might conceivably be due to a more ready penetration of the cells by the acid; on the other hand, since hemolysis is a complicated process, it is also possible that the salt might accelerate the destruction of the erythrocyte in some other way. It was in an attempt to decide between these two possibilities that advantage was taken of a method by which the penetration of the acid could be made directly visible. It was thought that in the presence of different concentrations of salt, a parallel between the time of color change and of hemolysis, with the M. H. JACOBS AND A. K. PARPART color change preceding hcmol\>is would indicate an effect of the salt on the actual rate of penetration of the acid. That such a parallel does in fact e\i-t i- -.In tun by tin- experiments to which reference ha-. already been made. In Table I, for example, it will be noted that concentrations of NaCl less than about 0.01 M have little or no effect on either process, but that at higher concentrations the acceleration of hemolysis very closely follows that of the color change. The question now arises why the entrance of acid and the onset of hemo!\>is occur more quickly in the presence of NaCl than in its .licence. Several conceivable explanations may immediately be di-missed as being decidedly improbable. For example, the case TAIU.I-; II Time (in seconds except where otherwise indicated) of Color Change and of Hemolysis with Different Concentrations of HCl in 0.15 M NaCl Concentration 01 HCl Color change (llC'lllolv/.r.] blood) i • K >t rli.mur (unhemolvzeil blood) Mi-L'inning of hemolysis 75 prr i-c'iii hemolysis pll after hemolysis 0.0 S \ Almost in- stantaneous 8 12 25 1.26 0.04 Almost in- stantaneous 25 29 35 1.55 0.02 Almost in- stantaneous 39 39 50 1.95 0.016 Almost in- stantam'ous 42 33 60 1.95 0.008 Almost in- stantaneous 46 44 90 2.25 0.004 AlmoM in- stantaneous 53 60 1 15 2.61 0.002 3 57 7J 194 3.06 0.001 9 72 75 232 3.78 0.0005 135 480 67(?) 232 4.62 0.0001 > 1 hours oo 00 oo — presents superficial analogies with the increased permeability of various cells in the presence of NaCl (Harvey, 1911 ; Osterhout, 1911, 1922, etc.); but it is easy to show that the accelerated entrance of acid into the erythrocyte and the subsequent hemolysis occur equally readily in the presence of pure isotonic Cad- or of physiologically balanced mixtures of N'aCI and CaCl-j. It has, in fact, been found iliat a considerable variety of salts are, without exception, effective in facilitating the entrance of HCl into the erythrocyte, though, as meniioiied abo\e, Millales in certain concentrations may inhibit the subsequent hemol\>i^. It is very improbable, therefore, that there i-- any close connection between this phenomenon and the older PERMEABILITY OF THE ERV'I'I I R( )C VTE 69 observations on the increase of permeability produced by sodium salts, which is a specific effect peculiar to these and perhaps a few other salts, and which is readily antagonized by calcium. Another explanation, applying specifically to the erythrocyte, was next considered, namely, that in the presence of salts some constituent of the cell surface might be removed, thus rendering the cell more permeable to ions. In this connection reference may be made to the work of Brinkman and van Dam (1920), who have reported that it is easy to remove lecithin from the erythrocyte by solutions of electro- lytes but not by those of non-electrolytes. An explanation of this sort was, however, rendered very improbable by a simple experiment TABLE III Time (in seconds except where otherwise indicated} of Color Change and of Hemolysis with Different Concentrations of HCl in 0.3 M Saccharose Concentration of HCl Color change (hemolyzed blood) Color change (unhemolyzed blood) Beginning of hemolysis 75 per cent hemolysis pH after hemolysis 0.32 N Almost 0.16 instantly Al most 12 31 52 0.82 0.08 instantly Almost 15 24 29 1.03 instantly 25 30 36 1.29 0.04 Almost 0.02 instantly Almost 56 59 69 1.50 0.01 instantly Al most 90 132 144 1.70 0.005 instantly 3 185 270 231 330 269 450 2.04 2.26 0.0025 75 720 960 — 2.56 0.0012 240 3 hours 4 hours — • 2.87 0.0006 2 hours 5 hours — — 3.30 which consisted in comparing the effectiveness of previous washings of the cells with isotonic NaCl solutions, with the actual presence of small quantities of the salt at the time of hemolysis. In one such experiment it was found that ox erythrocytes were no more readily hemolyzed in 0.3 M glycerol containing 0.015 N HCl after six previous washings in isotonic NaCl than before. If the observed effect were merely due to the removal of something from the cell surface, six washings ought to have been far more effective than was the presence in the solution at the time of hemolysis of as low a concentration of NaCl as 0.01 M or less; but this was not the case. From this and similar experiments the conclusion was drawn that it is necessary for 70 M. H. JACOBS AND A. K. PARPART the salt actually to be present with the HC1 in order to exert its characteristic influence. In an attempt to throw further light upon this question, systematic observations were made over a considerable range of acid concentra- tions both in the presence of NaCl (0.154 M) and in its absence, using for the latter purpose a 0.3 M solution of saccharose. Two such experiments, made on the same blood under as nearly as possible comparable conditions (except for a slight accidental difference, not believed to be significant, in the concentration of the blood in a few of the individual experiments) are summarized in Tables II and III. The data contained in Tables II and III, which were obtained before the subject had received any theoretical treatment, seemed at first sight somewhat puzzling; but, as will be shown, they have proved to be capable of a very simple semi-quantitative explanation on the basis of a hypothetical permeability of the erythrocyte to hydroxyl rather than to hydrogen ions. It will be noted that in the experiments in question, as in previous ones, color changes and hemolysis always occurred more rapidly in the presence than in the absence of NaCl, other conditions being the same. But not only were the rates of color change and of hemolysis slower in solutions of saccharose than in those of NaCl but in the absence of salt both of these processes entirely failed to occur at concentrations of acid that were otherwise always effective. In other words, it would appear that the salt, in addition to its other effects, influences the position of final equilibrium of the system. Another difference between the experiments represented in Tables II and III, whose meaning was at first far from clear but which we now believe to be of considerable theoretical significance, is that whereas in the non-electrolyte solutions, over a fairly wide range, the rate of color change and of hemolysis is roughly proportional to the concentration of acid (i.e., doubling the concentration of acid approximately halves the time required for the attainment of the chosen end-point), in the NaCl solutions the concentration of acid is ot much less importance. Thus, a forty-fold change in concentration (from 0.001 N to 0.04 N) is seen in Table II to decrease the time required for the color change only from 72 to 25 seconds and for the I ><•;: inning of hemolysis from 75 to 29 seconds. In an attempt to account for these various observations, we tunu • possible methods of pH change, would be determined by: [H]B0lution X [Cl]soiution " [H]ccll X and [OH]cdI X [Cl]solution "' [OH]soluti0n X [Cljccll respectively. Substituting in these equations the various concentra- tions given in Fig. 1, we have: .v- + xy - K)-14 - 4 ] !()-» 1 and (1H-14 \ C -y 1H-16 * + y- -}- -, (2) X / .A from which we may calculate, as has been done in Table IV, the relative tcmlcncic- for the erythrocyte to become more acid (or less PERMEABILITY OF THE ERYTIIROCYTK 73 alkaline) in solutions of different degrees of acidity in the presence of an approximately physiological concentration (0.16 M) of NaCl and in its absence. It is, of course, obvious that the figures in this table, taken singly, have no very great significance; and it is not even permissible at a given pH value to compare with each other the figures for hydrogen and for hydroxyl ions, since the actual rates of penetration of the cell by these ions will depend not merely on the mass law factors but upon the specific properties of the individual ions as well. But for each kind of ion separately over a series of different pH values, the figures have an important relative significance to which attention may now be directed. It wrill be noted in Table IV that according to the hypothes's of TABLE IV Mass law effect with 0.16 M NaCl and isotonic saccharose solutions at different pH values, for movement across the cell membrane (a) of H' with CV and (b) of OH' in ex- change for Cl'. Internal conditions as described in the text. 0.16 M NaCl Isotonic saccharose pH H- with Cl' OH' in exchange for Cl' H- with Cl' OH' in exchange for Cl' 7.0 12.00X10-9 24.00 X10~9 -4.00X10-9 -8.00X10-9 6.0 15.60X10-8 31.20X10-9 -4.00X10-9 -8.00X10-'° 5.0 15.96X10-7 31.92X10-9 -3.99X10-° -7.80X10-11 4.0 16.01X10-6 32.01 X10-9 6.00 X10-9 1.20X10-11 3.0 16.10X10-5 32.20X10-9 9.96X10-7 1.99X10-10 2.0 17.00X10-4 34.00 X10-;' i.ooxio-4 2.00 X10-9 1.0 26.00X10-3 52.00X10-" i.ooxio-2 2.00X10-8 permeability to hydrogen ions, the initial mass law factor, for the concentrations of acid actually employed in the experiments, should increase in the sodium chloride solution somewhat more rapidly than the concentration of hydrogen ions. In the saccharose solution, on the other hand, over the same range of concentrations the increase should be approximately proportional to the square of the hydrogen ion concentration. An examination of Tables II and III shows that in the salt solution the observed rate of penetration of the acid is only slightly affected by its own concentration, while in the saccharose solution it is roughly proportional to the first rather than to the second power of its concentration. Evidently the observed facts are in complete disagreement with the hypothesis of permeability to hydrogen ions. On the other hand, both in the presence and the absence of NaCl 7 74 M. H. JACOBS AND A. K. PARPART the rate of color change and of hemolysis are in good semi-quantitative agreement with the predicted mass law effects according to the hypothesis of permeability to hydroxyl ions. It is difficult to believe that this agreement is merely the result of chance. Admitting our ignorance of all but the probable beginning of the diffusion process, and making due allowance for the complicating effects of injury to the cells, especially in the more acid solutions, and the very rough nature of the calculations where so many simplifying assumptions have been made, it is nevertheless true that no other explanation of the facts has as yet been found which is at the same time so simple and so well in agreement with the other known properties of the erythrocyte. The entire failure of color changes and of hemolysis to occur in sugar solutions at acid concentrations which readily bring them about in the presence of NaCl is a necessary consequence of the general theory of ionic exchanges. It will be noted in Table IV that for sugar solutions of pH 7.0, 6.0 and 5.0 the mass law factors for both mechanisms of ionic exchange have negative signs. That is to say, in such solutions the total movement of hydrogen or of hydroxyl ions, as the case may be, must be in the opposite direction from that found in the remaining solutions. In other words, erythrocytes in such solutions should theoretically become more alkaline rather than more acid. This same problem may also be approached in a slightly different \\ay by introducing the idea of a final equilibrium. Though it is obviously impossible, because of imperfect knowledge of the behavior of hemoglobin and of the erythrocyte at decidedly acid reactions, to calculate the final equilibrium conditions in cells exposed to any desired acid solution, it is nevertheless possible to determine what external solution would cause no internal pi I change in normal cells. 1 or the simplified erythrocyte already dealt with, either equation (1) or equation (2) (omitting quantities which are negligibly small) yields the following equation for equilibrium: .v2 + .TV = 4 ! : K)-9. Solving for x, \ve have as the external hydrogen ion concentration in equilibrium with an internal pH value of 7 S: I'v means <.f ihi- equaiion the pi I values in Table Y have been calcu- lated. PERM KAMI 1. 1TY OF THE ERYTHROCYTE 75 It will be noted in this table that in the entire absence of salt the pH of equilibrium is 4.2, and therefore any solution less acid than this ought theoretically to cause the simplified erythrocyte to become more alkaline rather than more acid. For a concentration of NaCl of 0.0001 M, the critical pH is 4.5, and for one of 0.001, 5.4, etc. These figures, of course, cannot be expected to hold exactly in the case of actual erythrocytes where conditions are more complicated than those here considered. But the general principle itself, to which attention has already been directed by Netter (1928) and which has been put to a practical use by Bruch and Netter (1930), appears to be a sound one whose neglect has probably been responsible for con- siderable confusion in the past in experimental work with the erythro- cyte. TABLE V External pH in equilibrium with an internal pH of 7.3 with various external con- centrations of NaCl. Internal conditions as described in the text. Concentration of NaCl Equilibrium (mols. per liter) pH 4.2 0.0001 4.5 0.001 5.4 0.01 6.4 0.1 7.4 One further point about Tables II and III deserves mention- Not only is the color change of intact cells affected by the presence or absence of electrolytes, but a similar, though less marked, effect is observable in the case of hemolyzed blood. It is possible that we may here be dealing with a case similar to that reported by Adair, Barcroft, and Bock (1921), who found that even in blood hemolyzed by distilled water there is evidence that the cells, though invisible, may still be sufficiently well preserved to produce characteristic effects upon the dissociation curve of hemoglobin. It is not un- reasonable, therefore, to expect that even in hemolyzed blood there might be some evidence of the same salt-acid effect that is found with intact cells. Whether this is a complete explanation of the observed facts, however, or whether some additional principle is involved cannot at present be stated with certainty. SUMMARY 1. The penetration of acids into mammalian erythrocytes may be followed macroscopically by means of the color changes that occur when hemoglobin is converted into acid hematin. 2. The penetration of the acid precedes, rather than follows, 76 M. H. JACOBS AND A. K. PARPART hemolysis. In certain cases, penetration may be observed without subsequent hemolysis. 3. Over a considerable pH range, both in the presence and in the absence of NaCl, the rate of acid penetration into the erythrocyte, as inferred from the time of color change, is in semi-quantitative agreement with that predicted for a system permeable to hydroxyl and not to hydrogen ions. There is an entire lack of agreement with the theoretical behavior of such a system when the permeability to the t\\<> ions is reversed. The simplest explanation of the observed facts is that the hydrogen ion, like other cations, is unable to enter the- erythrocyte easily. I1IBLIOGRAP1IY ADAIR, G. S., J. BARCROFT, AND A. V. BOCK, 1921. Jour. Physicl, 55: 332. HKINKMAN, R. AMI I.. VAN D\M. W20. Biockcni. Zcitschr., 108: 35. I'.Ki . ii, H. AND H. XETTER, 1930. Pfliiger's Arch., 225: 403. HAKVI.Y, K. X., 1911. Jour. Exper. Zool., 10: 507. HENDERSON, L. J., 1928. Blood: A Study in General Physiology, Xe\v Haven. JACOBS, M. H., 1930s. Am. Jour. Med. -SY/., 179: 302. JACOBS, M. H., 19306. Biol. Bull, 58: 104. JACOBS, M. H., 1931. Ergebn. d. Biol., 7: 1. KERR, S. E., 1979. Jour. Biol. Chem., 85: 47. NETTER, II., 1928. Pfliiger's Arch., 220: 107. OSTERHOUT, \V. J. Y., mil. Science, X.S.. 34: 1X7 OSTERHOUT, W. J. Y., 1922. Injury, Recovery and De.ith in Relation to Conduc- tivity and Permeability. Philadelphia. VAN SLYKK, 1). D.. II. \Vr, AND F. C. M( I i N\. l(>-'3. Jour. Biol. Chan., 56: 765. WARBURG, !•-. J., \()22. Bi/iihcin. Jour., 16: 153. ASCIDIANS OF THE BERMUDAS N. J. BERRILL McGiLL UNIVERSITY, MONTREAL The ascidians of the Bermuda islands have already been described by Van Name (1902) and a revision published by the same author in 1921 under the title "The Ascidian Fauna of the West Indies." In view of this the present paper is confined to a brief account of the fauna as a whole, a description of several new species, and a few notes upon certain other forms. The ascidians of the Bermudas are almost identical, species for species, with those of the West Indies and Gulf of Mexico, and there is little doubt that, together with the bulk of the Bermuda fauna, they have been carried there from those regions by the west Atlantic drift. Like ascidians throughout the world, their habitat is determined as a compromise between two factors, their ability to withstand wave- action and exposure, and their need for a good flow of well oxygenated water. They may accordingly be divided into three groups, those to be found only under stones, those attached to the upper surfaces of stones, or to various corals and sea-fans, and those found beyond the outer reefs. Apart from such classification of habitat further description such as specified localities is practically unnecessary. Species found under stones in one locality are likely to be found in similar positions through- out the islands. The same holds true for forms that are attached to exposed surfaces. The material upon which this paper is based was obtained through the facilities of the Bermuda Biological Station, St. Georges, and also through the assistance of Louis Mowbray, Esq., of the Government Aquarium. FAUNA LIST Aplidium (A maroucium) berrmidae (Van Name) on corals Aplidium (A maroucium) exile (Van Name) under stones Trididemnum savignii (Herdman) on algae, under stones Trididemnum orbiculatum (Van Name) under stones Didemnum candid-urn Savigny under stones, on corals Didemnum (Polysyncratori) amethysteum (Van Xame) under stones Leptoclinum macdonaldi (Herdman) on corals, etc. Lissoclinum fragile (Van Name) under stones Polycitor (Eudistoma) olivaceus (Van Name) on stones, corals, etc. 77 78 N. J. BKRRILL Polycitor (Eudistoma) convexus (\'an Name) on corals, etc. Polycitor (Eiidistonni) flurus i Van Name i on stones, corals, , t, . Polycitor (Eudistomn^ t'nf>snlli<>ni riridis Verrill under stones Perophoru bcrnntdfiisis n.sp under stones einasi nlin turl>hiutIi'»nia viridt- I lerdman under stones Polyandrocarpa < l''.nsyii.\lyi-ln ) tint In (Van Name) under stones Polycarpa obtectti Traustedt on stones, corals, reefs .V/v(7(/ [>iirtilti (Stimpson) on stones, corals, reels Pyitra villain f'St impson) on stones Microcosmus i:\asperutns 1 Idler under stones, on n-i ! - The Bermuda fauna is exceptionally rich in members of the family Perophorida;. Two well-known forms, Ecteinascidia tnrbinata and I'rropliom riridis, are common. Their close relationship, however, is emphasized by the discovery of transitional types. These new forms may frequently be found growing under the same boulder as those two species. Inasmuch as the existing descriptions of Ectehuiscidiu tnrhiiiutn are of immature xooids, a somewhat detailed account is made here ot this species; such an account also forms a basis for comparison ot the other forms. Ecteinascidia liirhinntii I lerdman. This species was originally described by llcidman Irom material collected at P>ermuda by the Challenger expedition. Large orange colonies develop ,m,i< lied to the upper surfaces of rocks where currents are relatively strong. The essential organi/atioii of the /ooid is shown in Kig. 1. The structure shown here for the first time are the heart, and the gonads \\ith their dm ts. The heart is of the same relative size and length, and 1^ in t he same relat ive position as in species of Ascidia or Phallusia. The gonads develop iii the intestinal loop on the left side, as in Pero- plioi.i, the ovary bein^ surrounded by a ring of testicular lobes. The vas deferens during its distal course follows the intestine and they open together some distance short of the atrial siphon. The o\idii(t, on the other hand, is relatively short and wide. It passes ASCIDIANS OF I5KKMUDAS 79 Ml N. J. BERRILL from the ovary to the dot sal lamina, opening beyond it into the atrial cavity of the right side. A single egg at a time is forced up the duct. During this brief passage it becomes greatly elongated through compression, rapidly regaining its spherical shape as it emerges. The above features are generic, the characters of specific importance being as follows, — the tendency to grow on the exposed surfaces of rocks, the red or orange colour, the proximity of the atrial to the branchial siphon, the possession by the adult zooid of about thirty row- of stigmata, and the formation of tadpole larva? with twelve rows of stigmata. The breeding season extend- from about June first to August or September. A.'< friiiiis,-i(liitn. Probably the most definite feature distinguishing both E. conklini Ivpii/i and ininnln from /•-'. tnrbinatti is the curvature of the intestine in the former, tending to form a secondary loop. The intestine of I:, tnrhiiuitd is but very slightly curved in comparison. 1'rni/tlnirn In-rnindoisis n. >p. This species forms mixed colonies with the common Perophoraviridis, both growing in great profusion on the under surface of stones where llieie is .1 good circulation of water. ASCIDIANS OF BERMUDAS 81 82 N. J. BERRILL Perophora viridis has a greenish tinge, as its name implies, and both oozooid and blast o/ooid have four rows of stigmata. Perophora listen is similar in that the greenish colour is quite absent, but it forms by no means such dense colonies. Perophora annectens differs from both these species in that it has but three rows of stigmata. rcroplwni hcrmndctisis possesses blastozooids with five rows of sti.umaia .although the oo/oid has but four) and is accordingly unique. It may be readily distinguished from P. viridis with which it is usually entangled, by the absence of any green colour and the somewhat larger si/e of the zooids. Its structure is shown in Fig. 3, and it is seen that not only are there five rows of stigmata but that about a third of the stigmata of Fir,. 3. A. Id. i st( )/<)<") id of Perophora hcrnnult'iisis n.sp. 15, oo/ooid of J-'.ctcin- •sis In-rdniuni I.ahillr. the lirst and second rows are common to both. This condition prevails in the smallest and youngest blastozooids in which stigmata are discernible. Perophoropsis hcnlnuuii I.ahille. This form uas described by Lahille in 1S()() tVoin Rain'uls and has in it been rediscovered. It differs from specie-, of rcrophora in that theie are at least ten rows of stigmata, h Is a genus therefore inter- mediate between /'i'n>/>hi>nt and Ecteinascidid . At Bermuda, mixed \\itli colonies of /'I'ro/thoni viridis, P. bermn- densis, and Ecteinascidia <~<»iklini, were found individuals corresponding in structure to Perophoropsis; these are shown in Fig. 4. ASCIDIANS OF BERMUDAS There is, however, another possibility. It was determined in the case of Ecteinascidia turbinata, E. conklini, Perophora berrmidensis, Clavelina lepadiformis, Botryllus gigas, and Botrylloides leachii, that in developing blastozooids the number of rows of stigmata, while yet non-functional although perforate, equals or may slightly exceed the number of stigmata rows in the zooid from which they arose. In no case was it less than that number. This statement is probably true for ascidians as a whole and will form the subject of further investi- gation. This phenomenon accordingly rules out the possibility that the Perophoropsis-like forms were young blastozooids of Ecteinascidia; while the fact that the oozooid of E. turbinata possesses from the start twelve rows of stigmata makes it impossible that they are young oozooids of that species. With E. conklini the case is different. The oozooid first becomes active with but six rows of stigmata, and ac- cordingly it is possible that the forms in question are slightly-grown individuals of this type, especially as they were solitary and without budding stolons, although occurring not far from one another. An- other feature which supports such a contention is that the individual stigmata are but half the length of those of Ecteinascidia or Perophora species. Whether they are blastozooids or oozooids, they are probably but half-grown. Provisionally, therefore, these perophorids will be assumed to be young oozooids of Ecteinascidia conklini rather than the Perophoropsis of Lahille. If this is the case, then it should be noted that additional rows of stigmata appear in the oozooid when the stigmata of the original six rows are merely a fourth of full adult size, and cannot be distin- guished from those first six rows. Genus Clavelina. There is some confusion as to the number of species of Clavelina to be found in the Bermuda and West Indian region. From material taken on the Challenger expedition at Bermuda Herdman described a form that he named Clavelina oblonga. In 1900 Verrill, in his general account of the fauna of the Bermudas, described an ascidian under the name Diazona picta. In the account of the Bermuda ascidians published by Van Name in 1902 this last was changed to Rhodozona picta on the ground that it differed materially from the type of Diazona and was intermediate in character between Diazona and Clavelina. In his later account (1921) of the ascidian fauna of the West Indies Van Name includes both these forms under Clavelina oblonga, the Diazona picta of Verrill being considered to be large colonies and Herdman's Clavelina oblonga young colonies of 84 N. J. BERRILL one and the same specie-. This he describes as a form in which the zooids of a young colony are almost completely separated from one another but which in older colonies become enclosed for their greater part in a common gelatinous test. Such a ci inclusion is believed to be the result of two factors, the study mainly of preserved material, and of material collected not later than Mav, that is, of material somewhat immature. Preservation destrcn - pigmentation and in these two forms results in a tremendous shrinkage of the thoracic part of the zooids. These forms are believed to be both cla\ elinids but also to be quite distinct species. On this assumption then the one should be known as Clavelina oblonga Herdman and the other as Clnr/-liini f>i< tu (Yerrill). It is hoped that the following descriptions will show this to be the case. Clavelina oblonga Herdman. Habitat. Attached to the tinder-surface of stones near low-water level, very rarely in more exposed positions. Breeding season. End of April until August. Form of colony. • Number of zooids rarely exceeds 40, usually much less. They are at- t ached to a basal stolon, but other- ui-e are separated from one an- other. Budding occurs throughout breeding season so that zooids of all an- to be found. Pigmentation. Pesl < i •. stal-clear, l!i am-hia! s K- with Hecks of whin- pigment near iis anteiior end. Abdominal region yellowish. In living state whole colony perfectly transparent. :>nl tuilfwle size. 'i •! 1 mm. di.unei . i i gg). J.J5 mm. tui.il length (tadpole). liiri-nl development. 'r,id|ii)le .Hid young oo/.ouid h.ive Iroiu the hi'st. tuo row-, o| delmitive i Clavelina picta (Verrill). Habitat. Attached to sea-fans, corals, submerged wrecks, etc., from low- water level to a few fathoms; never found under stones; a good flow of water seems to be necessary. Breeding season. End of June until August or September. Form of colony. Colonies usually possess more than 40 zooids and may have as many as 1000. The zooids are embedded in a common test, which is usually divided into conns. During the summer the thorax of each zooid extends beyond the common test, at other times it is usually completely embedded. Budding does not occur during the breeding season or the months preceding it and so the size of the zooids is very uniform in any one colony. Pigmentation. Test unpigmented but slightly opaque. En- dostyle, dorsal lamina, and pel ipliaryngeal bands densely coloured with purple or carmine cells. They also extend throughout the ab- domen and then accumulate in t he ends of 1 he tesl vessels. They appear in the late embryo in such places and the oo/onid is as highly coloured as the Egg and tadpole size. 0.4() mm. diameter (egg). 3.30 mm. total length (tadpole). Post-larval development. Tadpole and young oozoiiid have from first, four rows of definitive stigmata. other species of Clavelina have but two. the All ASCIDIANS OF BERMUDAS 85 Clavelina oblonga usually occurs in groups of two or three during May and increases in number throughout the breeding season until about forty zooids of various size may be found in one colony. This was found to be the case in both 1930 and 1931, and the implication is that the oozooid forms a few winter statoblasts which regenerate and give rise to the type of colony just described during the following spring, while such a colony dies off at the end of the summer, failing to form a second generation of statoblasts. That is, the life-cycle is completed within 18 months; otherwise, large colonies should be found similar to those of Clavelina lepadiformis, which is not the case. That Claielina picta was originally taken to be a diazonid is due FIG. 4. A, colony of Clavelina oblonga Herdman. B, C, egg and tadpole of same. D, part of a colony of Clavelina picta (Verrill). E, F, egg and tadpole of same. G, H, egg and tadpole of Clavelina lepadiformis. The eggs and tadpoles are drawn to the same scale. to a superficial similarity in type of budding. In Diazona violacea budding by abdominal strobilation occurs after the breeding season has ended, and only then. So that every autumn each zooid forms eight or ten compact bodies which regenerate slowly through the winter and spring to become sexually mature in the late summer; the process is then repeated. Therefore the zooids are uniform in size and remain embedded in the common test, although the anterior part of the thorax of each extends separately during the late summer. In Clavelina picta winter statoblasts are formed after the breeding season and apparently at no other time. The oozooid probably forms about ten such bodies. These regenerate, grow and become sexually 86 N. J. BERRILL mature during the winter and spring. They carry on these processes close together within the test of the oozooid, and so remain within a common test. During the late summer these processes are repeated, resulting in the form.n ion during the following year of a colony composed of about ten corms, each containing ten or more zooids. This may be repeated a third year to form ma>H\e colonies. During maturity the anterior parts of the zooids extend from the common test just as in Diazonu, and as in Ditizona they contract and degenerate after the breeding season while the thoracic extensions of the test are sloughed off. Thus, apart from the difference in budding, strobilation of the abdomen in Diuzona and the formation of post-abdominal extensions in Clai'dina, the cycle is much the same in the two genera .ind it is understandable that a superficial similarity results. The separation of the zooids in large colonies of Clavelina lepadiformis in contrast to the condition in C. picta is due probably to the very thin test of the former, so thin that when the statoblasts grow into new individuals they form independent vertical extensions of it from the first. Budding in these forms wrill be the subject of a much more detailed investigation. riride Herdman. This species was described first by Herdman from material collected at Bermuda by the Challenger expedition, and has since been dis- covered throughout the West Indies, and in the East Indies from the Philippines to the Red Sea. A full description with illustration of a xooid is given by Van Name (1()21). This form has at various times been included in the StyelicLe and in the Botryllida'. A few additional observations made at Bermuda merely emphasize its relationship to both these groups. In fact it so completely bridge^. the gap, both in its adult structure and organization and in its development, between these two families that their main- tenance a> distinct families is an unnatural classification. It is .icioidingly proposed that the Botryllid.e be subordinated within the Styelid.i , either as a sub-family or merely as constituent genera. The method of budding in its fundamentals is identical in the I'olystyelid.e, Symplegma, and Botryllidic. Syniplf^mn resembles polystyelids such as Polyandrocarpa, Dis- , or Stnlntiii-ii inasmuch as each zooid has its own atrial siphon, clo.ical cavities not being formed. Its general anatomy Mibles that of Polynndrocarpa or Polycarpa except that its in- dividuals are much smaller. Correlated with this reduction is a ASCIDIANS OF BERMUDAS 87 diminution in number of polycarps (to a single polycarp) and in number of rows of stigmata. Its structure differs from that of botryl- lids only in the absence of common cloacal cavities, and consequently in the lack of obvious systems in the colony. In other words, common cloacal cavities and the arrangement of zooids into systems are the only major features that separate the botryllids from other styelids. Apart from these two characters Symplegma is so similar to Botryllus and Botrylloides that it exhibits the same degree of variation in colour and form of whole colonies. These varieties are even more definite and constant in Symplegma than in the Botryllidae, the commoner A CD FIG. 5. A, colony of Symplegma viride stolonica n.v. B, colony of Symplegma viride typica. C, zooid of stolonica. D, zooid of typica. E, F, G, tadpoles of Symplegma viride, Distomus variolosus, and Botryllus gigas, respectively, drawn to the same scale. colour varieties being, — green, green and white, green and brown, black and white, and orange. With the exception of the last, all colonies are very compact and sheet-like as in botryllids, the adult zooids being pressed together, young buds appearing in large numbers together with clusters of ampulke at the colony margin. The orange variety, however, is different. Instead of the budding stolons being very short, they are even longer than in Stolonica socialis, with the result that the zooids are often widely separated from each other wrhile the individual zooids assume a comparatively vertical position. It may, in view of this difference, be worth recognizing the orange N. J. BERRILL form as a definite variety. If so, it is suggested that it be named Symplegma viride stolon ica. The essential unity of such forms as Symplegma, Botryllus, and Distomus is perhaps best shown in the study of the development. The tadpole of Symplegma is almost identical with those of Hix- tonuis, Stolonica, and Styelopsis, except that it is somewhat smaller. It has tin- same degree of organization ; it has the same single but coinpoHte sense organ (probably developed from the otolith after the primitive eye was lost, an otolith and eye being found only in Styela itself); and in Symplegma and Distomus there is a long anterior mental process surrounded by a ring of ampulla-. Botryllus and Botrylloides both produce tadpoles with a similar mental process and ring of ampulhe, there being eight ampulla' in each of these genera and also in Symplegma. While the botryllid tadpoles are the more highly organized, the young oozooids developing from these and from tadpoles of Symplegma are hardly to be distinguished. Altogether there are greater differences between a genus such as Styela and the polystyelids than there are between these last and the botryllids. REFERENCES I. Aim. LI-:, F., 1800. 1\( . In n lies sur les Tunirirr^ des cotes de France. Toulouse. VAX XAMH, \Y. ('-., \<)()2. The Asridi.ms of the Bermuda Islands. Trans. Conn. A fad. Arts and Sfi., 11: 325. VAX XAMI-, \\ . G., 1921. .Widiansof the West Indies. Bull. Am. MHS. Xat. Hist., 44: 283. THE LIFE HISTORY OF EPIBDELLA MELLENI MACCALLUM 1927, A MONOGENETIC TREMATODE PARASITIC ON MARINE FISHES THEO. L. JAHN AND L. ROLAND KUHN DEPARTMENT OF BIOLOGY, UNIVERSITY COLLEGE, NEW YORK UNIVERSITY AND THE NEW YORK AQUARIUM At present our knowledge concerning the life histories of the members of the order Monogena is extremely scanty. Although the life histories of the monogenetic trematodes are apparently much simpler than those of the digenetic forms, they have been less exten- sively studied, and definite information concerning them is, with a few exceptions, fragmentary. Of the sub-order Monopisthodiscinea Fuhrmann the only life history known is that of Gyrodactylus (Katha- riner, 1904). Of the sub-order Monopisthocotylinea Odhner, which includes Epibdella, no life histories have been reported. In the sub- order Polyopisthocotylinea Odhner two life histories, those of Poly- stomum integerrimum (Zeller, 1872, 1876) and Diplozoon paradoxum (Zeller, 1872a), have been reported. The dimorphic development of Polystomum integerrimum on the gills of the tadpole and in the urinary bladder of the frog as given by Zeller (1872) has been questioned by several workers, including Stunkard (1917) who states that "the findings of Zeller are so unusual that one is led strongly to suspect he confused two different species." This leaves our definite knowledge of monogenetic life histories limited to two genera, Gyrodactylus and Diplozoon. The present investigation of Epibdella melleni MacCallum 1927 presents the third definitely-known life history of the Monogena and the first of the sub-order Monopisthocotylinea. The adult of Epibdella melleni was first described by MacCallum (1927) as parasitic on the Pacific puffer (Spheroides annulatus], the spadefish (Chaetodipterus faber), and various species of angel fishes (Angelicthys and Pomacanthus) from the New York Aquarium. Due to the necessarily closed salt water system of the Aquarium, the infection spread rapidly to all tanks which contained susceptible fishes, and for several years it has been a very grave source of danger to all susceptible fishes. This study was undertaken at the suggestion of Mr. Charles M. Breder of the New York Aquarium and Professor H. \Y. Stunkard of 8 -89 90 THRO. L. JAHN AND L. K. KUHN \e\v York I'niversity. The writers wish to thank Mr. Breder for his constant aid in the collection of material, for the identification of the fishes, and for information regarding their susceptibility, and Professor Stunkard for his helpful suggestions during the preparation of the manuscript. This study was made possible by a grant of the New York Zoological Society in the form of a Research Fellowship which was held l>y the senior author during the summer of 1930 and by the junior author during the* academic year 1930 31. MATERIAL AND MKTIIODS All material used for this study was obtained from infected fishes at the New York Aquarium. Adult specimens were obtained by scraping the body and cornea of the fish with a scalpel. In this way considerable mucus was obtained along with the parasites. This material was placed in small stender dishes, covered with sea water, and allowed to stand for ten minutes or longer. It was found that the parasites became firmly attached to the bottom of the dish, and that the mucus material could easily be removed with a pipette. After several changes of sea water the organisms, attached to the dish, could be studied with the aid of a binocular. In this way observations of the process of egg-laying were possible. It was found that the organisms remained attached to the dish after fixation. Therefore they were fixed, stained, destained, and cleared while attached by changing the fluid in the stender dish. They were then removed to a slide with the aid of a pipette and covered with balsam. Schaudinn's fluid and saturated aqueous solution of mercuric chloride plus five per cent acetic acid were used as fixatives. Whole mounts were stained with paracarmine, and sections with I H-lalield's h.i-mato\ylin and eryth rosin. Larvae were obtained by collecting eggs and allowing them to hatch. Whenever larv.e were wanted, adult worms were collected and allowed to remain in stender dishes. The worms rarely lived more than twenty-foul- hours, but during this time numerous eggs weir usually laid. These were removed to .mother dish in which the water was changed several times a day. At the end of five to eight days the eggs hatched, and the larva: could be isolated with the aid ot a mouth-controlled pipette. MORPHOLOGY OF TIIK Ann/r The adult of J-'.f>i/>t ribution of >licll material in the ootype at the time the egg i^ -haped. In >ome instances slightly larger and somewhat globiform ire ] i! od need i Fig. 11). These seem to result from a lack of LIFE HISTORY OF RPIBDELLA MELLENI MACCALLUM 97 sufficient shell material or from the presence of too much yolk in the ootype at the time of egg formation. The normal eggs are yellow in color when formed but turn to a golden bronze shortly after they are laid. HATCHING OF THE EGG, AND BEHAVIOR OF THE LARVA Eggs of Epibdella melleni, isolated into small stender dishes in which the sea water is changed daily, hatch at room temperature in five to eight days. On the fourth to sixth days the larvae may- be seen within the eggs, and if an egg is broken open the larva is found apparently completely formed but non-motile. Just before hatching, the larva may be seen squirming about within the shell. A small circular opening is formed in the rounded corner of the egg, that is, on the corner which bears neither filament nor hook, but a preformed operculum was not observed. The larva emerges anterior end first. The anterior ciliated portion is thrust out, and the cilia meanwhile beat very rapidly. The larva may remain in this position for fifteen minutes. Then the beating of the cilia slowly pulls the organism out until only the posterior sucker remains within the shell. At the end of another fifteen minutes the larva is usually completely emerged. However, the time required for hatching is presumably affected by numerous undetermined factors. The larva, after emergence, swims very rapidly through the water, pausing only momentarily now and then on the bottom of the dish or on solid objects in the water. Within six hours, if the larva does not become attached to a suitable host, it apparently becomes exhausted, the rate of movement decreases, and the organism finally settles to the bottom, capable of only a slow crawling movement. This is presumably due to a loss of the ciliated epithelium, since such larva? are always almost completely deciliated. Up to this time the posterior sucker does not appear to be functional. MORPHOLOGY OF THE LARVA The ciliated larva of Epibdella melleni (Fig. 4) is approximately 225 microns in length and 60 microns in width. It is slightly flattened dorso-ventrally in the anterior region, and, otherwise, is fusiform in general shape except for a constriction in the region of the mouth. The posterior third of the larva is composed of the posterior sucker which, at the time of hatching and for some time afterward, remains folded in such a manner as to be non-functional as a sucker. The anterior two-thirds of the larva is composed of what becomes the body of the adult, and it is, in general structure, similar to that of the adult except that the digestive and excretory systems are simple and the reproductive systems are not yet developed. The oral sucker is round 98 THEO. L. JAHX AND L. R. KUHX and muscular and is located in the anterior part of the middle third of the body proper. It opens into a very short oesophagus which leads into two relatively large digestive caeca. These continue latero- posteriorly almost to the end of the body. No diverticula are present. Anterior to the mouth on the dorsal surface are two pairs of eyes. These are cup-shaped masses of pigment from the cavity of which protrude spherical hyaline lenses. Kach of the posterior pair of eyes is approximately 16 microns in diameter and is directed anterio- laterad. Those of anterior pair are 12 microns in diameter and are directed posterio-laterad. The anterior suckers are not distinct sucking disks but are pad-like muscular areas, easily distinguishable in the living form. The excretory system is composed of two relatively large excretory vesicles which are located slightly posterior and laterad to the mouth, four ducts which lead from them, and ten pairs of flame cells. The vesicles are quite distinct in the living organism and appear as large vacuolar structures. The excretory system opens by dorsal pores located on either side of the mouth as in the adult. The excretory system is shown diagrammatically in Fig. 7. Two large excretory channels extend, one from each vesicle, posteriorly into the posterior sucker. They are joined with each other by a cross-channel at the posterior end of the body proper. Each of these channels gives off an anteriorly directed branch which ends in a single flame cell about halfway between the vesicle and the cross-channel. The longitudinal channels continue into the sucker where each branches and end< in five flame cells as shown in Fig. 7. Leading anteriorly from the excretory vesicles is another pair of channels, one from each vesicle. These unite anterior to the mouth to form one median channel which continues forward between the eyes and branches laterally, and then each branch bifurcates anterio-posteriorly and ends in flame cells. In the region of the mouth are four flame cells, one on each side of the anterio-lateral and posterio-lateral sides of the oral sucker. The channels of these cells are probably branches of the anterior longi- tudinal channels, but this could not be determined with certainty due to the thickness of the oral sticker and the proximity of the ducts to it. 'I hese supposed connections are shown in Fig. 7. The arrangement of the excretory ducts of the larva is virtually a 'ring' system with two Literal excretory pores. The longitudinal channels and cross-con- nections of the larva have been checked in older specimens and have 1-een tiiuiid to he the main excretory channels of the adult. The organism bears cilia in the anterior, middle and posterior The anterior ciliated region extends from the anterior pair LIFE HISTORY OF EPIBDELLA MELLENI MACCALLUM of eyes forward, and cilia cover practically all of the anterior region except the sucking pads. The middle ciliated region, extending from the posterior edge of the excretory vesicles almost to the posterior end of the body proper, is covered with cilia on the lateral, latero-dorsal, and latero-ventral surfaces. No cilia were seen on the mid-dorsal and mid-ventral surfaces in this region. The posterior ciliated region includes the lateral and dorsal surfaces of the posterior two-thirds of the posterior sucker. The cilia are relatively long and arise from an epithelial layer which can be seen distinctly in living specimens. This epithelial layer contains many large highly refractile granules of unknown function. When cilia are lost, the epithelial layer from which they arise is shed with them. The posterior sucker, as folded when the larva is free-swimming, contains the definitive spines characteristic of the adult. These lie in a longitudinal position with the curved ends directed mediad as seen from the ventral surface. At the margin of the folded sucker, and lying ventrad to the definitive spines, may be seen the accessory or larval hooks. These are all the same size and measure approximately 10 microns in length. A single hooklet is shown in Fig. 13. When maintained under a sealed coverslip, the ciliated larva may be seen to lose its cilia and to assume a shape and position more characteristic of the adult. The anterior sucking pads are usually very active in fresh preparations; these are stretched forward and are attached to the slide, and the body may be drawn afterward. The organism is capable of moving about in this fashion as well as by the use of the cilia. After this type of movement has been continued for some time, the ciliated epithelium begins to slough off, leaving areas of the normally ciliated region devoid of cilia. Apparently the first regions to become deciliated are those in the vicinity of the anterior sucking pads and those which cover the posterior sucker. However, the other ciliated epithelium is shed shortly afterward. If free- swimming larvae are selected at random and examined, many are found which do not bear their full quota of cilia due to the sloughing of the epithelium. This may give rise to considerable confusion concerning the normal distribution, and it was necessary to examine a number of specimens in order to obtain the distribution shown in Fig. 4. Concomitantly with the sloughing of the ciliated epithelium, the posterior sucker is unfolded. As it is spread out, the definitive spines turn upon their longitudinal axis so that the curved ends are directed laterad, and the accessory hooks are pointed radially around the margin of the sucker. At this time the posterior sucker is functional and becomes firmly attached to the slide. Such an 100 THEO. L. JAHN AND L. R. KUHX organism is shown in Fig. 5. This drawing was made of a larva which was isolated when ciliated and which had undergone the above trans- formations in about forty -live minutes while under observation. In an attempt to interpret these transformations in relation to the normal life history of the organism, it is assumed that the larva is free-swimming until it comes in contact with a susceptible fish. Attachment is first by means of the anterior sucking pads. Then the cilia are lost, and the posterior sucker is unfolded, and a firmer attachment is afforded by means of the fourteen accessory hooks of the posterior sucker, aided perhaps by the most posterior pair of definitive spines. DKYKI.Ol'MKXT OF THE ATTACHED FORM After attachment the first morphological change to be noticed is the development of the anterior sucking pads into definite suckers. Then the mouth, which is round in the ciliated form, becomes lobate in outline as is that of the adult, and the digestive caeca show signs of becoming diverticulated. The posterior sucker becomes slightly larger in proportion to the size of the body, and the large spines increase in size and change in shape. A specimen in this stage that was obtained from an infested fish is shown in Fig. 6. No increase in size of the larval or accessory hooks could be noted, even when measurements of these hooks in young larvae and in adult organisms were compared. All accessory hooks measured were between 9.8 fj. and 10.7 /j. in length. For this reason it is believed that these hooks are not of especial importance in the adult. Further development of the organism seems to involve principally a further bifurcation of the digestive caeca, an elaboration of the excre- tory pattern, and the development of the reproductive system. The relatively large amount of space occupied by the reproductive system in the adult is responsible for the great differences in general appearance of the young ( Fig. 6) and of the adult forms. The sizes of the eyes change very little during the growth of the organism. Measurements made of the eyes of newly attached larvae and of adult specimens showed no significant difference in size. The more anterior pair of eyes was found to be 10 1 S /i in diameter, and the posterior pair was 15 l! t lie adult s are usually not spherical but are flattened in one direct ion or another. The eyes of some adults even appear conical with the base oi t he rone adjacent to the lens. LIFE HISTORY OF EPIBDELLA MELLENI MACCALLUM 101 During the growth of the individual the relative amounts of growth of the three pairs of spines are not equal. The middle and the more anterior pairs grow at a rate that is relatively about twice that of the most posterior pair. Thus, the sizes of the first two pairs in the adult are over six times that of the same pairs in the youngest attached form, while the size of the most posterior pair is only three times that of the young form. Table I shows the average measurements of the three pairs of spines in young, medium-sized (about 1.5 mm.), and adult (over 3.5 mm.) specimens, and also the ratio of the sizes of the spines in the various sized specimens, the youngest forms being used as unity. The figures are the averages of the measurements of six specimens, and all three pairs were measured on the same individual. Therefore the ratios are strictly comparable. OCCURRENCE AND PATHOGENICITY When the adult of Epibdella melleni, obtained from the tanks of the New York Aquarium, was first described by MacCallum (1927), TABLE I Average size of spines in mm. Ratio of the size of the spine to that of the young attached form Pair of spines Anterior Middle Posterior Anterior Middle Posterior Young .035 .109 .216 .045 .174 .297 .038 .086 .118 1.0 3.1 6.2 1.0 3.9 6.6 1.0 2.3 3.0 Medium Adult.... it was stated that the infection was probably introduced by a Pacific puffer (Spheroides anniilatus] from California. However, it is the belief of Mr. Breder and his associates that there was no Pacific puffer in the Aquarium for some time preceding the discovery of the parasite and that the parasite was originally introduced and is being continually reintroduced with shipments of fishes from Key West and Nassau. Furthermore, the parasite occurs in both the Chicago and Philadelphia Aquariums, neither of which has Pacific fishes. Also, it is known that E. melleni will not survive very long exposure to the acid water of New York harbor and that it will multiply rapidly in the neutral tank water of the Aquarium. For many years no attempt was made to control the chemical composition of the water in the tanks of the New York Aquarium, and only after the installation of an efficient means of chemical control did the parasites become numerous, probably due to the very high acidity of the water previous to that time. For these reasons it is believed that the parasite is a West 102 THEO. L. JAHN AND L. K. KUHN Indian species and that it might have been continually present in small numbers in the New York Aquarium for many years before its discovery in 1927. The li>ln> which have been found to be susceptible to infection with Epibdclla melleni are as follows: Subclass Teleostomi ( )nli-r Acanthopteri Family Carangidae Caranx crysos (Mitchill), Runner (X) Caranx hippos (Linnaeus), Common Jack (X) Naucrates d actor Linnaeus, Pilot Fish Trachinotus carolinus (Linnaeus), Common I>oni])ano '1'rachinotus glaucus (Hloch), Old Wife or I'alomeia Vomer setapinnis (Mitchill), Moonfish Family Pomatomidae Pomatomus saltnlrix (Linnaeus), Bluefish Family Serranidae Centropristus striutns (Linnaeus), Common Sea Bass (X) Dermatolepis pnnctalns Gill, Spotted ( '.rouper (Pacific) (X) Epinephelus adscensionis (Osbeck), Rock Hind Epinephelus guttatus (Linnaeus), Red Hind Epineplielns morio (Cuvier and Valenciennes), Red Grouper (X) Kf'iin-f'hi'lns si rial us (Bloch), Nassau Grouper (X) Paralabrax maculatofasciatus (Steindachner), Spotted Cabrilla (Pacific) Promicrops itaiara (Lichtenstein), Jewfish Family Lutianidse Li/tinniis iiHulis (Cuvier and Valenciennes), Muttonfish Lutianus apodus (Walbaum), Schoolmaster Lutianus jocu (Bloch and Schneider), Dog Snapper Luliuniis synagris (Linnaeus), Spot Snapper (X) Family I hcmulidse Anisotreniiis surinamensis (Bloch), Black Margate Anisotrcnnts virginicus (Linnaeus), Porkfish Iltcnnilon album Cuvier and Valenciennes, Margate Family Sciii-nida- Mi'nlii irrlnis saxatilis (Bloch and Schneider), Kingfish Miiropogon nndnluliis (Linnaeus), Croaker Family Labrida1 Lachnolaimus uinxhuns f\\';.lbaum), Ilogtish Tdutnyi diiiti\ ( l.innaen^:, 'I'autOg l-.imily L|ihip|ii(].r Chcetodipterus faber (Broussonet), Sjiadeiish Family < 'li.Hodnnt id.r Angelichthys n'linris (Linnaeus), nm-en .\n-e]ti>h Angelichthys isnliclitn Jordan and Ritter, Blue An^ellish Angelichthys townsendi Nichols and Mowbray, Townsend's Angelfish todon ocellatus Bloch, Common Butterfly Fish Pomacanthus an nntns (Linnaeus), Black Angellish Pomacanthus paru (Bloch), French Angelfish 1 amily Acanthuridae .luiiilliiirus candeus Bloch and Schneider, Blue Tang .\iiinllniriis hepatus (Linnaeus), Brown Tang or Doctor Fish Lamilv Ballot id.r Hiilittrs vt-tnlii Linnaeus, Queen Triggerfish Mi-lii lilltyx hispinosits Gilbert, Pacific Black Trigger (X) LIFE HISTORY OF EPIBDELLA MELLENI MACCALLVM 103 Family Monacanthidae Ceratacanthus sclicepfii (\Valbaum), Orange Filefish StephanoUpis his/>idus (Linnaeus), Common Filefish Family Ostraciida> Lactophrys tricornis (Linnaeus), Cowfish Lactophrys trigonus (Linnaeus), Common Trunkfish Lactophrys triqueter (Linnaeus), Smooth Trunkfish Family Tetradontidae Spheroides annidahis (Jenyns) Pacific Puffer (X) Spheroides maculatus (Bloch and Schneider), Common Puffer or North- ern Swell fish Family Diodontidae Diodon hystrix Linnaeus, Porcupine Fish Family Triglidse Prionotus evolans (Linnaeus), Striped Sea Robin Family Malacanthidae Malacanthus plumeri (Bloch), Sandfish The species marked "X" seems to have developed a partial im- munity after a short period of susceptibility. Most of these species are usually present in the tanks of the Aquarium, and newly arrived specimens always show a marked susceptibility to infection. How- ever, after being present in the tanks for several weeks these species seldom show a slight and never a serious infection although they are continually exposed to reinfection. Some of the other susceptible fishes (e.g., Chaetodipterus faber, the spadefish) seemingly retain their infections, continually become reinfected, and die if they do not receive treatment. The central members of the spiny-rayed fishes (Acanthopteri), especially members of the families Serranida? and Lutianidae, are extremely susceptible, and the possibility of the develop- ment of an immunity seems to be more strongly suggested in these families although it is not shown by all members. The other species showing a distinct susceptibility are rather scattered phylogenetically but are all within this order. Epibdella melleni has never (prior to June 1931) been observed on any of the following fishes, all of which have been continually exposed to infection while at the Aquarium: Subclass Elasmobranchii Order Asterospondyli Family Ginglymostomidae Ginslymostoma cirratum (Bonnaterre), Nurse Shark Family Galeida? Mustelns canis (Mitchill), Smooth Dogfish Family Carchariidae Carcharias littoralis (Mitchill), Sand Shark Order Batoidei Family Rajidae Raja eglanteria Lacepede, Clear-nosed Skate 104 THEO. L. JAHN AND L. R. KUHN Family Dasyatida? Dasyatis centrum (Mitchill), Northern Sting Ray Family Aetobat ida- Rhinoptera (jididriloba (l.e Sueur), Cow-nosed Ray Subclass Teleostomi Order Apodr^ I amily M ura-iii>fr Gymnothorax funebris Ranzani, Green Mora> Gymnothorax moringa (Cuvier), Spotted Moray ( )rder I laplomi Family Poeciliidae /•'iiinl it/us liftcnxlitns Cl.imiaeus), Common Killihsh l-'itndulus majalis (Walbaumi, Striped Killifish Order Lophobranchii Family Syngnathiche Hippocampus hitdsonius 1 )e Kay, Northern Sea Morse < >nler Acanthopteri Family Serranida? Mycteroperca bonaci (Poey), Black Grouper Mycteroperca mirrolepis (Goode and Bean), Gag Roct'iis Ihictiliis (Bloch), Striped Bass l;.unil\- I.utianiihr l:.i'<>t>lites viridis f\'alenciennes\ Blue-striped Snap]icr i P.icific) Lutiiimis Arisen* (Linnaeus), Gray Snapper Family ILnemulidae Hamulon sp., Grunts Family Sparid.i- Art'linxur^iis fxnirtn/i'sii CSteindachner), Pacific S.dcm.i Archosargus probatocephalus (Walbaum), Sheepshead 1-amily Sciajnidse xiiiillmriis Laceprile. Spot < romis (Linnaeus), Black or Sea Drum Family Pomacentridae /'onin.ifiitrnx rectifrcenum (Gill), Pacific Beau Gregory l-'amily I )iodont ida- Chilmnyclerns stlrrf>'ii (\\'albaum), Spiny Boxfish 1 aiiiiK Li hciicidse /•'.< lii'in-is nnncrntcs Linnaeus, Shark Sucker I amilv Bat raclioidida- tun ( l.innacii>), Toadtish It is to he noted that although six elasmobranchs, representing six dirTerent families, were exposed, not one was ever observed to be infected. Also, live orders of the teleosts .ire represented on the list of non-susceptible fishes, and only one order, the Acanthopteri, is represented on the list of susceptible fishes. Therefore it seems as if .ill ^usceptible fishes may belong to the order Acanthopteri. Of this older .1 hiv.h HIM ept ibilit y is shoun by members of the families Serr.mid.r, Lutianid.r, and Kphippid.e, and no infections have been obseixed in the families Spaiid.r, Pomocentridae, Echeneida\ and Batrachoididae. Some of the other families are represented by LIFE HISTORY OF EPIBDELLA MELLENI MACCALLUM 105 members on both the susceptible and supposedly non-susceptible lists, and others are represented on the susceptible list only. The injury produced to a susceptible fish is usually quite con- siderable, and if the infection is not treated, death often results. The trematodes attach themselves to the epidermis and to the con- junctiva, and in some cases they have been found in the gill and nasal cavities. Young specimens may be found on almost any part of the epidermis, but in some species (e.g., Chxtodipterus faber and Promicrops itaiara] the adults seem to be concentrated on the eyes. This concentration may be brought about either by the migration of young forms to the eyes, or by a high mortality rate of those which have become attached to the epidermis, or by both of these possible factors. The relative ease with which nourishment could be obtained from the soft tissue of the conjuctiva as compared with the difficulties encountered with the firmer epithelium between the network of scales seems to offer an obvious explanation of why such a concentration might occur. In mild infections the cornea is attacked and sometimes destroyed. If the infection is not treated, destruction of the eye follows, probably due to the effects of both Epibdella and secondary bacterial invaders. In very heavy infections the epidermis may suffer such severe injury as to cause the falling off of scales and the exposure of large areas of connective and muscular tissues with subsequent death of the fish. In one case, that of a Galapagos labroid which had escaped attention for some time, over two thousand adult worms were removed from the entire surface of the body. In the case of another similar infection several thousand eggs were found in the gill and nasal cavities. Both of these severe infections resulted in the death of the fishes. DISCUSSION The life history of Epibdella melleni differs widely from the other definitely-known life histories of the Monogena. The life history of E. melleni seems to offer no homologies to the polyembryonic or paedogenetic condition found in Gyrodactylus (Kathariner, 1904) or to the characteristic fusion of the diporpa larvae of Diplozoon (Zeller, 1872a). The direct development of a ciliated larva into the adult seems to be more closely related to the type of development described for Polystomum integerrimum (Zeller, 1872, 1876). However, the dimorphic development described for this species is probably, as pointed out by previous workers (e.g., Stunkard, 1917), the result of the confusion of the life histories of two different species. It seems probable that the ciliated larva described by Zeller may develop into 106 THEO. L. JAHN AND L. R. KUHX one of the two forms which he describes as the adult. If the develop- ment is similar to that found in Epibdella, it is quite probable that the ciliated larva of Zeller developed into the adult which was found on the gills of the tadpole. Also, if the ciliated larva of Zeller is shown to be the larva of the form in the urinary bladder, it is possible that a comparison of the two cases of direct development may still be drawn. The wide differences between the life histories of Epibdella and dvi-odiiftylus seem to have an important bearing on the taxonomic grouping of the members of the order. Fuhrmann (1928) removed from the suborder Monopisthocotylinea Odhner the three families Protogyrodactylidae, < '.yrodactylidae, and Calceostomida? and created for them a new suborder, Monopisthodiscinea, of which the distin- guishing characteristic was the absence of a definite posterior sucker TEXT FIG. 2. Diagrams showing the possible development of the intra-ovarian seminal receptacle. A. Relationships of seminal receptacle (sr), oviduct (od), and ovary (ov) in some of the other species of Monogena. H. A hyp<>t In 1 1< -d inter- mediate stage. C. The relationship as found in f-'.f>if>ilrll. 1('JS. Trcmatoda. Handbuch der Zoologie. Edited by \\". Kukenthal and T. Krunibarh. Berlin and Leipzig. JAHN, I. L., AND L. K. Ki HN, 1'MO. Tin- Life History of Epibdella melleni M..,-- Callum 1927, a Monogenetic Trematode Parasitic on Marine Fishes il'-tract). Anat. AVr., 47: 3<>5. K \ 1 1! \ki\KR, L., 1904. IVl>er die Entwicklung von Gyrodactylus elegans v. Nordm. Zool. Jalirb., Suppl., 7: 519. M M ( M.i.t \i. G. A., 1('27. A New Ecto])arasit ir Treinatode, l;.;.ili(lc-lla nu-lli-ni, sp. nov. Zoopathologica, 1: 291-300. Sn NKAKD, II. \\'., 1917. Studies on North American Polystomidie, Aspidogastridae, and Paramphistomidae. ///. Hi«I. Monographs, 3: 285. XMIII;, E., 1872. Untersuchungen iiber die Entwicklung und den Ban de> Poly- stomum integerrimum Rud. /eitxchr. f. wiss. Zool., 22: 1. XKI.I.KK, E., 1872i;. Untersuchungen iiber die Entwicklung des IHploxoon para- doxuiii. Zcitxclir. f. iciss. Zool., 22: 168. ZELI.EK, E., 1876. \Yeiterer Beitrag zur Kenntniss dcr Polystomcn. Zeitxchr. f. wiss. ZouL, 27: 238. LIFE HISTORY OF EPIBDELLA MELLKNI MACCALLUM 109 EXPLANATION OF PLATES 1. Camera lucida drawing of the adult of Epibdella melleni showing the prin- cipal organ systems. The vilellaria fill most of the interstitial space throughout the body and are not shown in the figure. Size of specimen is 4 millimeters. 2. Sagittal section through anterior half of adult specimen. 3. Sagittal section through the posterior sucker showing the secondary margin. 4. Free-swimming larva of E. melleni, drawn from living material, shortly after hatching. Size of specimen, 225 n. 5. Larva which had been isolated when ciliated and maintained under a sealed coverslip for forty-five minutes, during which time the ciliated epithelium was shed and the posterior sucker opened. 6. Young individual obtained from an infected fish. Size of specimen, 320 ^i. 7. Diagram of the excretory pattern of free-swimming larva. 8. Most common type of egg of E. melleni. Length of body of egg, 150 //. 9. Somewhat less common type of egg. 10. Rather unusual form of egg which differs from the others in the ratio of the lengths of the edges of the tetrahedron. 11. Abnormal egg, shape probably the result of insufficient shell material in the ootype at time of formation. 12. Filament as cast out when apparently still less shell material is available for formation of the body of the egg. 13. Larval hook. Specimen 10 n in length, a, edge view; b, side view. ABBREVIATIONS ag anterior nervous ganglion as anterior sucker c cirrus cs cirrus sac dc digestive caecum e eye ep excretory pore ev excretory vesicle gs genital sinus Ih larval hooks In longitudinal nerve trunk m mouth mb muscular band through testis ml metraterm od oviduct oes oesophagus ol oral lips ot ootype os oral sucker ov ovary pg prostate gland pog post oral ganglion pr prostatic reservoir sg 'shell gland' sgd 'shell gland' duct sm secondary margin of posterior sucker sr seminal receptacle sv seminal vesicle t testis vd vas deferens yd vitelline duct yr yolk reservoir 110 THEO. L. JAHN AND L. R. KUHN 1'I.ATK 1 as - sm LIFE HISTORY OF EPIBDELLA MELLENI MACCALLUM 111 PLATE 2 dc- -Mh • THE VAPOR TENSION RELATIONS OF FROGS EDWARD F. ADOLPII (From the Physiological I.iilmrntory, the 1'nirersity of Rochester School of Medicine and Dentistry) INTRODUCTION An account of the water relations of frogs would be incomplete without a description of the exchanges of vapor between the animals and environing atmospheres. Some of the same properties of osmotic pressure and permeability that control water exchanges when frogs are surrounded by liquid might be expected to exhibit themselves in the presence of air. But whereas liquid water may serve as a medium for the exchange of other substances, water vapor is lost from the body without being accompanied by any dissolved material. The two chief objects of the observations were: first, to find whether the rate of evaporation from the frog's body is proportional to the relative humidity of the atmosphere, or at least fixes a vapor tension curve for the organism; and second, to find whether a frog can come into equilibrium with a definite vapor tension. In theory ever\ object has a measurable vapor tension; in practice the relation to water vapor is greatly complicated by the thermal properties of the body. It therefore becomes necessary to interpret data on rates of evaporation in terms of heat production, body temperature, and thermal conductivity. Of practical consequence is the finding that under no circumstances can a frog absorb water from the atmosphere. Taken in connection with the fact that the skin offers no unusual obstruction to the loss of uater into the atmosphere, it is evident tli.it with respect to water balance a frog is unsuited to non-aquatic existem e. Ml-/l HODS In the course of the experiments six different procedures were n-ed to establish the- relationship between the frog and the atmosphere. I .K li method \\as in>t i IK i i \ e upon certain points and each was useful to evaluate some factors of water exchange. All the method- depended upon weighing the frog at frequent intervals; in these intervals only water was lost in appreciable amounts. The loss of carbon b\ a 1 1 'o^ of SO -rains \\ei-lil amounts to about 1 ..? milligrams per hour, calculating from the mean rate of carbon dioxide production, 112 EVAPORATION FROM FROGS 113 measured by Smith (1925) on Rana pipiens. The same species was used in all the present measurements of evaporation. The six methods of weighing evaporation losses that were used will be designated by letters. The last one is to be recommended for most general use in the study of atmospheric relations of organisms. A. Frogs were exposed to the air of the room in a wire basket resting on a table. The temperature, relative humidity, and dew-point of the air were read at intervals. With inappreciable air motion and small changes of humidity during any one test, the weight changes could be related to the average humidity for the period. B. A single frog was enclosed in a 400 cc. glass chamber through which condi- tioned air was recirculated. The chamber was suspended from a triple-beam balance, and connected by flexible rubber tubes with wash-bottles containing sulphuric acid mixtures. The air was pumped by raising and lowering a mercury bulb by means of a "wind-shield wiper"; the wash-bottles, in the bottoms of which were layers of mercury, serving as the valves of the pump. The apparatus was run in a room of constant air temperature, but this did not prevent water from condensing in the chamber when high humidities were used. C. A single frog was placed in a screen cage in a 4-liter jar equipped with a fan. The fan was spun very rapidly by a belt and motor outside the jar, the fan shaft piercing a brass top fitted by a groove to the jar. The fan was stopped and the jar was opened each time the frog was to be weighed. In the bottom of the jar was a sulphuric acid mixture which controlled the relative humidity of the air. The fan and the air movement produced considerable heat so that temperature gradients always existed. The steady state was temporarily destroyed at each removal of the frog for weighing. The apparatus was operated in the constant temperature room. D. The fan chamber was used, but the cage containing the frog was suspended from the balance arm by three wires passing through three holes in the chamber lid. The holes were closed by felt washers except when weighings were being made, at which time the fan was also stopped. The chamber was immersed in a regulated water-bath up to the rim, but most of the heat gradients persisted. E. The frog rested on a screen platform in an ordinary desiccator above a sulphu- ric acid mixture. The desiccator was immersed in the water-bath up to the rim and was opened to remove the frog at each weighing. In weighing, the frog was exposed to another atmosphere and to handling, and the results were therefore unreliable at slow rates of drying. The air was quiet except as it was moved by the frog's breath- ing, but the frog remained at a constant distance (5 cm.) from the equilibrating solu- tion. F. A frog was suspended in a jar by a single wire from a balance arm. The jar and top were completely immersed in a water bath. The wire passed through a vertical tube in entering the jar, and the tube was ordinarily closed by a washer on the wire; when a weighing was taken the wire was raised slightly. The weights were reproducible enough so that an analytical balance was used. The cage proved to be a complicating factor because water coated its wires through capillary attraction away from the frog's skin. The most constant results were obtained by pithing the frog and suspending it by fine wires in a horizontal position. This method was accurately checked by exposing solutions of diverse vapor tensions in a glass dish in place of the frog. High humidities were obtained by exposing the air to water or salt solutions of known concentrations. Lower humidities were controlled by keeping the air in contact with mixtures of sulphuric acid and water; their concentrations were estimated from their specific gravities measured with a Westphal balance. The vapor 114 EDWARD F. ADOLPH tensions of all these solutions were obtained from chemical handbooks. Recovery from desiccation was studied in about half of the experi- ments. The frogs were placed in tap water and weighed at frequent intervals in the manner usual for wet frogs (Adolph, 1931). RATES OF EVAPORATION Frogs were gently blotted with a towel before each experiment so that no water would drip from their surfaces. Urine was pressed out of the bladder in the course of handling them, and it is well known that urine formation ceases when water is no longer being taken into the body (Adolph, 1927). Under these circumstances regular changes 12 1 2 3 4 5 Time in hours FIG. 1. Changes of body weight during evaporation (by method 7^) ami recovery' in tap water. The evaporation occurred in an atmosphere saturated with moisture at 20° C. during 25 hours, with slight det rease in rate throughout. For comparison <>| methods, the rate of evaporation by method /•', during 26 hours ex- posure to a saturated atmosphere at 20° ('., is also shown. In each case 7 and II represent successive (\.\\-*, each experiment being continued through the intervening night. of body weight were observed. Temperature adjustments all occurred within the first halt -hour; then weight was lost rapidly for one or two hours, after which the rate of loss was quite constant from hour to hour. This series of events is shown for one experiment in Fig. 1. \Yithin the next 1\ hours, the rate of weight loss in any humidity usually decreased very slowly. This change of evaporation rate was EVAPORATION FROM FROGS 115 as great as 50 per cent when the atmosphere was saturated, but was less than 20 per cent when the humidity was nearly zero. The responses of a single individual upon successive exposure to diverse humidities is shown in Fig. 2. The various experiments were compared by plotting the rates of evaporation against the relative humidities that prevailed. It may be stated that no object is known for whose vapor equilibrium the absolute humidity has significance apart from the relative humidity. This principle follows from the kinetic behavior of gases at uniform temperature. In the experiments it was assumed that the relative humidity was that which would have prevailed if the atmosphere were completely in equilibrium with the equilibrating liquid; this liquid always exposed more surface than the frog. In many experi- ments a hair-hygrometer was placed in the chamber with the frog and this assumption was found to be nearly true. to 03 28 27 CQ 26 F14 96 f. 0 1 7 23456 Time in hours FIG. 2. Changes of body weight during successive exposures (by method F) to three relative humidities. The rates of evaporation may be estimated in a number of ways; by finding the percentage of the original body weight lost in 24 hours, or the average number of grams of water lost per hour in the first six hours, or the grams per square centimeter of body surface lost during the steady state of the third to seventh hours of an experiment. Actually the experimental results were analysed in these several ways, and the last one was adopted, both as giving the most consistent and reproducible data and as being the most rational. All the results by the six methods are shown as averages in Fig. 3. 116 EDWARD F. ADOI.PH Since the rates of evaporation by three different methods fall on curved lines, it is probable that the rates are not exactly proportional to relative humidities. This may be due to some feature of the experi- ment such as the gradient of vapor near the frog's skin and is not necessarily to be ascribed to the supply of moisture on the surface of the body. One important factor is that, as Hall and Root (1930) observed, the body temperature is much lower than the air temperature .1- the humidity declines. A serious attempt was made to relate the rates of evaporation in high vapor tensions to the rates of water exchange by the frog im- 0 o 15 E 0 20 25 30 CO o 35 0 20 40 60 60 100 Relative humidity in percent I'K,. .v Me. 111 rat( "I < \ .i|><>rat ion ;it v.iriuus n-l.ilivr luimiditics as obtaiin.l liy i In- MX methods. mersed in salt solutions of the corresponding tensions. It is well known (l)iirig, 1901; Adolph, 1925) that in a certain range of sodium chloride solutions that are hypotonic when compared with the body fluid^ of tlu- frog, water is gained by the body faster than it is gained in t.ip water, \\.is it possible that the vapor tension of the frog was higher .it (>9.8 per cent relative humidity than at 100 per cent? \Yhen EVAPORATION FROM FROGS 1 1 7 compared (by method C), no significant difference in rate of evapora- tion was found at these two humidities. But this merely meant that the precision of the method was too low. In fact none of the methods was nearly good enough to decide this point, because the evaporation was slow even with high velocities of air motion, and because the production of heat by the frog as well as by the air motion could not be sufficiently corrected for. VAPOR TENSION EQUILIBRIUM Is it possible to establish conditions in which a frog will neither gain nor lose water? The only procedure sufficiently accurate to answer this question was method F. Some 30 experiments were run with saturated atmospheres; the air temperature being constant to ± 0.01° C., and fluctuations of weight due to all causes being reduced to ± 0.5 milligram per hour. The average result was a loss of 4.3 milligrams per hour by the frog, or perhaps 3.0 milligrams if the loss of carbon is allowed for. In 5 of the experiments slight gains of weight were shown and in 6 more experiments no change of weight occurred. But in every one of these 11 tests water was later found on the wire cage, and in all experiments where the wire cage was omitted and the pithed frog merely hung in the chamber some weight was lost. A few experiments where the humidity was reduced to 99.7 and 99.3 per cent showed similar losses of weight from the frog. It is believed that no means could be devised of bringing a living frog into vapor tension equilibrium. The reason for this is, of course, that the organism is producing heat, and that at the frog's surface exists therefore a slight vapor deficit. A frog weighing 30 grams produces 12.5 calories per hour (Smith, 1925), and has a body surface of about 75 sq. cm. (Adolph, 1931) ; hence in a steady state it is losing 0.17 cal. per sq. cm. per hour. The evaporation from the frog of 4.3 milligrams of water per hour is equivalent to an expenditure of latent heat of 0.034 cal. per sq. cm. per hour. In other words, even this rate of evaporation dissipates only one-fifth of the heat that is being basally produced. Moreover, it eliminates between two and three times the amount of water that is being basally produced in the frog's body by oxidation, which is 1.6 milligrams per hour. The loss of 20 per cent of the frog's metabolic heat by evaporation in this atmosphere happens to be similar to the loss of 24 per cent of a man's metabolic heat by evaporation under basal conditions. It is obvious that no vapor equilibrium can be approached more closely than this by the metabolizing organism. 118 EDWARD F. ADOLPH EFFECTS OF PITHING AND OF REMOVAL OF SKIN The best determinations of evaporation rate, as already stated, could be made when the frog was totally quiescent and when the frog could be suspended by fine wires instead of being put into a cage. A few experiments were therefore made to compare the pithed frog with the normal frog. This was best done at high rates of evaporation, because of the smaller importance of absolute errors under such conditions. One experiment is shown in Fig. 4, and it is evident, as was true in other similar experiments, that no consistent difference existed in rates of evaporation between the pithed and the normal frog. A similar conclusion was reported by Hug (1927). 18 17 16 .£ 15 •r CO 13 12 H tK "K F30 8 10 12 Time in hours Fi<;. 4. ( h,m-i s uf II(K|V weight during successive exposures of the same frog under t lin-r mm lit ions to a rd.it \\ c humidity of 50 per cent. /, norm.il; //, pithed; ///, skinlrsx Method F. After c;ich experiment it was ascertained that the circulation of the Mood persisted in the pithed individual. Two tests in which the eirnilation was completely stopped showed no detectable difference from the normal frog, and it is likely that the circulation is not a limiting factor in the rate of evaporation. It was concluded by Hug L927) that dead frogs evaporated at the same rates as living ones. EVAPORATION FROM FROGS 119 During and after the evaporation tests it was noted that the appearance and feel of dryness in the skin was highly variable. But it proved impossible to correlate this condition with the rate or the amount of desiccation suffered. In a short series of experiments the entire skins were removed from pithed frogs. It seemed possible that the external surface of the skin, being in equilibrium with fresh water, would naturally have the vapor tension of pure water, while the deeper tissues would have vapor tensions similar to that of a Ringer's solution, which corresponds to 99.7 per cent relative humidity. It was found that in low humidities the rate of evaporation of a skinless frog was little different from the rate of a normal frog, as Fig. 4 shows. This is in marked contrast to the protection against evaporation furnished by the skins of reptiles (Gray, 1928). In saturated atmospheres also no difference of rates could be measured. Whereas normal frogs lost 4.3 milligrams per hour, six skinless pithed frogs lost on the average 3.6 milligrams per hour, which is a much better agreement than could be expected. Obviously the heat production of the frog is sufficient to prevent water from condensing on the superficial tissues even though its vapor tension be slightly lower than the tension of pure water. In a number of experiments frogs were first desiccated by 15 to 35 per cent of their body weights and then placed in saturated atmos- pheres. In no case was there a significant gain of weight; on the average the rate of loss was the same as for a normal frog. Even when the desiccated frogs were pithed and skinned no gains of weight were found. Evidently the vapor tension of the body cannot by this means be lowered sufficiently to overcome the vaporization due to dissipation of metabolic heat. RATE OF REGAIN OF WATER AFTER DRYING But when put into water, a desiccated frog regains fluid at a rapid rate. The course of this regain is shown for one experiment in Fig. 1. The rate is fairly uniform for the first hour or two hours, though some gradual diminution in rate occurs. After the original weight of the frog has been attained, the gain ceases quite sharply. The average initial rate of gain (38 experiments) was 0.8 gram per hour or 11 milligrams per square centimeter of body surface per hour. This is more rapid than the fastest desiccation in still air can be accomplished. The rate of regain is not correlated with the amount of desiccation, provided at least 5 per cent of the body weight had been lost, nor with the velocity of the desiccation. 120 EDWARD F. ADOLPH Partial contact of the body with moisture is sufficient to supply water for regain (Stirling, 1S77; Dung, 1901). If the dried-out frog is merely placed on a damp towel, water will be imbibed through the skin at the average rate. So far as is known to investigators generally, frogs never inge>t water through the mouth when immersed in it. It is of interest that the rate of respiratory metabolism increases with moderate desiccation of the frog and decreases markedly with extreme desiccation (Caldwell, 1925). HEAT EXCHANGES It has been demonstrated that the exchanges of water between frog and atmosphere do not correspond to an ordinary vapor equi- librium. The explanation is found in the continual production of heat in the body. In an atmosphere saturated with moisture at 100 Relative humidity in percent FIG. 5. Jhmiiciily .nxl tciiipiT.it HIT n-l.it ions at 20n ('. (), dry-bulb temporal urr <.f tin- ,iir; T, rectal t.-mprr.ii urr of R,ina t>i(>irns. shows that a frog resembles a wet-bulb + 6 o X CJ b Q- _• c^3 CJ -2 03 d CJ -6 98 100 H V o 20 40 60 80 100 Relative humidity in percent FIG. 6. Partition of virtual heat exchanges by frogs at 20° C. R, gain of heat by conduction, convection, and radiation from the surroundings; H, gain of heat by oxidative production in the frogs; V, loss of heat by evaporation of water from the frog's surface. The inset at the top is a ten-fold enlargement of the right-hand edge of the graph. thermometer very closely in high humidities. In low humidities the effect of convection in slinging the psychrometer is more pronounced. Hall and Root (1930) had only slight air movement when they meas- ured the rectal temperatures of the frogs. 122 EDWARD F. ADOLPH Evaporation as a complication of measurements of heat production has been discussed, at least for isolated tissues (Fischer, 1927; Hill, 1930). Heat production as a complication of measurements of evaporation deserves equal recognition. From the rates of evaporation into air of diverse humidities it is now possible to estimate the proportions and total amounts of heat di-si] >ated by vaporization on the one hand, and conduction, convection ,ind radiation on the other hand. It is assumed for this purpose that all the water evaporated from the body gained its latent heat from the frog; this is very nearly true because the specific heat of the body is much higher than that of the air surrounding it. The view is equally sound that the frog really receives no heat from the surround- ings while in the steady state, but merely acts as a converter of kinetic heat into latent heat within the atmosphere. The calculations of heat production are facilitated by the data of Hall and Root (1930) on the body temperatures of frogs in various humidities (Fig. 5), and by the numerous data, as those of Vernon (1897) and Krogh (1914), on the relative rates of respiratory metabo- lism at various temperatures. The partition of heat losses from the frog is indicated in Fig. 6, the rates of loss by evaporation being calculated from the measure- ments by method F. It will be seen that under nearly all atmospheric conditions evaporation alone removes heat much faster than com- bustion generates it. Ordinarily, therefore, the frog is virtually taking up heat from the surroundings by conduction, convection and radia- tion. It has been ascertained in the present experiments that in 100 per cent humidity, however, the evaporation accounts on the average for only one-fifth of the heat loss. Since the cooling is all produced by evaporation in proportion to the difference of vapor tensions between the frog's surface and the air, it is easily understood why there is no condensation of water on the surface of the frog even though it is much cooler than the atmosphere. The temperature of the frog never decreases to the dew-point of the atmosphere that surrounds it. It is also possible to calculate roughly a coefficient of heat flow for the combined virtual losses by conduction, < i m\ ection and radia- tion, excluding evaporation, from the data of Fig. 6. The heat dissipation < II') i-- proportional to the bod\ sin lace (5) to the time (/), and to the temperature difference (0). < )r W---- kSlO. The b(-> t value of /;, for the range of low humidities where k is actually EVAPORATION FROM FROGS 123 constant, calculated from the slopes of line R in Fig. 6 and line T in Fig. 5, is 1.0 cal. per square centimeter per hour per degree centigrade. It is more than possible that the curves present in the line for body temperatures (Fig. 5) and in the line for evaporation rates (Fig. 6) are significant, in which case k is modified at diverse high relative humidities, and the thermal properties of the frog's body differ at various humidities or body temperatures. Such differences might be due to vasomotor shifts or other physiological responses. The amount of this heat flow that is due to the single factor of radiation can be calculated. It is assumed that the frog has maximal radiation (as for an ideal black body) such as is believed to hold true for human skin (Cobet, 1924), and that 70 per cent of its surface is exposed to radiation. It is then found that this form of heat transfer might account for half of the combined heat flow (R) at low humidities and for all of it at humidities above 70 per cent. Further, it can be calculated from the constant for heat conduction through air that a still atmosphere would be unable to conduct much of the other half of the heat from the water of the bath to the suspended frog. Hence convection currents set up by the frog's breathing and by temperature differences near the body must be important in bringing heat to the animal. COMMENT The failure of frogs to absorb water from moist atmospheres means that these animals cannot survive long away from liquids. While the habits of frogs are such as usually to keep them in or near water, toads are ordinarily regarded as terrestrial. Toads, when subjected to similar vapor tensions, likewise showed no ability to absorb vapor from a saturated atmosphere. Their survival away from water evidently depends upon their taking up water while in contact with wet objects; it has been seen that mere moisture held in towels can supply this. Soil is a sufficient natural source of supply. So far as is now known, a toad has no properties fitting it for water conserva- tion or accretion that are not possessed by most aquatic animals. The amounts of desiccation endured by frogs have always been matters for remark ever since the first observations were made by Edwards (1824), Chossat (1843) and Kunde (1857). Various investi- gators have attempted to find how much loss of water is consistent with subsequent recovery; loss of roughly 40 per cent of the body weight, which is 50 per cent of absolute water content, allows of survival (Snyder, 1908; Hall, 1922; Smith and Jackson, 1931). Studies have been made of the relative losses by the various organs and tissues of the frog's body during desiccation (Durig, 1901 ; 124 EDWARD F. ADOLPH Ueki, 1924; lizuka, 1926; Smith and Jackson, 1931). At present little relation can be deciphered between the partition of water losses and the water economy of the body as a whole. It is possible that the marked loss of water by the skin helps to diminish the rate of subse- quent (.'v.ipi'ration to the small extent found above. SUMMARY 1. Frogs lose water by evaporation at rates that are nearly in- versely proportional to the relative humidities of atmospheres. 2. Functioning of the central nervous system, of the blood's circulation, and of the skin made no significant differences in rates of evaporation. 3. In saturated atmospheres evaporation still goes on, which is explained by the fact that the production of heat keeps the body slightly warmer than the atmosphere. 4. In unsaturated atmospheres heat may be regarded as being lost by evaporation until the lowered temperature of the body comes into a steady state with the gain of heat by conduction, convection, and radiation from the surroundings. 5. No equilibrium of zero evaporation can be established for the living frog, and so the vapor tension of the frog's surface cannot be measured. HIHLIOGRAPHY ADOLPH, E. F., 1925. The Regulation of Body Volume in Fresh-water Organisms- Jour. Exfx-r. Zool., 43: 105. AUDI. I'll. I-'.. I-".. 1027. The Excretion of Water by the Kidneys of Fi'o-s. Am. Jour. riiysiol.. 81: 315. AnoLPH, E. F., 1931. The Water Exchanges of Fnnjs With and Without Skin. Am. Jour, rhysidl., 96: 569. CALDWELL, < .. I.. l'»25. A Reconnaissance of the Relation bet \\eeii Desiccation ami (ail >on Dioxide Production in Animals. H/'ol. /in!/., 48: 259. CHOSSAT, C., 1843. Recherehes experiment, lies sur 1'inanit ion. Mi' in. acad. roy. sci. savinils t'tn i>/\\, Paris, 8: 202 pp. COKET, R., r\D F. P>K.\\iii,K, 1924. I her Messun- der Warmest rahlung der menscli- lichen II. ml mid ihre klini^i In- P.edcutung. Pent. Arch. klin. Alcd., 144: 45. DIKK,, A., 1901. Wasseri;ehalt mid Organfunct ion. I. Arcli.f. d. gcs. Physiol., 85: 401. FIA\ \ku-~, \\'.. 1S24. I >e I'inlluenee des ;it;ents physi(ine> Mir la \'ie. Paris. I [S( III.K. F., 1927. Ober thermoelektrische Me-.Mmi;cn am I ler/en. Arch. f. d. ges. Physiol., 216: 123. 1 .K\Y, J., 1(>2S. Tin- Role of U'.iter in the Evolution of the Terrestrial Vertebrates. Hrit. Jour, l-'.xficr. Hint., 6: 2(>. MALI., 1 . (... 1('22. '1 he Vital Limit of Exsiccation of Certain Animals. Biol. Bull., 42: 31. ll.u.i,, !•'. (, , \\n R. U'. Root, 1930. The Inlliieiice of Humidity on the Body Tempei.ii in i- 'it ( ertain Poikilotherms. Biol. null., 58: 52. HILL, A. V., VND P. S. KUPALOV, 1''30. The Vapour Pressure of Muscle. Proc.Roy. . Ser. M, 106: 445. EVAPORATION FROM FROGS 125 HUG, E., 1927. L'evaporazione dell'ucqua attraverso la cute dclla raua in varie condizioni d'ambiente. Arch. sci. biol., 10: 322. IIZUKA, N., 1926. Recherches sui la dcshydratation de la grenouille et son retentisse- nient sur les crhanges respiratoires. Ann. physiol. phvs.-chim. biol., 2: 310. KROGH, A., 1914. The Quantitative Relation between Temperature and Standard Metabolism in Animals, lntcrn.it. Zeitschr. f. phys.-chem. Biol., 1: 491. KUNDE, F., 1857. Ueber Wasserentziehung und Bildung voriibergehender Kata- rakte. Zeitschr. v/.v.v. Zoo/., 8: 466. SMITH, 11. M., 1925. Cell Size and Metabolic Activity in Amphibia. Biol. Bull., 48:347. SMITH, V. D. E., AND C. M. JACKSON, 1931. The Changes during Desiccation and Rehydration in the Body and Organs of the Leopard Frog (Rana pipiens). Biol. Bull., 60: 80. SNYDER, C. D., 1908. Der Temperaturkoeffizient der Resorption bei tierischen Membranen. (Vorlaufige Mitteilung.) Zentralbl. Physiol., 22: 236. STIRLING, W., 1877. On the Extent to Which Absorption Can Take Place through the Skin of the Frog. Jour. Anal. Physiol., 11: 529. UEKI, R., 1924. Uber den Wassergehalt der Organe trocken gehaltener Frosche. Arch. f. d. ges. Physiol, 205: 246. VERNON, H. M., 1897. The Relation of the Respiratory Exchange of Cold-blooded Animals to Temperature. Part II. Jour. Physiol., 21: 443. 10 A < < >M|'. \RISON OF THE PLASTID WITH THE GOLGI ZONE T. ELLIOT WEIER1 There has always been a desire on the p;trt < >l c\ tologists to homolo- the structural elements of plant and animal cytoplasm. Of late much attention has been given to finding an element in the plant cell which might be comparable with the Golgi zone of the animal cell. Guilliermond (1929), because of the network structure assumed under certain conditions by the plant vacuoles, believed that these might be compared with the Golgi body. This idea found support in the work of Parat (1928), which seemed to show that the Golgi apparatus was mainly vacuolar in nature. Hall (1931), in a recent work on Euglena, maintains that it is not the sap vacuole which is comparable to the Golgi body but certain small granules staining with neutral red. Bowen (1920, 1922a, b, c), in an extensive study of insect spermatogenesis, has given a very definite criterion for recog- nizing the Golgi zone. It is that differentiated mass of cytoplasm which during spermatogenesis gives rise to the acrosome. Working with this in mind, he (1927) undertook a study of spermatogenesis in the moss, Polytriclnim jitniperinmn. Unfortunately this investigation was never completed, but Bowen's belief was that the limosphere, a body which because of its development as described by Allen (1917) was thought by Wilson (1925) to be comparable to the acroblast, was elaborated by the osmiophilic platelets. Bowen's results, though differing slightly from the known formation of tin- acrosome in insects, made it seem possible that the osmiophilic platelets might be compared with the Golgi body. The problem was attacked independently and in the same manner by Weier (1931, 1932) who, however, arrived at conclusions quite different from those of Bowen. During sporogenesis and spermato- genesis in I'olytrichum commune reactions of the plastid to osmium .UK! silver techniques and .1 strikingly comparable structure and behavior of the plastid to that presented by the Golgi body during comparable periods of development led this author to believe that the plastid and the Golgi zone had certain characteristics in common. 1 National Ruse-arch Fellow in the- Biological Sciences 1930-1931, Department of I 'mi any, < <>ni<-ll I Hi\ rrsity. 126 COMPARISON OF PLASTID AND GOLGI ZONE 127 This idea was not new, Bowen having suggested it in 1926 to withdraw it later in favor of the osmiophilic platelet hypothesis. A study of spermatogenesis in Polytrichum commune and Catha- rinaea undulata furnished clear evidence that the limosphere is a derivative of the plastid. Indeed the history of plastid development during spermatogenesis is, in all important details, strikingly similar to the history of the development of the Golgi body during spermato- genesis in insects. When one adds to this the similarity in structure of the living Golgi zone (Morelle, 1926) with the living plastid (Zirkle, 1926), the similarity in reactions of both structures to certain fixing fluids (Morelle, 1926 and Weier, 1932), the similarity of function (the elaboration of specialized products for use by the organism), the similarity of relation to mitochondria in certain cells (Pensa, 1925; Parat, 1928), and the similarity in relation to chemically allied pigments (chlorophyll and hemoglobin — see Weier, 1932), it seems that there must be some significant relation existing between the plastid and the Golgi zone. Furthermore, it has been shown (Weier, 1932) that the previously supposed homology of mitochondria with plastid and of Golgi zone with vacuoles was not sound. It is well known that osmium and silver impregnation techniques give preparations with a great variety of fixation pictures. Parat in his "vacuome" hypothesis has variously interpreted these to be "vacuoles," "active mitochondria," and "diffuse lipoids." During the course of the sporogenesis and spermatogenesis studies on Poly- trichum commune and Cathannaea fixation pictures surprisingly similar to those interpreted by Parat as " vacuole" and "active mitochondria" were encountered in the plastid. This paper is a record of these fixation pictures. It is believed that the similarity found between Parat's "zone de Golgi" and the plastid is significant in two ways. Firstly, it further establishes the theory that the plastid is comparable to the region of cytoplasm which the zoologists have generally called the Golgi apparatus. Secondly, it lends strong support to the interpre- tation of the Golgi body expressed in the work of Bowen, Nassonov, Gatenby, Hirschler, Beams, Morelle and others, namely, that the Golgi body or zone is composed of specialized living cytoplasm and not of a liquid vacuolar sap of a non-living nature as Parat holds. Figure 1, A and B, is redrawn from a portion of Parat's (1928) Fig. IX, a and b. Both A and B represent, according to Parat, different fixation pictures of the "systeme vacuolaire" of the salivary gland of the Chironomus larva. In A fixation has been excellent; the "vacuoles" have impregnated evenly, presenting the aspect which one observes them to have in the living condition. In B, however, 128 T. ELLIOT WEIER 'Telement apparait incontestablement retracte. . . . Les aspects les plus divers de- \rsicules chromophobes a parois chromophiles sont observables. ..." The plastid in the living cell of Polylrichum is formed of specialized cytoplasm sharply delimited from the remainder of the cell and, if it is a chloroplast, in some manner impregnated with chlorophyll. This B FIG. l. A and B. Two fragments of the salivary inland of the Cliin»n>»ins lar\ra after fixation according i» I >a I'.mo. A. Cytoplasm insulin lent Iv fixed: the impregnated vacuoles r«seml>le dictyo- somes. After I 'a rat. B. Cytoplasm well fixed: the impregnated vacuoles appear as they do in the living cell. After 1'arat. C and D. Two aspects of plastid fixation in I'olytricliuni commune. ('. Cytoplasm \\ell fixed: ihe plastid presents a homogeneous structure similar to that observed in living cells. D. Cytoplasm poorly fixed: the structure of the plastid has changed so that it may appear similar to diciyos<>st-chromatization of 72 hours. The plastid appears as a densely staining m.i^ containing "vacuoles" in which polarized light demonstrates the presence of starch. depositories for the secretory products of the specialized mass of <•> loplasm. There is a difference, however, between Parat's description of the C.ol^i /one and the plastid: the presence of "active mitochondria" in the ( ,ol^i /one. According to Parat, a post-chromatization of six COMPARISON OF PLASTID AND GOLGI ZONE 131 hours is not sufficient mordanting to make possible differential staining of the "active mitochondria." They do, however, show after treat- ment with osmium and silver. Parat describes these "active mito- chondria" as being long filaments. Since they appear only after prolonged fixation and in both longitudinal and oblique sections (Figs. Ill a, IV and V, Marg. and M. Parat, 1930) or in longitudinal and cross sections (Figs. X, XVII b, c, f, g, Parat, 1928) as filaments s y ' // r/i , ^ j' ' * B FIG. 3. A. Silver impregnation of an intestinal cell of Triton marbre. Active mitochon- dria ramify through the diffuse lipoids of the Golgi zone. The vacuoles are frequently in contact with the active mitochondria. After Marg. and M. Parat. B. Plastid of Polytrichum commune after osmium impregnation. Structures similar to the active mitochondria ramify through a denser region of cytoplasm and frequently border vacuole-like regions. and seldom as granules, is it not possible that these "active mito- chondria" are in reality distinct plates such as Morelle and Bowen have described in the Golgi zone? The plastid in Zea mays is evidently more resistant to the action of fixing fluids, for even after seventy-two hours post-chromatization it does not show any evidence of the platework structure which fixation usually imparts to the plastid as well as to the Golgi zone. 132 T. ELLIOT WEIER Figure 3, A and B, furnishes another striking comparison between plastid and Golgi zone. In this case the example of Golgi zone structure is taken from Hate XIV, d of Marg. and M. Parat, 1930. It represents a silver impregnation of an intestinal cell of Triton marbre. Figure 3, B, is an osmium impregnation of moss plastids. FIG. 4. A and B. Kpididymis colls of the mouse; osmium impn-n.it ion. .1. Transvi ection. Perivacuolar chondriome impregnated. After Parat. H. llc-.i\y blackened ribbons, formed by the .literal ion of the perivacuolar ' hondriome, surround iln- \.imole, which is deformed but not impregnated. After Parat. C and D. Types of osmic acid impregnation similar to A and B may be found to occur in the plastid of I'olylrichum commune. It will lie noticed that in each case the "active mitochondria" lie parallel to r.ich other and are most certainly in the case of the d and in all pi < >\ >.il >ilit y in Triton mnrbre delicate irregular plates. The-r ramify tlnou-h a region of "diffuse lipoids" or specialized cxtoplasm which has some power of reducing the osmium, thus COMPARISON OF PLASTID AND GOLGI ZONE 133 appearing faintly gray. The "vacuoles" of the Golgi zone are clear and round, while the "cavities" of the plastid are longer and not so sharply and clearly delimited from the darker cytoplasm. In Figs. 2 and 3 we have considered what apparently are good fix- ations of the " active chondriome " and " vacuome." However, after certain techniques, particularly osmium impregnation, this "active chondriome" may be greatly altered. Figure 4, A (Parat, 1928, Fig. XXVIII, a) is a cross section of an epididymis cell of the mouse. Here the "chondriome " is well preserved. By making a comparison with the longitudinal section shown in Fig. 3, A it may be seen that the "chondri- osomes ' ' are most frequently much more ribbon-shaped or plate-like than filamentous. They border on the " vacuoles ' ' which have not blackened with the osmium. In Fig. 4, B (Parat, 1928, Fig. XXVIII, &) the "ac- tive chondriome " has been altered so as to form coarse black plates bor- dering the "vacuoles" which have not impregnated. Under a similar treatment with osmium the archesporial plastid of Polytrichum commune shows a corresponding reaction. In Fig. 4, C an "active chondriome" or filamentous strands of what appears to be plastid cytoplasm border upon a non-blackened portion of the plastid. In Fig. 4, D heavy black plates are in contact with the region of the plastid cytoplasm which reduced no osmium. Both plastid and Golgi zone, then, may under certain conditions precipitate osmium in the form of more or less heavy filaments or plates. These filaments or plates border on a non-reducing mass of plastid or Golgi substance. A comparison of the living plastid with this fixation picture shows that the reduction of the osmium is mainly confined to the borders of the plastonema of the archesporial plastid and that the plastosome, which is certainly of cytoplasmic nature, does not reduce the osmium. It appears lighter than the surrounding cytoplasm, or as a "vacuole." This comparison shows further that the structure of the plastid is not faithfully preserved in that the reduction of the osmium occurs most frequently at the interfaces of the plastonema, the plastosome and the surrounding cytoplasm. This suggests that the blackened regions, instead of marking formed cytoplasmic bodies, may rather indicate interfaces between certain specialized cytoplasmic regions where great cellular activity is taking place. The difference in amount of osmium reduced would depend upon either the length of time the tissue was treated with osmic acid or upon the amount of metabolic activity going on within the cell at the time of fixation. Since it seems that in the plastid this type of impregnation is an artifact in the sense that it does not preserve the exact details of the physical structure of the living cell, may it be 134 T. ELLIOT WEIER suggested that perhaps a similar phenomenon has taken place in the Golgi zone? In other words, the "active mitochondria" are not mitochondria, permanent cellular elements, but mark instead, regions of intense cellular activity. This activity, as would be expected, is taking place at interfaces between differentiated regions of the cyto- FIG. 5. .1, B, and C. Metallic impregnation of the pan< n-as cell of Scyllium catiilns. A and B. The vacuome alone has impregnated. After Parat. C. In this case only the active chondriomc has impregnated. After Parat. D, E, and F. Osmic acid impregnation <>l the androgonal plastid of Poly- Irichum commune. The same types of impregnai inn are encountered here as in A, B, and C. plasm, either between the Ciolgi /.one and the surrounding cytoplasm or between differentiated regions within the zone. It is, of course, known that mitochondria do invade the Golgi re- ion (Morelle, 1926, Parat, 1928) but the "active mitochondria" are >o different from the ordinary mitochondria in structure and to fixation reagents and so similar to the artifacts due to COMPARISON OF PLASTID AND GOLGI ZONE 135 fixation and osmification within the Golgi zone (Morelle) and within the plastid that one cannot avoid questioning the interpretation given to these structures by Parat. In Fig. 5,A,B and C three different aspects of metallic impregnation of pancreatic cells have been reproduced from Parat's (1928) Fig. XXVI. In A and B only the "vacuome" has impregnated, while in C the "active mitochondria" alone reduced the osmium. The criterion for separating these two elements lies merely in the length and the thickness of the bodies which are able to be oxidized by the osmium. Corresponding types of osmium impregnation are frequently en- countered in the plastids of androgones within the same antheridium (Fig. 5, D, E and F). In the case of the plastid, one is dealing with FIG. 6. A. Nervous cell of the tadpole of Bnfo vulgaris. Silver impregnation. Vacuome is impregnated as well as a portion of the active chondriome. In the satellite cell the vacuome is impregnated and exhibits a definite polarity. After Parat. B. Androgonal cell of Polytrichum commune. Osmic acid impregnation. Vacuole-like regions impregnate within the plastid. neither "vacuoles" nor "active mitochondria" but probably with especially active regions of plastid cytoplasm. Even the Golgi bodies of nerve cells are not without a fixation picture which may be compared to the plastid. Silver treatment on this type of tissue in Bnfo vulgaris results in an impregnation of the "vacuoles" and "active mitochondria" embedded in a mass of cytoplasm which is more dense than that of the remainder of the cell (Fig. 6, A, redrawn from Parat, 1928, Fig. XIV, b). At the lower left-hand corner of the large cell is a small satellite cell with its "vacuome" impregnated and showing a definite polarity strikingly similar to that shown by the plastids when there are but two per cell (Fig. 6, B). 136 T. ELLIOT WEIER The plastid of Polytrichum commune (Fig. 6, B) after treatment with osmic acid not infrequently presents an image quite similar to that assumed by the < iolgi /one in the nerve cell of Fig. 6, A. Many small regions reduce the osmium intensively so that the plastid appears to have a very similar structure to the Golgi zone of Bufo vulgaris. However, the regions in the plastid which reduce the osmium so strongly are apparently not the "cavities" which previously have been compared to Parat's "vacuoles" and within which the starch first appears. They probably represent regions of cellular activity rather than any definite morphological cellular element. The two plastids possess a definite polarity similar to that shown by the "vacuoles" in the satellite cell of Fig. 6, A. May it not be that the "vacuoles" too are regions of active cytoplasm rather than B FIG. 7. A. Salivary glaml of the Chironomus larva. Vital. The vacuoles appear to be in the form of a network. After Parat. B. Proplastids in microsporocyte of Zea mays. Benda. The proplastids are reticular. cellular elements perpetuating themselves only by division as the nucleus doe-.' Parat's classical example of vacuole formation is the salivary gland of the Chirononnts larva (Parat, 1(>2S, Fig. Y, 1 to 8). In the a< i mint of this development there are present in all eight stages small discrete "vamoles." At stage six, ho\\e\er, some of these flow together, forming the secretion. It seems th.it Parat may have confused the secretorj products with the discrete Golgi elements, the region of cytoplasm which elaborated the products. At any rate young plastids frequently may be observed to have network aspect very similar to Parat's "vacuoles" found in the salivary gland of the Chirouomns larva. I'ensa has reported young pla-tid> of this same structure as occurring in the cells of the young COMPARISON OF PLASTID AND GOLGI ZONE 137 ovary of Tulipa gessneria. In Fig. 7, B young plastids found in Zea mays microsporocytes are seen to compare very favorably with the "vacuoles" of the salivary gland of the Chironomns larva (Fig. 7, A). During spermatogenesis the Golgi zone is formed, according to Parat, of "lepidosomes" or specialized mitochondria, "rhagiocrine and plasmocrine vacuoles," and " diffuse lipoids." The pro-acrosomic granules are associated with the ' ' rhagiocrine vacuoles. ' ' Here the ' ' le- pidosomes," rather than the vacuoles, as in yolk formation, take on the form of the dictyosome. If we compare the dictyosome-like "lepido- some" (Fig. 8, A} with the archesporial and androgonal plastids (Fig. 8, B), we find no appreciable difference in structure; both are formed of a convex chromophilic border and an internal chromophobic substance. FIG. 8. A. Idiosome from the spermatocyte of Helix pomatia. Osmic acid impregnation. The lepidosomes with convex chromophilic and internal chromophobic portions re- semble the classical dictyosomes. After Parat. B. Archesporial plastid from Polytrichttm commune. Benda or osmium fixation. The plastid closely resembles the dictyosomes. This might be taken as an indication that the plastid, too, is a specialized sort of chondriosome comparable to the "lepidosome." However, in view of the previous comparison between the plastid and the Golgi zone and the part played by the plastid in the elaboration of the limosphere, this is hardly possible. The role of the plastid in the latter respect is identical with that of the classical Golgi body, it elaborates the apical body or acrosome. Parat claims that the "lepidisomes" do not enter into the elaboration of the acrosome. He, however, stresses the fact that they are in an intimate relation with the Golgi zone up until the deposition of the acrosome. Exactly at this time they move away from the zone and pass down the tail of the maturing sperm. It seems that Parat may have missed one of the most important steps in spermatogenesis, the elaboration of the acrosome by the acroblast. 138 T. ELLIOT WEIER It does not seem possible that the "lepidosome," which is so strikingly similar in behavior and structure to the dictyosomes and androgonal plastid, does not play some role in the formation of the acrosome. Nor does it seem possible that the excellent preparations of the late Professor Bowen which the present author has had the privilege of examining can all be either special cases or all distorted to show a relationship between acroblast and acrosome which does not exist in life. It seems that the similarity shown by the plastid to the Golgi /i me cannot be without some significance. It must mean that these two regions of the cell are comparable, possibly even homologous, although the great evolutionary distance between the mosses and animal forms under consideration makes such a suggestion extremely hypothetical. Since we know that the plastid is essentially cytoplasniic, that it possesses a few "cavities" (probably also cytoplasmic in nature) in which the starch is deposited, that it gives rise upon fixation to a peculiar platework picture, in all likelihood marking particularly active regions within the plastid, and that it gives rise to the apical body in spermatogenesis, we cannot but question Parat's interpretation of the plate-like "active mitochondria," the secretory "vacuoles" and the "diffuse lipoids" of the "zone de Golgi." If the similarity shown by the plastid to the Golgi zone in structure, function and behavior may be used as evidence for the structure of the Golgi zone, it lends strong support to the interpretation placed upon that cellular element by Morelle, namely, that it is a mass of specialized cytoplasm containing a number of "vacuoles," and that after prolonged treatment with fixing agents a coarse, heavily-staining network appears within the speciali/ed cytoplasmic mass. I wish to acknowledge my indebtedness to Professor L. \V. Sharp for the interest which he has shown in the progress of this work, and for his stimulating and helpful criticisms of the many perplexing questions which have arisen in connection with the study; and to Pro- fessor E. B. Wilson for his kindness in loaning me some of the prepar- ations of the late Professor Robert Bowen. UTKk.YlTKK CITED ALLEN, C. K., 1('17. The Spermatogenesis of Polytrichum juniperinum. Ann.Bot., 31: 269. BOWEN, K. II. ,1020. Studies on Insect Spennatogenesis. Biol. Bull., 39: 316. BOWEN, K. II., l(>22a. On Certain Features of Spermatogenesis in Amphibia and .1 ni. Jour. Anal., 30: 1. COMPARISON OF PLASTID AND GOLGI ZONE 139 BOWEX, l\. II., 1922/;. Studies on Insect Spermatogenesis. II. The components of the spermatic! and their r61e in the formation of the sperm in I Icmiptera. Jour. Morpli. and Physiol., 37: 79. BOWEN, R. II., l°22r. Studies on Insect Spermatogenesis. V. On the formation of the sperm in Lepidoptera. Quart. Jour. Micros. Sci., 66: 595. BOWEN, R. H., 1926. The Golgi Apparatus — its Structure and Functional Sig- nificance. Anat. Rec., 32: 151. BOWEN, R. II., 1927. A Preliminary Report on the Structural Elements of the Cytoplasm in Plant Cells. Biol. Bull., 53: 179. GuiLLIERMOND, A., 1929. The Recent Development of Our Idea of the Vacuome of Plant Cells. Am. Jour. Bot., 16: 1. HALL, R. P., 1931. Cytoplasmic Inclusions of Menoidium and Euglena, with Special Reference to the Vacuome and Golgi Apparatus of Euglonoid Flagellates. Ann. de Protist., 3: 57. MORELLE, J., 1926. Les constituants du cytoplasme dans les pancreas et leur in- tervention dans le phenomene du secretion. La Cellule, 37: 77. PARAT, M., 1928. Contribution a 1'etucle morphologique et physiologique du cyto- plasme; chondriome; vacuome (appareil de Golgi); enclaves; etc. Arch. d'Anat. Micros., 24: 74. PARAT, MARG. ET M., 1930. Essai d'analyse histochimique et morphologique de la zone de Golgi. (Cellule de la gland pelvienne du Trition, cellules intestinales de Trition et d'Axolotl.) Arch. d'Anat. Micros., 26: 447. PENSA, A., 1925. Les questions les plus discutees sur le cytoplasm des vegetaux. C. rend. Assoc. Anat. 20e reunion. Turin. WEIER, T. E., 1931. A Study of the Moss Plastid after Fixation by Mitochondria!, Osmium and Silver Techniques. I. The plastid during sporogenesis in Polytrichum commune. La Cellule, 40: 261. WEIER, T. E., 1932. A Study of the Moss Plastid after Fixation by Mitochondrial, Osmium and Silver Techniques. II. The plastid during Spermatogenesis in Polytrichum commune and Catharinaea undulata. La Cellule, 41: 51. WILSON, E. B., 1925. The Cell in Development and Heredity. New York. ZIRKLE, C., 1926. The Structure of the Chloroplast in Certain Higher Plants. Am. Jour. Bot., 13: 301. Vol. LXII, No. 2 April, 1932 THE BIOLOGICAL BULLETIN PUBLISHED BY THE MARINE BIOLOGICAL LABORATORY PHYSICAL AND CHEMICAL CONSTANTS OF THE EGG OF THE SEA URCHIN, ARBACIA PUNCTULATA E. NEWTON HARVEY PRINCETON UNIVERSITY For purposes of calculation it would frequently be convenient to have readily accessible data regarding the physical characteristics of much -used animal and plant cells. The unfertilized egg of the sea urchin, Arbacia punctilio ta, is a cell in a relatively stable condition, which has been studied perhaps more widely than any other marine material. Accordingly, I have collected the available data regarding this egg, with the hope that others will do a similar thing for different cells and organisms. A table of physiological constants for man should prove most valuable. The figures for the unfertilized sea urchin egg are given in metric units for the temperature indicated (mostly "room" temperature, about 23° C.) with temperature coefficients where these are known. Although the cm. -gram-second units are common in physics, the mi- cron-milligram-minute units give more convenient figures in dealing with single cells and have been more widely used by physiologists. It must always be remembered that there is considerable variation in the eggs of different females and less variation in the eggs of a single female (see Goldforb, 1917; for effects of ageing, Goldforb, 1918, a and b). The most probable average value is, therefore, taken as a basis for derived values. Some determinations cannot be made with great accuracy and consequently represent approximations, such as the di- ameter of granules and percentage of granules, and determinations based on such measurements. Where our knowledge of data is not complete, this compilation may serve as a guide to future investigation. 1. DIMENSIONS Egg diameter,1 74 /JL, average of 550 cells from four females. Eggs do not flatten under their own weight (McCutcheon, Lucke and Hart- 1 The European Arbacia pustulosa egg has a diameter of about 100 p.. 141 11 142 E. XEWTOX HARVEY line, 1931); 66 to 94 M (Glaser, 1914). Loss of diameter on fertiliza- tion, 2.3 n for 74 // egg (Glaser, 1924).2 Egg surface, 1 7200 n~. Egg volume, 21 2000 n*. Diameter and volume of eggs separated by centrifugal force. Colork^- and pigmented half separate from whole egg and clear and granular quarter from colorless half (E. B. Harvey, 1932a). Diameter Volume H M3 Whole egg 73 203690 (197400 ju3 I-ucke, 1932) Colorless half egg <><) 113100 (108600 M3 Lucke, 1932) Pigmented half egg 56 91950 (90700 M3 I^ucke. 1932) Clear quarter egg 53 77950 i .ranular quarter egg 40 33510 Thickness of jelly, 28 to 32 M (E. B. Harvey). :! Thickness of perivitelline space, 3-5 ju (E. B. Harvey). Thickness of hyaline plasma layer (ectoplasmic layer), 1-2 ju (E. B. Harvey). Diameter nucleus, 11.5 AI (E. B. Harvey). Diameter oil globules, 1 /u (E. B. Harvey). Diameter fifth layer granules,4 0.6-1 /JL (E. B. Harvey). Diameter yolk granules,'"' 0.7-1.1 M (E. B. Harvey); 0.3 n (Heil- brunn, 1926). Diameter pigment granules, 1.1-1.6/x (E. B. Harvey); 0.92 /Z (Heilbrunn, 1926). Percentage formed bodies ( E. X. Harvey).6 Nucleus 0.4; oil globules 1.0; fifth layer 4.8; yolk 27.2; pigment 5.5; fluid 61.1. Number of chromosomes (diploid), probably 38 (Morgan, 1927). 2. DENSITIES Sea water, 1.024 (Lyon, 1907). 1.02426 21.5° C./21.50 C. (Car- rey, 1905). 2 This has been questioned by others. 3 The undated references to E. B. Harvey apply to data furnished by her for this article but not published elsewhere. 4 This layer appears on top of the yolk layer when eggs are centrifuged for a long time. The granules are very variable in si/e and shape. The figures for granule diameter are only approximate. '" These granules are often irregular and frequently polyhedral in centrifuged eggs. 6 The percentage nucleus was calculated from its volume based on diameter inr.iMirements. The percentage of other formed bodies was made by measuring the diiin -n-ions of the spherical segment occupied by the layers in Fong centrifuged eggs, ulating the volume of the spherical segment and taking 74 per cent of this, since spline^ of equal size lying in a volume occupy 74 per cent of that volume. The urea must l.c considered only approximate, as it is not certain how much packing of tl es oo urs. PHYSICAL AND CHEMICAL CONSTANTS ARBACIA EGG 143 Egg with jelly, 1.090 (E. N. Harvey, 1931).7l Egg without jelly, 1.084 (K. N. Harvey, 1931). 7J (Lyon, 1907).H 1.0485-1.0656 (Heilbrunn, 1926£).9 Density oil? 10 Density nucleus? Density clear layer, 1.0358-1.0514 (Heilbrunn, 19266). n Density fifth layer granules? Density yolk granules, 1.1035-1.1269 (Heilbrunn, 19266). 12 Density pigment granules? Density clear half of egg separated by centrifugal force, 1.076 (E. N. Harvey, 1931). 7 Density yolk half of egg separated by centrifugal force, > 1.100 (E. N. Harvey, 1931).7 f 3. VISCOSITY Water ---- 0.01 poise (dyne-seconds per cm.2) at 20° C. Granule-free protoplasm, 0.018-0.025 (approximately .02) poise13 (Heilbrunn, 1926, a and b). Granule-free protoplasm, 0.04 (Heil- brunn, 1928).14 (See this book for effect of substances on viscosity.) Entire protoplasm, 2-3 times above. Fertilized egg considerably more viscous. For changes during de- velopment see Heilbrunn, 1920. For effect of CC>2 see Jacobs, 1922; for effect of fatty acid-salt buffers see Howard, 1931; for effect of acids and alkalies see Earth, 1929; for effect of temperature see Heilbrunn, 1924. 4. TENSION AT SURFACE Air/water surface tension - 73 dynes per cm. at 20° C. Unfertilized egg less than 0.2 dynes per cm., with considerable 7 Determined by centrifuging eggs in mixtures of sea water and .95m cane sugar at 22° C. and measuring the density of the sea water-sugar mixture in which they remain suspended with a hydrometer graduated for 60° F./600 F. There is con- siderable variation in density of different eggs, m = molal, 342 grains cane sugar to a liter of water. 8 Determined by centrifuging eggs in gum arabic in sea water. No difference in density up to 16-cell stage but pluteus has a density of 1.055 to 1.066. 9 Determined by centrifuging eggs in cane sugar solutions whose densities were determined from tables. 10 The relative density of oil, nucleus, etc., is in the order given, oil and nucleus being lighter than clear layer, the remaining granules heavier. 11 Calculated from density of whole egg and volume (19 per cent) occupied by heavy granules of egg. 12 Determined by crushing eggs and centrifuging crushed material in sugar solutions of various densities. 13 Determined by centrifuging granules and applying Stokes' law with Cun- ningham correction. 14 Determined by Brown ian movement method. 144 E. NEWTON HARVEY variation, for 25 per cent increase in surface area.15 (E. N. Harvey, 1931.) Initial tension 0.08 dynes per cm. for undistorted egg (fertilized or unfertilized) and approximately proportional to deformation (flat- tening) and to increase in surface area.16 Surface has properties of elastic membrane. Increase of surface force for 1 per cent increase of surface area is 0.005 dynes per cm. for unfertilized and 0.7 dynes per cm. for fertilized egg. Internal pressure about 40 dynes per cm.2 (Cole, 1932.) 5. OSMOTIC PROPERTIES Osmotic pressure equals that of sea water (22 atmospheres at 0° C.), whose depression of freezing point, A, : 1.81° C. (-• 1.805 to - 1.84). NaCl 0.52M, MgCl2 0.29M and cane sugar 0.73M (342.2 Drains per liter of solution) have same A as sea water. (Carrey, 1905, 1915.) Equilibrium given by (F0 -- b)P0 =: (Vex -- b}PtI, where F0 =: vol- ume in sea water, P0 -- osmotic pressure sea water, I volume in concentrated or diluted sea water, Pfx -- osmotic pressure in concen- trated or diluted sea water, /; == osmotically inactive material, 11 per cent (7 to 14 per cent) (McC'utcheon, Lucke and Hartline, 1931). No effect of narcotics on equilibrium. Permeability (k) to water defined as: where d\' dt -- rate of change in volume, A • - surface area, I' -- os- motic pressure at time /, Pex =r osmotic pressure within egg or of solution with which cell is in equilibrium. For water entering eggs at 20° ('., k == 0.087 ^ per /JT surface per atmosphere difference of pressure per minute; for water leaving eggs, k - 0.141 /A At 15° C., = 0.05-0.06 for endosmosis and 0.07 0.08 for exosmosis. Values are independent of osmotic pressure but depend on kind and propor- tions of salt in medium, injury, narcotics, etc. Q10 -'- 2 to 3 between 12° and 24° C. Thermal M = 15200. (Lucke, Hartline and Mc- ( "utchcon, 1931.) See also Northrop, 1927; McCutcheon and Lucke, 1926 and 1927. I or fertilized eggs k =: 0.12 to 0.1 7 at 15° C. for endosmosis (Lucke, unpublished). Lillie (1916; 1917) finds water enters fertilized eggs and butyric a< ill .,( ii\ated eggs four times more rapidly than unfertilized. I >i tci mined by centrifugal force necessary to pull egg into two halves. I >< •(( Tinim-d 1)\ force necessary to deform eggs by flattening. PHYSICAL AND CHEMICAL CONSTANTS ARBACIA EGG 145 Absence of ions (glucose solution) increases k from 0.05 to 0.1 at 12° C. and 0.0001M CaClo or MgCl2 added to glucose solution main- tains k same as in sea water (McCutcheon and Lucke, 1928). Cations decrease permeability to water, the effectiveness increasing with the valence of the cation. In 0.38m dextrose solution containing 0.005 A I K3 citrate (in which solution cells have high water permeaV bility), the following concentrations of cobaltammine chlorides were required to reduce permeability to the value obtained in sea water:— 0.00005M of the 6 valent salt, more than twice as much of the 4 valent salt, more than eight times as much of the 3 valent, and 64 times as much of the 2 valent salt, while this amount of the 1 valent salt was incompletely effective. Temperature 12° ± 0.5° C. (Lucke and Mc- Cutcheon, 1929). Anions increase permeability to water, the effectiveness increasing rapidly with the valence of the anions. In 0.38/w dextrose solution containing 0.0005M CaClo, 0.001M of potassium ferrocyanide was required to definitely increase permeability, twice as much ferri- cyanide, four times as much potassium sulphate, and eight times as much chloride. Temperature 12° ± 0.5° C. (Lucke and McCutcheon, 1929). Narcotics (urethanes and carbamates) in sea water do not decrease permeability to water beyond the value normally found in sea water. When dissolved in a non-electrolyte solution, they tend to decrease permeability to water. Thus, 0.025M w-butyl carbamate in 0.3Sm dextrose solution decreased permeability from 0.096 (the value in dex- trose solution alone) to 0.062. Temperature 15° ± 0.5° C. (Lucke, 1931). For effects of HCN and KCN on permeability to water see Blu- menthal, 1927; for ether see Heilbrunn, 1925; for anaesthetics and KCN see also Lillie, R. S., 1918, and Blumenthal, 1928. 6. PERMEABILITY Permeability to a solute (S) may be defined as where dS/dt = rate of change of amount of solute; A • - surface area; Cs =- external concentration; V - volume of the egg. k -- number of mols that will penetrate I//2 of surface in 1 minute with a concentra- tion difference between exterior and interior of 1 mol per liter and is 17 (Jacobs and Stewart, 1932): 17 Determined by change in volume of eggs in hypertonic solutions. 146 E. XEWTOX HARVEY Ethylene glycol . 3.6 X ICT15 Acetamid 5.8 X 10~1S Propionamid. . ... .14.2 X 10~15 Butyramid. . . 36.6 X 10~u Glycerol 0.5 X 10~15 For fertilized eggs k for ethylene glycol = 9.8 X 10~15 (Stewart and Jacobs. 1932). For permeability to non-electrolytes and NH4 salts see Stewart, 1(M1 ; for fatty acid-salt buffers see Howard, 1931. 7. ELECTRICAL PROPERTIES Electrical resistance." Interior 90 ohm-cm, or 3.6 times that of sea water at room temperature for 1000 to 15 X 106 cycles. Im- pedance of surface high below 1000 cycles and behaves like a " polari- /,n ion-capacity." No measurable change in resistance on fertilization (Cole, 1928). Dielectric constant? Cataphoretic potential,19 - 0.035 volts with jelly; - 0.021 volts without jelly for zeta potential (K. Dan, 1931). 8. HYDROGEN-ION CONCENTRATION pH of sea water, 8.2. pH of egg? Probably 6.6-6.8 from studies of other echinoderms by indicator method. No change on fertilization. Change to 5.4-5.6 on injury. Nucleus of immature egg 7.6-7.8 (Chambers and Pollak, 1027; Needham, 1926). Buffer value? 9. METABOLISM Heat production,20 - 0.08 gram-cal. per 10(i eggs of 74 p. diameter per hour; 0.88 gram-cal. on fertilization, falling to 0.58 gram-cal. 20 minutes after fertilization and 0.52 gram-cal. per 10fi eggs per hour at first cleave '50 minutes after fertilization). 3.34 ergs per egg per hour for unfertilized and 20 ergs per egg per hour for fertilized eggs (Rogers and Cole, 1925). Carbon-dioxide production? Sec Lynn, 1904. Oxygen consumption, 33.6 mm.* (17-51 mm.3) measured at 24.7° C. per K)'1 eggs (diameter 72-80 ju; average 77 /i) per hour at 24.7° C. or 0.0023 mm.3 O2 per mm.3 egg per minute. Fertilized eggs four limes with no \.ni.ition during cleavage21 (Tang, 1931). 30 mm. per 10' eggs (diameter 72 //) per hour at 25° C. .0025 mm.3 O2 per mm.:: I >ftrnnHici| hy vacuum thermocouple voltmeter-ammeter method. I irirniiinrd liy deflection from vertical fall in electric field. I 'fti-riiiiiD-d 1>\- thermocouple and twin-calorimeters. 1 Warburg manometer method. PHYSICAL AND CHEMICAL CONSTANTS ARBACIA EGG 147 egg per minute. Fertilized eggs 5 times greater with no variation during cleavage 21 (Tang and Gerard). Temperature coefficient of oxygen consumption -'• (Loeb and Wasteneys, 191 la). °c. Qio °C. Qio °C. Qio 3-13 2.18 10-20 2 17 17-27 2 0 5-15 2.16 13-23 2.45 20-30 1.96 7-17 2.00 15-25 2.24 Oxygen consumption of unfertilized eggs at different O2 tensions. 760 to 50 mm. O2, 100 per cent; 45 mm., 95 per cent; 20 mm., 90 per cent; 10 mm., 70 per cent; 6 mm., 55 per cent; 2 mm., 40 per cent (Tang, 1931) 21 at 24.7° C; see Gerard, 1931. Oxygen consumption of fertilized eggs at different O2 tensions. 228 to 80 mm. O2, 100 per cent; 20 mm., 90 per cent; 13 mm., 80 per cent; 9 mm., 60 per cent; 4 mm., 40 per cent; 2 mm., 20 per cent. Below 11 mm. division rate slows; below 4 mm. division rate ceases 23 (Amberson, 1928). Temperature about 22°. 760 to 50 mm., 100 per cent; 28 mm., 80 per cent; 20 mm., 60 per cent; 15 mm., 40 per cent; 9 mm., 20 per cent (Tang and Gerard).21 Temperature 25°. For effect of methylene blue alone, with cyanide and with nar- cotics, see Barren, 1929. For effect of CO2 and HC1 on oxygen consumption, see Root, 1930. For oxygen consumption under various conditions see McClendon and Mitchell, 1912; Loeb and Wasteneys, 19116, 1913, 1915; Waste- neys, 1916. 10. OXIDATION-REDUCTION POTENTIAL Not determined for Arbacia but from other sea urchin eggs. rH < 7.9 in nitrogen and 12 (£„' -- - 0.06 volts for pH = = 7) (Cham- bers, Pollak and Cohen, 1929, 1931) or 21-22 (Needham, 1926) in air by indicator method. No change on fertilization and during seg- mentation. Red pigment, echinochrome. E0 -- +.1995 volts at pH = 0, de- creasing 0.06 volts for each pH unit except above pH : - 8.78 where slope changes due to acid-base dissociation (Cannan, 1927). 11. COMPOSITION Total nitrogen? Total fat, carbohydrate and protein? - \Vinkler method. 23 Haldane analysis of gas in equilibrium with sea water. 148 E. NEWTON HARVEY Lactic acid, 3.14 mg. per gram egg protein (N2 X 6.25). 81 per cent increase if treated with 0.003M KCN for 3 hours.24 19 per cent more lactic acid in 4 to 8-cell fertilized eggs (Perlzweig and Barren, 1928). Free sugar absent (Perlzweig and Barren, 1928). Reducing sugar by acid hydrolysis == 50 mg. glucose per gram egg protein ( Perlzweig and Barron, 1928). Total solid, 18.1 per cent (McClendon, 1909). Ash, 8.5 10 per cent dry weight of eggs (Page, 1927). Calcium,--"' 1.9 mg. per 10''' eggs; 0.047 millimols; 106 eggs dried u.-igh 124 mg. (Page, 1927). Magnesium,2'5 4.48 mg. per 10fi eggs; 0.182 millimols (Page, 1927). Sodium,-7 1.301 mg. per 10'5 eggs; 0.056 millimols (Page, 1927). Potassium,2* 2.445 mg. per 10(i eggs; 0.063 millimols (Page, 1927). Iron,2" 0.030 nig. per 106 eggs; 0.0005 millimols (Page. 1927). Sulphate,™ 0.00046 mg. per 10° eggs; 0.00004 millimols (Page, 1927). Chloride,-'1 0.1864 mg. per 10" eggs; 0.0053 millimols (Page, 1927). Total phosphate,'- 0.9064 mg. per 10" eggs; 0.0291 millimols (Page, 1927). Acid-soluble P, 0.40-0.50 mg. per 10r> eggs; 0.0161 millimols (Page, unpublished). Lipoid P, 0.498 mg. per 106 eggs; 0.160 millimols (Page, unpub- lished). Nitrate, trace (Page, 1927). Silica, moderate amounts (Page, 1927). Copper, 17 micrograms per cc. eggs for unripe ovarian eggs; 175 micrograms per cc. for unfertilized and 21 micrograms per cc. for fertilized eggs (Glaser, 1923). Cholesterol, present (Mathews, 1913). Oil; iodine number 146-148; saponification value 606 (Page, 1927). Echinochrome; see McClendon, 1912. Catalase content of unfertilized and fertilized eggs same (Amberg and YVinternitz, 1911); see Lyon, 1909. Composition of layers of crushed and centrifuged unfertilized eggs (with jelly) in percentages of weight of whole egg mass (P phos- phorus, N -- nitrogen) (McClendon, 1909). -4 Nitrogen determined by micro-Kjeldahl and lactic acid by Clausen's method. ;4 Precipitated by NH4 oxalate and ignited to CaO. 6 Weighed as Mg2P,O7. Kramer and Gittlemann pyroantimonate method. Kr.inici and Tisdall method. -"' Allen-Scott colorimetric method. Weighed as BaSO... Mnhr met hod. linn-die i I'lu-i- method. PHYSICAL AND CHKMICAL CONSTANTS ARKACIA EGG 149 Layer Centripetal 32.5 Centrifugal 67.5 •'« Whole Egg \Yater 28.6 53.3 81.9 Solids 3.9 14.2 18.1 Kt her ext. 0 308 1 946 2 254 P in ether ext 0 00154 0 06760 0 069 1 4 Alcohol ext 1 60 3 48 5 08 P in alcohol ext. . . . 00434 0 0814 0 1248 Water ext 0.78 1.42 2.20 P in water ext 0 130 0 182 0312 Residue of water ext. . . 1 309 7 29 8 59 P in residue 0.0392 0.1167 0.1559 N in residue 0.1625 0.7750 0.9375 Ash in residue 0.0162 00264 0 0426 Total P in layers 0 21414 0 4478 0 66194 33 Contains jelly. Analysis of eggs washed with isotonic glucose (K. C. Blanchard, unpublished). Ash 7-8 per cent dry weight of eggs Calcium 6.45 ± 0.3 per cent of ash Magnesium 1.77 ± 0.02 per cent of ash Potassium 15.79 ± 0.25 per cent of ash Lactic acid 0.4 mg. (when eggs first shed) to 6 nig. (on standing) per gram egg protein Glycogen 50-80 mg. per gram egg protein For analysis of autolysed eggs see Lyon and Shackell. 1910. For egg secretions see Lillie, F. R., 1913; Glaser, 1914, 1921, 1922; Woodward, 1918; Clowes and Bachman, 1921. 12. RATE OF DEVELOPMENT Time in minutes from fertilization to first cleavage (Loeb and Wasteneys, 191 la). °c. Time °C. Time °C. Time 7 498 16 85 26 33 8 410 17.5 70 27.5 34 9 308 18 68 30 33 10 217 20 56 31 37 12 147 22 47 32 No cleavage 15 100 25 40 35 No cleavage °c. QlO °C. Qio °C. Qio 7-17 7.3 10-20 3.9 15-25 2.6 8-18 6 12-22 3.3 16.26 2.6 9-19 > 4 13-23 3.3 17.5-27.5 2.2 20-30 1.7 150 E. XEWTOX HARVEY Time from fertilization to 50 per cent first cleavage, 62.5 minutes at 20.6-20.8° C. (Haywood and Root, 1930); 55 minutes to 50 per cent first cleavage and 83 minutes to second cleavage at 22° C. (Lillie and Cattell, 1923); 50 minutes to 50 per cent first cleavage at 22° C. (Whitaker, 192<>). Time from fertilization at 23° C. to various stages of development (E. B. Harvey). ( dinpletion of fertilization membrane, 1 to 2 minutes. Hyaline plasma layer begins 2 minutes. 1'iiion of pronuclei (monaster), 10 minutes. Nuclear streak begins 20, ends 40, marked 25-30 minutes. Nuclear membrane disappears, 40 minutes. \mphiaster, 45 minutes, l-irst cleavage (50 per cent), 50 minutes. Second cleavage (50 per cent), 78 minutes. Third cleavage (50 per cent), 103 minutes. For effect of sea water diluted with .73M cane sugar see Lillie and Cattell, 1923. For effect of CO2 and acids on development see Clowes and Smith, 1923, Smith and Clowes, 1924, a and b, and Smith, 1925. For effect of CO2 on cleavage rate see Haywood, 1927, and Hay- wood and Root, 1930. For effect of lack of oxygen see Harvey, E. B., 1930, and Lyon, 1902. For effect of HgCl2 see Hoadley, 1930. For effect of mechanical shocks see Whitney, 1906. For cleavage rates of centrifuged egg fragments see Harvey, E. B., 1932a, and \\ hitaker, 1929. 13. NARCOSIS Anaesthetic and Lethal Concentrations (Heilbrunn, 1920) of Anaestlict ic Lethal per cent per cent Ether 2 4 Chloroform 0.1 .* 1 Chloral hydrate O.J5 1 Nitromethane 2 3 I'aral.l.-hyde 4 Acetone 5 10 I iliyl nitrate 0.4 l.iliyl acetate 2 5 1 .1 liyl butyrate 0.25 0.5 A< rtonitrile 4 5 \ l'n>|>\ 1 alcohol 1 ? Amyl alcohol 0.6 1 I'hriivl nrcthaiir 4/5 sat. sat. I i liyl urcthaiic 1.5 3 PHYSICAL AND CHEMICAL CONSTANTS ARBACIA EGG 151 Critical narcotic concentrations for reversibly suppressing cleavage (Harvey, E. B., 19326). Ethyl urethane 0.15-0.2 M N-propyl urethane . . .0.07 M Isopropyl urethane . .0.1 M N-butyl carbamate . . .0.025 M Isoamyl carbamate . .0.01 M Phenyl urethane 0.00125-0.0025 M Critical narcotic concentrations for reversibly suppressing cleavage (Lillie, 1914). Methyl urethane 2-2.5 per cent; .29-.S3 M Ethyl urethane 1.5-1.75 per cent; .15-.19 M Phenyl urethane 08-.06 per cent; .005-.006 M Ethyl alcohol 5 vol. per cent; .87 M Propyl alcohol 2 vol. per cent; .27 M Isopropyl alcohol 3 vol. per cent; .4 M H-Butyl alcohol 8 vol. per cent; .086 M Isoamyl alcohol 4 vol. per cent; .037 M Capryl alcohol 015 vol. per cent; .001 M Ethyl ether 5-.6 vol. per cent; .05-.06 M Chloroform 06 per cent; 1/12 saturated; .005 M Chloral hydrate 1-.12 per cent; .006-007 M Chloretone 2-.25 per cent; .008-.01 M See also Baldwin, 1920, for effect of alcohols on dividing eggs. 14. MISCELLANEOUS For effects of light see Lillie and Baskervill, 1921, 1922, and Hin- richs, 1926, 1927. For effect of salts see Lillie, R. S., 1910, 1911; Lillie and Basker- vill, 1921. For effect of supersound waves see Harvey, Harvey and Loomis, 1928, and Harvey and Loomis, 1931. For rhythms of susceptibility during development see Lyon, 1904, Baldwin, 1920, and Page, 1929; Lillie, R. S., 1916 and Just, 1928, for hypotonic sea water; Moore, A. R., 1915, for hypertonic sea water. For initiation of development and related phenomena see Greeley, 1902; Hunter, 1903; Harvey, E. N., 1910; Loeb, 1913, 1915; Heil- brunn, 1913, 1915, 1920, a and b, 1925 ; Lillie, F. R., 1914, 1921 ; Glaser, 1914; Moore, C. R., 1916, 1917; Just, 1922, 1928, a and b, 1929. For surface precipitation reaction see Heilbrunn, 1930. For agglutination see Goldforb, 1929. For nature of fertilization membrane see Kite, 1912; Heilbrunn, 1913, 1915, 1924a; Harvey, E. N., 1914; McClendon, 1914; Carrey, 1919; Chambers, 1921. 152 E. XEYYTOX HARVEY LITERATURE AMBERG, S., AND M. C. WINTERNITZ, 1911. Jour. Biol. Client., 10: 295. AMBERSON, YY. R.. 1»Js. Biol. Bull., 55: 79. BALDWIN, E. M.. I'JJd. Biol. Bull., 38: 123. BARRON, E. S. < ,., 1020. Jour. Biol. Chem., 81: 445. EARTH, L. ('-.. 1929. Protoplasma, 7: 505. BLUMEVMIM . R., 1027. Biol. Bull., 52: 313. BLUMENTHM.. k.. 1028. Physiol. Zool., 1: 269. CAN\A\, K. K., 1027. Biochem. Jour., 21: 184. CHAMBERS, R., 1021. Biol. Bull., 41: 318. A vi. II. POLLAK, 1927. Jo«r. G?H. Physiol., 10: 739. B. COHI-.V, .111,1 II. Pnl.LAK, 1929. ( tiiii.v AVI. 1 1. POLLACK AND E. BACHMAN, 1921. A , AND H. \Y. SMITH, 1923. Jour. Gen. Physiol., 12: 37. Jour. Cell, and Com par. Physiol., 1:1. ( 'n \\IHEK^, R., ClIAMMEIO. K.. ( 'HAMMER^, ( '!.' .U ES, ' .. ( I i i\\ I S, < .. ( :oi i , K. S., 1928. COLE, K. S., 1932. R.. II. H. B. A., Brit. Jour. Exper. Biol., 6: 229. 1031. Jour. Exper. Biol., 8: 1. Jour. Biol. Chem., 46: xxxi. Am. Jour. Physiol., 64: 144. DAN, K., 1931. CARREY, \\ . I.., V, CARREY. \\". H.. GARREY, \Y. E., GERARD, R. \\"., GLASER, O. C. ( ,1 \-I-.K GLASER .1 nat. RI-I'.. Suppl. O. C., O. C., . O. i .. GLAH:K, O. C., 1905. 1915. 1919. 1031. 1014. 1921. 1922. 1023. 1924. Biol. Bull, Biol. Bull., Biol. Hull., , 51 : 2S. 8: 257. 28: 77. 37: 287. Hull., 60: 245. Biol. Bull. Biol. Bull. Biol. Bull. Biol. Bull. Biol. Bull. GOLDI ORB, A. J., 1017. GOLDFORB, A. J., 1918«. 26 41 43 44 47 84, 367, 368. 63. 175. 73. I H >I.DH)KH, A. J., GOLDFORB, A. J., GREELEY, A. \Y., HARVEY, E. !'•. HARVEY, E. B. HARVEY, I-:. B. HARVEY, E. X HARVEY, E. X. HARVEY, E. X. 1918&. 1929. 1002. 1030. 1932a. 19326. IOHI. 1914. 1931. 274. Inst. l}ni'. Xo. 251, p. Biol. Bull., 34: 372. Biol. Bull., 35. 1. Biol. Bull., 57: 333, 350, 389. Biol. Bull.. 4: 120. Biol. Bull., 58: 288. Biol. Bull., 62: 155. I npul.lishcd. Jour. Exper. Zool., 8: 355. Biol. Bull., 27: 237. Biol. Bull., 61: 273. HARVEY, E. V. I-:. B. HARVEY AND A. L. LOOMIS, 102S. Biol. Bull,, 55: 459. HAKVI s. I.. V. \VD A. L. LOOMIS, 1931. Jour. (!en. Physiol., 15: 147. Biol. Bull., 53: 450. 2t<. Biol. Bull., 50: 1, 455. HIM M n-. M. A., 1027. Biol. Bull., 53: 416. HOADLEY, I .. l''Sn. Biol. Bull., 58: 123. HAYWOOD, C., 1027 . Biol HAYWOOD, ( '., AND \V. S. 1 HKILURI NN, ,. V., 1913. HEILHIU-NN, , V., 1915. HKILHRUNN, .. V., 1920«. HEILBRUNN, .. V., 1920/>. HEILHRUNN, .. V., 1924a. 1 ll II.MRUNN, .. V., 1924/;. 1 Ir.ll.MKrMN, .. V., 1925. I ll- II l:|.:r.N'N, .. V., 1926a. 1 ll.ll.lik CNN, .. V., 1926/A III ll i ! in NN, ,. V., 1928. 1 ll II likl NN, .. V., 1930. PHYSICAL AND CHEMICAL CONSTANTS ARBACIA EGG 153 HOWARD, E., 1931. Biol. Bull., 60: 132. HUNTER, S. J., 1903. Biol. Bull., 5: 143. JACOBS, M. H., 1922. Biol. Bull., 42: 14. JACOBS, M. H., AND D. R. STEWART, 1932. Jour. Cell, and Compar. Physiol., 1 : 71, 83. JUST, E. E., 1922. Kwl. Hull., 43: 384, 401, 411. JUST, E. E., 1928. Protoplasma, 5: 97. JUST, E. E., 1928o. Biol. Bull., 55: 358. JUST, E. E., 192S6. Phvsiol. Zool., 1: 26. JUST, E. E., 1929. Biol. Bull., 57: 422. KITE, G. L., 1912. Science, 36: 562. LILLIE, F. R., 1913. Jour. Kxper. Zool., 14: 515. LILLIE, F. R., 1914. Jour. Kxper. Zool., 16: 523. LILLIE, F. R., 1921. Biol. Bull., 41: 125. LILLIE, R. S., 1910. Am. Jour. Physiol., 26: 106. LILLIE, R. S., 1911c. Am. Jour. Physiol., 27: 289. LILLIE, R. S., 19116. Jour. Morpli., 22: 695. LILLIE, R. S., 1914. Jour. Biol. Ctiem., 17: 121. LILLIE, R. S., 1916a. Am. Jour. Physiol., 49: 13, 249. LILLIE, R. S., 19166. Jour. Exper. Zool, 21: 369. LILLIE, R. S., 1917. Am. Jour. Physiol., 43: 43. LILLIE, R. S., 1918. Am. Jour. Physiol., 45: 406. LILLIE, R. S., AND M. L. BASKERVILL, 1921. Am. Jour. Physiol., 57: 110. LILLIE, R. S., AND M. L. BASKERVILL, 1922. Am. Jour. Physiol., 61: 272. LILLIE, R. S., AND W. CATTELL, 1923. Jour. Gen. Physiol., 5: 807. LOEB, J., 1913. Artificial Parthenogenesis and Fertilization. Chicago. LOEB, J., 1915. Biol. Bull,, 29: 103. LOEB, J., AND H. WASTENEYS, 191 la. Biochem. Zeitschr., 36: 345. LOEB, J., AND H. WASTENEYS, 19116. Biochem. Zeitschr., 37: 410. LOEB, J., AND H. WASTENEYS, 1913. Jour. Biol. Chem., 14: 355, 459, 469, 517. LOEB, J., AND H. WASTENEYS, 1915. Jour. Biol. Chem., 21: 153. LUCRE, B., 1931. Biol. Bull., 60: 72. LUCRE, B., 1932. Unpublished experiments. LUCRE, B., AND M. MCCUTCHEON, 1929. Jour. Gen. Physiol., 12: 571. LUCRE, B., H. K. HARTLINE AND M. MCCUTCHEON, 1931. Jour. Gen. Physiol., 14: 405. LYON, E. P., 1902. Am. Jour. Phvsiol., 7: 56. LYON, E. P., 1904. Am. Jour. Physiol,, 11: 52. LYON, E. P., 1907. Arch. f. entw.-mech., 23: 151. LYON, E. P., 1909. Am. Jour. Physiol,, 25: 199. LYON, E. P., AND L. F. SHACRELL, 1910. Jour. Biol. Chem., 7: 371. MATHEWS, A. P., 1913. Jour. Biol. Chem., 14: 465. MCCLENDON, J. F., 1909. Am. Jour. Physiol., 23: 460. McCLENDON, J. F., 1912. Jour. Biol. Chem., 11: 435. McCLENDON, J. F., 1914. Internal. Zeitschr. Physik. Chem. Biol., 1: 163. MCCLENDON, J. F., AND P. H. MITCHELL, 1912. JOUY. Biol. Chem., 10: 459. MCCUTCHEON, M., AND B. LUCRE, 1926. Jour. Gen. Physiol., 9: 697. MCCUTCHEON, M., AND B. LUCRE, 1927. Jour. Gen. Physiol., 10: 659. MCCUTCHEON, M., AND B. LUCRE, 1928. Jour. Gen. Physiol., 12: 129. MCCUTCHEON, M., B. LUCRE AND H. K. HARTLINE, 1931. Jour. Gen. Physiol., 14: 393. MOORE, A. R., 1915. Biol. Bull., 28: 253. MOORE, C. R., 1916. Biol. Bull., 31: 137. MOORE, C. R., 1917. Biol. Bull., 33: 258. MORGAN, T. H., 1927. Experimental Embryology, p. 627. NEEDHAM, J., AND D. M. NEEDHAM, 1926. Proc. Roy. Soc., London, B, 99: 173. NORTHROP, J. H., 1927. Jour. Gen. Physiol., 11: 43.' 154 E. NEXYTON HARVEY PAGE, I. H., 1027. Biol. Bull., 52: 164, 168. PAGE, I. H., 1929. Brit. Jour. Exper. Zool., 6: 219. PERLZWEIG, XX'. A., AND K. S. ( ,. BARROX, 1928. Jour. Biol. Chem., 79: 19. ROGERS, C. G., VND K. S. COLE, 1925. 5zW. 5«//M 49: 338. ROOT, \Xr. S., L930. Biol. Bull., 59: 48. SMITH, H. XX".. 1'>J5. Am. Jour. Physiol, 72: 347. SMITH, H. XX'., VND G. II. A. ("LOWES, 1924a. Am. Jour. Physiol., 68: 183 SMIIII, H. \\'., AM. G. H. A. CLOWES, 1924ft. Biol. Bull., 47: 304, 323. 335. STEU \ki, D. K., 1931. Hiol. Bull., 60: 152, 171. STEWART, 1 >. R., AND M. H. JACOBS, 1932. /owr. Cell, and Compar. Phvsiol., 1: 83. TANG, I'., 1931. Biol. Bull.. 60: 242. TANG, I'., AND R. ( '.KUAKI>. To he published. WASTENEYS, H., 1916. Jour. Biol. Chem., 24: 2S1. XVmiAKER, D. M., 1929. Biol. Bull., 57: 159. XX'HITNEY, D. D., 1906. Jour. Exper. Zool., 3: 41. XX'oomvARD, A. E., 1918. Jour. Exper. Zool., 26: 459. THE DEVELOPMENT OF HALF AND QUARTER EGGS OF ARBACIA PUNCTULATA AND OF STRONGLY CENTRIFUGED WHOLE EGGS ETHEL BROWNE HARVEY (From the Marine Biological Laboratory, Woods Hole, and Washington Square College, New York University) When the unfertilized eggs of Arbacia punctulata are strongly centrifuged in a sugar solution of the same (or graded) density, they become dumb-bell shaped and then separate into two nearly equal spheres or half-eggs, one colorless and the other pigmented (E. N. Harvey, 1931). Further, when the colorless spheres are again cen- trifuged, more strongly, in a sugar solution of the same (or graded) density, they separate into two spheres or quarter-eggs, one without visible granules and the other with granules. All of these half- and quarter-eggs, as well as the deformed total eggs just before pulling apart, can be fertilized and develop. We can in this way obtain "egg fragments" of very definite size and content in great numbers. This obviates the former tedious method of cutting individual eggs with a glass needle (Harnley, 1926; Horstadius, 1928; Plough, 1929; Tennent, Taylor and Whitaker, 1929; Whitaker, 1929) and insures a much more accurate division of the egg into parts of known structure than the old haphazard and harmful method of shaking the eggs into pieces (Hertwigs, Boveri, Driesch, Morgan, etc.).1 Sea urchin eggs centrifuged in sea water alone do not separate into spheres owing to the fact that they sink to the bottom and are crushed by the force exerted on them. The method employed was to put two parts 0.95 molal sugar (95 per cent of 342 grams cane sugar added to 1 liter tap water, an isotonic solution) in the centrifuge tubes and above this one part of sea water containing eggs; the tubes were rolled gently to obtain partial mixing and then centrifuged at about 7000 r.p.m. (11 cm. radius) for four minutes. On removal of the tubes, the colorless spheres formed a whitish layer just beneath the surface, the whole eggs, which had not broken apart, formed a reddish layer about halfway down the tube and the pigmented spheres rested at the bottom of the tubes. One could then pipette off any one of the three kinds of eggs with practically no admixture of the other varieties. The sugar solution was found to have no ill effect on the eggs, since eggs could be kept in this solution for at least five hours 1 The centrifuge method of separating eggs supplements rather than supplants the cutting method, since the granular constitution of fragments is different in the two cases. 155 156 ETHEL BROWNE HARVEY and when returned to sea water could be fertilized, and developed normally. Eggs cannot be fertilized, however, while in the sugar solution although the sperms are active and surround the egg. The rapid centrifugini; also has no ill effect upon the eggs, as eggs subjected to this treatment develop into normal plutei when fertilized in sea water. To separate the colorless half-eggs into two parts, they were taken from the centrifuge tubes and put into other centrifuge tubes above a small amount of cane sugar solution (3 parts of 0.95 molal sugar to 1 part sea water) and a little sea water placed above. The tubes were then rentrifuged for 45 minutes at about 10, 000 r.p.m. (radius 11 cm.) Unfertilized Whole, Half- and Quarter-eggs When normal Arbacia eggs are centrifuged as described above, they become stratified while still spherical into five layers, (1) oil on top, (2) a clear layer without visible (i.e., under ordinary illumination) granules, in which lies the nucleus, (3) a thin granular layer separated out only with strong centrifugal force, and not heretofore described and which we have, therefore, termed "the fifth layer," (4) a large yellowish yolk layer, and (5) a layer of red pigment concentrated at the heavier pole (Fig. 25). The fifth layer can be very beautifully demonstrated by staining with methyl violet or methyl green when it becomes purple, or with Janus green when it becomes green. The eggs after stratification become elongate, then dumb-bell shaped (Fig. 25) and then break into two slightly unequal spheres, usually through the upper part of the yolk. The colorless sphere (cf. Fig. 1), slightly larger, contains therefore the oil, below which always lies the nucleus; the clear layer, without visible granules; the granular or fifth layer; and a little yolk at the heavier pole. The pigmented sphere, (cf. Fig. 13) slightly smaller, contains usually only yolk and pigment, the pigment being massed at the heavier pole; there is, of course, no nucleus in this sphere. In a few batches of eggs, the two spheres were almost equal in size, the separation having taken place in the fifth layer, a little of which appeared as a light cap on the pigmented sphere. Usually, however, there is quite an appreciable difference in size of the two half-eggs, and their relative size is fairly constant in any one centrifuged lot. When the colorless spheres are centrifuged again, as described above, they separate into (1) a larger sphere with oil, nucleus and clear layer — this quarter-egg now having practically no visible granules (Fig. 38); and (2) a quite small sphere '•'imposed entirely of the granules of the fifth layer and yolk, about hall of c.ich (cf. Fig. 46). Die ,i\ t rage measurements obtained in typical lots of eggs are given m I iMi I. A photograph of unfertilized whole eggs and the two 12 Photograph 1. Unfertilized whole and half-eggs. Photograph 2. 2 and 4-cell whole and colorless half-eggs. Photograph 3. Blast ulae of whole and colorless half-eggs. Photograph 4. Five tyi>es of eggs, whole, half- and quarter-eggs. DEVELOPMENT OF HALF AND QUARTER ARBACIA EC.CS 157 kinds of half-eggs is given in Photograph 1. Photograph 4 is a picture of the five types of eggs, whole, half- and quarter-eggs. Development of Half -eggs Both the colorless and pigmented half-eggs can be fertilized in sea water immediately after centrifuging or at any later time. They both form fertilization membranes at the same time as the whole egg. These are often well separated from the surface of the egg but some- times, especially in the pigmented spheres, rather closely investing. There is in both half-eggs a well marked ectoplasmic or hyaloplasmic layer formed on fertilization. The development of the two half-eggs must now be followed separately. All observations were made on living material, and the times given are for 23° C., when 50 per cent normal eggs cleave in 50 minutes. The development of the colorless half-egg is quite normal. Many of the nuclear phenomena accompanying fertilization and cleavage can be seen with great clearness, and they parallel those described by TABLE I Diameter Volume Sums of Volumes Whole Egg M 73 M3 203700 M3 Colorless Half Egg 60 113100 Pigmented Half Egg. . 56 91950 205050 Clear Quarter Egg 53 ' 77950 Granular Ouarter Egg 40 • 33500 111450 Wilson (1895) in the living Toxopneustes egg, which is devoid of pigment and quite transparent; these phenomena cannot be observed in the normal living Arbacia egg on account of the pigment. The sperm aster can be seen in the granular area a few minutes after fertilization, and its approach toward the female pronucleus can be followed (Fig. 2). The female pronucleus travels down from its position under the oil cap toward the center of the egg about eight minutes after fertilization (Fig. 2), and as it moves down some of the oil spherules also move down from the oil cap. After the union of the pronuclei (Fig. 3) astral radiations extend through the granular area; in fact all the granules are often arranged in rays extending in a half circle from the nucleus (Fig. 4) ; there is no indication of rays (in the living egg) in the clear area free of granules. In the normal uncen- trifuged egg at this period, the astral radiations extend throughout the cell, this being the "monaster stage" (Fig. 36). These radiations 158 ETHEL BROWNE HARVEY gradually fade out ami the nucleus enlarges from a diameter of 12 ju to 16 n (Fig. 5). This is the "streak stage" of the normal Arbacia egg, lasting from 20 to 40 minutes after fertilization, characterized by a curved band extending on either side of the nucleus (Fig. 37) ; this stage is referred to by Wilson in the Toxopneustes egg as the "pause." The nuclear wall now breaks down (Fig. 6) and soon afterwards there are again radi.iti< ms, now from the two poles of the amphiaster, but present only in the granular zone (Figs. 7, 8). The half-eggs divide about the same time as normal control eggs, or sometimes a little earlier as \Vhitaker (1929) found for his diploid fragments; the plane of cleavage iiMially comes in perpendicular to the stratification (Fig. 9), but sometimes parallel with it (Fig. 10) or in any intermediate position, but it divides the egg into two equal parts (photograph 2) as in the normal egg. The following cleavages come in at right angles to the preceding and divide the blastomeres equally (Figs. 11, 12). Blastulae (photo- graph 3) are formed from the half-eggs, quite normal except for size and coloring. In many of the cultures the larvae remained for several days as very actively swimming blastulae; in some they developed into gastrulae and in some into plutei with well developed skeleton and arms exactly like the normal ones except that they were colorless and only half the size. The pigmented spheres have no nucleus at the time of fertilization. The aster accompanying the sperm nucleus may very often be seen in the yolk 15 20 minutes after the fertilization membrane has been given off (Fig. 13). The nucleus, at first very small, enlarges and is quite noticeable about 30 minutes after fertilization (Fig. 14), and about the time of first cleavage of the whole and colorless half-eggs. This single nucleus now usually without radiations, after enlarging considerably (Fig. 15), breaks down about one hour after fertilization (Fig. 16). At about 80 minutes after fertilization (Fig. 19), when the total and colorless spheres are in the 4-cell stage, many of the pigmented ^pheres h.i\<- i\vo nuclei; either a dumb-bell shaped nucleus or a very small amphiaster can sometimes be seen preceding the binucleate stage (Figs. 17, 18). Usually no cleavage plane comes in, probably owing to the mechanical difficulty in cutting through the dense material. After several successive divisions of the nuclei, the pigmented half-eggs appear as reddish spheres containing many white circles, some with radiations, giving somewhat the appearance of pictures of the moon with craters on the surface (Fig. 20). There are often slight indi- • aliens of cleavage planes running in from the periphery as indentations <»r not (lies (Figs. 21, 22). Rather rarely, the division planes come in quite normally and the egg divides into 2, 4 and 8 equal blastomeres DEVELOPMENT OF HALF AND QUARTER ARBACIA EGGS 159 (Figs. 23, 24). The first cleavage plane in these cases may come in in any relation to the stratification, and the following planes at right angles to the preceding. Although it is difficult to determine in living material, it would seem that when cleavage planes do not come in at first, they come in later on after some 20 or 30 nuclei are formed, for the blastulae seem quite normal and appear, as far as I could tell, multi- cellular. The blastulae are quite active, but are not very viable, and relatively few in any batch develop much further. I have, however, had a number of quite normal gastrulae and some plutei with skeletons, but only two with well developed arms. None have survived more than nine days. There seems no doubt that these merogonic eggs can give rise to dwarf embryos similar to normal ones except for containing more pigment. Whether the failure of the majority of these eggs to develop far is due to lack of certain formative stuffs or to the lack of the female pronucleus, or to the over-crowding with dense material is not certain, but I should judge from the appearance and behavior of the developing eggs that the last explanation is the correct one. It may be, however, that only those half-eggs containing some of the fifth layer develop, but this awaits further investigation. It would seem then that both half-eggs can develop into plutei, and that neither food material (or very little) nor a female pronucleus is necessary for development. Development of Strongly Centrifuged Whole Eggs When the unfertilized Arbacia eggs are strongly centrifuged, as mentioned before, they elongate, then become dumb-bell shaped before breaking apart (Fig. 25). When these are left either in the sugar solution or in sea water, they soon become oval and then spherical, often within an hour. But if they are removed to sea water and fertilized immediately, they retain their dumb-bell shape and the fertilization membrane follows the contour of the surface with usually a bulge at one or both poles. This must be due to a "setting" or •gelation of the protoplasm following fertilization, for if the fertilization membranes are removed by drawing the eggs into a capillary pipette, the eggs retain their aspherical shape (and develop). This is another indication of increased viscosity following fertilization. One can often observe the male aster in the yolk or granular zone, and this apparently pulls some pigment granules along as it travels toward the center (Figs. 26, 27). The descent of the female pronucleus can be observed and the pull of the oil spheres toward the center, the union of the two pronuclei and the gradual fading out of the sperm aster whose radi- ations have been visible only in the granular zone (Fig. 27), the enlarge- LLJLIBRARY: 160 ETHEL BROWNE HARVEY ment of the nucleus and its rupture (Figs. 29, 30). A very striking stage is shown in Fig. 28, a sharp line of demarcation running between the clear zone and the granular zone; this corresponds to the "streak" stage in the normal egg (Fig. 37). A streak is only rarely seen running perpendicular to the stratification in well centrifuged eggs, probably owing to the denseness of the material. Later, rays from the amphiaster are seen in the yolk and granules of the fifth layer, which by this time are thoroughly mixed (Fig. 31); either a whole aster whose mate is sometimes faintly distinguishable in the clear zone into which some mules have spread, or less commonly two half asters in the granular and ynlk zone. It would seem that there must be a change of axis in the mitotic figure, for in normal eggs the long axis of the spindle is the same as the long axis of the "streak," whereas here it is usually perpen- dicular to it. The first cleavage plane comes in at almost the same time as in normal eggs, usually through or near the constriction and parallel (or at a slight angle) with the stratification, separating one colorless blastomere from the one containing yolk and pigment (Fig. 32). These two cells are usually unequal in size, more frequently the colorless cell is the smaller, though the two blastomeres are sometimes of size corresponding with the two half-eggs, the cleavage plane following the future separation plane (Photograph 2). The first cleav- age plane comes in rarely in these elongated eggs perpendicular to the stratification, and this is sometimes followed by a second cleavage plane also along the long axis resulting in four sausage shaped cells. The second cleavage plane always comes in perpendicular to the first (Fig. 33, Photograph 2). Many years ago, Lyon (1907) and Morgan and Lyon (1907) found that the first cleavage plane in centrifuged Arbacia eggs was usually perpendicular to the stratification. The apparent contradiction is explained by the difference in shape of the eggs. In elongated eggs, the first cleavage plane comes in usually parallel with the stratification, in the short axis. In spherical eggs it comes in perpendicular with the stratification; this is true both for eggs which are spherical because not centrifuged sufficiently to become elongate, and for eggs which have been elongate but have resumed a spherical shape on standing in sea water before fertilization. In slightly elongate or oval shaped eggs, the first cleavage plane comes in one way or the other in about equal numbers. The later cleavage planes often come in fairly regularly except that the clear cells often divide in advance of the pigmented cells, and are -in, ill. T (Fig. 34). Very frequently, the original first cleavage plane ivin.iins quite prominent, the first two blastomeres developing almost independently i Fig. 35). So much so, in fact, that double embryos are DEVELOPMENT OF HALF AND OUARTER AKHACIA HOGS 161 often produced, one colored and the other colorless, at first within the same membrane, and later swimming attached together. Slipper shaped blastulae (Photograph 3), normal gastrulae and plutei arise from the elongate and dumb-bell shaped eggs, but the stratification of pigment and yolk remains, and may be in any relation, apparently, to the axis of the embryo. Individual eggs have not been studied with reference to polarity. Development of Quarter Eggs The colorless half-eggs, when centrifuged again, are drawn out into dumb-bells (Fig. 51) and are then separated into spheres. The stratification is the same as before, the nucleus lying just below the oil cap in a large clear layer without visible granules; at the heavier pole is the granular or fifth layer and a layer of yolk. The half-egg breaks usually at the line between the clear layer and the granules, so that we obtain one perfectly clear sphere, with oil cap and nucleus (Fig. 38) and one granular sphere containing about an equal amount of granules (fifth layer) and yolk (cf. Fig. 46). This granular quarter-egg is much smaller than the clear quarter, about one sixth the volume of the whole egg. Both of these quarter-eggs, as well as the dumb-bell shaped half- eggs can be fertilized in sea water, and throw off fertilization mem- branes. The ectoplasmic layer of the granular quarter is much thicker than that of the clear quarter, where it is thinner than in normal eggs. Owing to the absence of granules in the clear quarter, nothing can usually be seen of the speim aster; the female pronucleus can be observed migrating from the oil cap to the center of the egg about 25 minutes after fertilization, pulling some of the oil spheres along. This gradually enlarges from 12 p. to 16 /*, but no other change occurs for six hours or more (Fig. 39). In some eggs the nucleus becomes enormous, as large as 22 /j. (nearly half the diameter of the egg, and an increase of six times in volume), but this is probably abnormal. The nucleus later breaks down and disappears (Fig. 40) and cleavage takes place some seven hours (or more) after fertilization (Figs. 41, 42) and in any plane with regard to the oil cap. The very slow cleavage of these diploid quarters is not in accord with Whitaker's (1929) expla- nation of cleavage rates. Other cleavages follow slowly, but usually the membrane breaks and there is a loose mass of cells, some perfectly clear and some with oil drops (Figs. 43, 45). I have obtained a few intact later cleavages, particularly in eggs left for several hours after recentrifuging before fertilizing them (Fig. 44). The granular quarter-eggs (Fig. 46), on the other hand, although having no female nucleus, develop quite normally. The male aster can 162 ETHEL BROWNE HARVEY be seen in the granules (Fig. 47), the nucleus enlarges, disappears, and the egg cleaves a little later than the control eggs (Fig. 48). This then divides into 4, S, 1(> approximately equal cells usually retaining the fertilization membrane (Figs. 49, 50). Whether the quarter-eggs could give rise to swimming plutei will be investigated further. This part of the work was done for a short period late in September when the eggs are not in the best condition. The granular quarters went as far as the stage just before they become free-swimming and looked quite normal at that time. The recentrifuged half-eggs which have become dumb-bell shaped \-"\'^. 51 ) retain their shape if fertilized immediately, and round up if left unfertilized just as the whole eggs do. The rounded eggs develop in the same way as before recentrifuging. The dumb-bell shaped half-eggs form a fertilization membrane following their contour and the nuclear phenomena accompanying fertilization can be clearly seen 1 ;ii;s. 52-54). The first cleavage usually comes in near the junction of the clear with the granular area, giving one clear cell with oil cap and one granular cell (Fig. 55). The granular cell often precedes the clear cell in division just as it does when completely separate. The fertil- ization membrane usually breaks after several divisions, giving a loose mass of cells, some clear and some granular (Figs. 56-59). Micro-dissection The difference in the material of the white and red half-eggs can be well demonstrated by micro-dissection. When the colorless halt eggs are punctured by a needle they immediately explode, the granules and nucleus flowing out and leaving the membrane empty. When the I lamented half-eggs are punctured, there is no How of granules, the material is quite pliable and elastic; it can be pulled out in strands which will go back again and resume a spherical form, or it can be cut in parts, each of which may round up. The stratified whole egg responds in a similar way. When the clear zone is punctured, the granules How out. When the yolk or red layers are punctured, the material can be pulled out and re'u a>e<| without any explosion or loss of material, and it behaves like an elastic and pliable substance. When the clear quartet -e^ i> punctured, it explodes immediately. When the granular quarter is punctured, the granules flow out but quite >lo\\ 1\ . Parthenogenesis The question of parthenogenesis in the half- and quarter-eggs has been studied only slightly. Just (1928), and others previously, found that by treating unfertilized Arbacin eggs for a few seconds with DEVELOPMENT OF HALF AND QUARTER ARHAriA EGGS 163 distilled water and then returning them to sea water, the eggs formed fertilization membranes and developed to the stage just before cleavage. When the half- and quarter-eggs are thus treated, they all form beautiful fertilization membranes and good ectoplasmic layers. In the centrifuged whole eggs and in colorless half-eggs, the nucleus descends toward the center of the egg just as in fertilized eggs; this then is not an attraction by the male pronuncleus. The nucleus enlarges and breaks just as in fertilized eggs. Astral radiations characteristic of fertilized eggs at the time of union of the pronuclei are, of course, absent, but the astral radiations from the amphiaster are later seen in the granular zone. In the pigmented half-eggs, no development further than the formation of the fertilization membrane and ectoplasmic layer has been observed, nor would it be expected since there is no nucleus of any sort. The clear quarters start to develop just as the colorless half-eggs, as indicated by the descent of the nucleus to the center. The granular quarters, like the pigmented half- eggs, show no further development after the formation of the fertil- ization membrane and ectoplasmic layer. SUMMARY 1. With strong centrifugal force and the proper medium, Arbacia eggs can be separated into two half-eggs, one colorless containing oil, nucleus, clear layer, fifth (granular) layer and a little yolk; the other slightly smaller containing yolk and pigment. \Vith greater centrifugal force, the colorless half-eggs can be separated into quarter-eggs, one perfectly clear with oil and nucleus; the other, smaller, with fifth layer (granules) and yolk. All of these half- and quarter-eggs can be fertilized, form fertilization membranes and cleave. 2. Nuclear phenomena accompanying fertilization and cleavage, quite normal, can be observed with great clearness in the colorless half-eggs. Astral rays occur only where granules are present. These half-eggs cleave regularly and form swimming blastulae and plutei, normal except for color and size. 3. The pigmented half-eggs develop with only the male nucleus which divides repeatedly, usually without cell division. Some blastulae and a few plutei developed but these eggs and larvae are not very viable. 4. Whole eggs, centrifuged till dumb-bell shaped, retain their shape if fertilized immediately, even if the fertilization membrane is removed. The first cleavage in elongate eggs is usually parallel with the stratifi- cation, in spherical eggs it is usually perpendicular to it. Slipper shaped blastulae develop from the elongate eggs, and normal plutei. 164 ETHEL BROWNE HARVEY 5. Clear quarter-eggs begin to cleave very slowly (after 7 hours), and usually form loose clusters of cells owing to the breaking of the fertilization membrane. 6. Granular quarter-eggs develop with only the male nucleus and a little more slowly than the normal whole eggs; cleavage is quite regular but no swimming blastulae were obtained. 7. The pigmented half-eggs can be drawn out with a micro- disM-rtion needle, and the material is pliable and elastic; the colorless half-eggs explode wrhen punctured, pouring out granules. The clear quarter-eggs collapse immediately when punctured, and the granular quarters pour out their granules slowly. 8. All of the half- and quartcr-rui^ will start to develop partheno- vriirtically, i.e., throw off a fertilization membrane, if treated with distilled water. Only those with a nucleus develop further, till just before cleavage. LITERATURE LIST HAKM.KV, M. IL, \f>l(i. Localization of the Micromere Material in the Cytoplasm of the Egg of Arbacia. Jour. Exper. Zool., 45: 319. II \K\ KY, E. X., 1931. The Tension at the Surface of Marine Eggs, especially those of the Sea Urchin, Arbacia. Biol. Bull., 61: 273. HORSTADIUS, S., 192S. Uber die Determination des Keimes bei Echinodermen. . 1 i In Zoologica, 9:1. JUST, E. E., 1928. Initiation of I U-vdnpment in Arbacia. IV. pp. 1-191. Pruto- /'litsnin, 5: 97-126. LYON, E. P., 1907. Results of Centrifugalizing Eggs. Arch. /. Enlw. Afech., 23: 151-173. MORGAN, T. H., AND E. P. LYON, 1907. The Relation of the Substances of the Egg, Separated by a Strong Centrifugal Force, to the Location of the Embryo. Arch.f. Entw. Mech.,24: 147-159. PLOUGH, H. H., 1929. Determination of Skeleton-forming Material at the Time of the First Cleavage in the Eggs of Echinus and Paracentrotus. Arch. f. Entw. Mech., 115: 380-395. TENNENT, D. H., C. V. TAYLOR, AND D. M. WHITAKICR, 1929. An Investigation on < trganization in a Sea Urchin Egg. Carnegie Inst. Pub. Wash., Xo. 391, pp. 1-104. WHITAKER, D. M., 1929. Cleavage Rates in Fragments of Centrifuged Arbacia Eggs. Biol. Bull, 57: 159-171. WILSON, E. B.( 1895. An Atlas of the Fertilization and Karyokinesis of the Ovum. Xew York. DESCRIPTION OF PLATES The drawings have been made from living eggs entirely, and are magnified 266 X. The large solid dots represent pigment, the coarse stippling yolk granules, line stippling the granules of the fifth layer, small circles oil drops. The times given limes after fertilization, approximate for 23° C. (controls cleave in 50 minutes). FlGS. 1 12. Colorless half-eggs; Figs. 13-24 pigmented half-eggs; Figs. 25-35 iitu^i-'l whole eggs; Figs. 36, 37 normal whole eggs; Figs. 38-45 clear quarter- I igs. \<> 50 granular quarters; Figs. 51-59 recentrifuged colorless half-eggs. DEVELOPMENT OF HALF AND QUARTER ARBACIA EGGS 165 80 min. 5 hr. 3 hr. 24 166 ETHEL BROWNE HARVEY 29 ^-^ 30 35 ml-n 38 min. 6 hr. 25 min. DEVELOPMENT OF HALF AND QUARTER ARBACIA EGGS 167 59 2 hr. hr. 57 3 br. 68 4 far. MIGRATION OF THE PROXIMAL RETINAL PIGMENT IN THE CRAYFISH IN RELATION TO OXYGEN DEFICIENCY kl'DOLF BENNITT AND AMANDA DICKSON MERRICK ZOOLOGICAL LABORATORY, UNIVERSITY OF MISSOURI The primary factors determining the position of the pigment in the proximal and distal sets of pigmented cells of the crustacean retina l;ig. 1) are light and darkness. Whether they act directly on the cells or indirectly via the circulation or the nervous system has not been definitely determined. But ever since Exner (1891) called attention to the optical importance of ensheathing the visual axis (Fig. 1, rhabdome and cone) of the ommatidium with opaque pigment in the light and of withdrawing this sheath in darkness, thereby allowing the photoreceptor (the rhabdome) to make use of light other than that which enters along its axis, most investigators have regarded this as a more or less adaptive mechanism, though of somewhat doubtful effectiveness. It became apparent, however, that under some circumstances factors other than light and darkness might induce changes in the proximal and distal sets of pigmented cells of the crustacean retina. For example, Congdon (1907) found that temperature was such a factor in certain Crustacea. Demoll (1911) and Bennitt (1924) found that anaesthesia and death, in noctuid moths and crustaceans re- spectively, were always accompanied by the extreme "light-adapted" position of the retinal pigment, regardless of the surroundings. Welsh (1930a) recently noticed that anaesthesia had this effect on the distal retinal pigment of the shrimp Macrobrachium. He further observed a "diurnal rhythm" in the migrations of the distal pigment cells in the retina of this animal, even under constant illumination, and this he believed to be associated with the same sort of "metabolic periodicity' to which had been ascribed numerous other cases of periodic change- in color, luminescence, etc. Bennitt (19326) observed a similar, though less extensive, diurnal rhythm in the proximal pigment cells of the retina in crayfishes which were kept for several days in total darkness. Finally, Bennitt (1924, 1932a) obtained evidence that under certain conditions stimulation of one eye might result in bodily changes affecting the position of the proximal pigment in the other (unstimu- lated) eye; whether these changes are nervous or vascular or both is still undecided. 168 MIGRATION PROXIMAL RETINAL Pir.MKNT 169 It would seem, therefore, that in addition to the usual photic changes associated with the action of the proximal pigmented cells of the retina there is an internal mechanism of some sort whose operation may, under certain conditions, bring about changes which were at first thought to be solely photomechanical. The experience of Demoll, Bennitt, and Welsh with anaesthetics, referred to above; Arey's (1916) discovery that oxygen deficiency was associated with retraction of the L D CUT HYP. CM. DST. RTN. NL.PRX PRX.RTN RHB. ACC; B.MB. M. FBR, FIG. 1. Longitudinal sections of two ommatidia of Cambarus mrilis Hagen, showing the arrangement of retinal pigment in the light-adapted (L) and dark- adapted (D) conditions. Only two of the seven proximal retinal pigment cells are shown in each ommatidium. ACC., nuclei of accessory pigment cells; B.MB., basement membrane; CN., cone; CUT., cuticle; DST.RTN., distal retinal pigment cells; HYP., cells of the hypodermis; N.FBR., prolongations of the proximal retinal cells which form nerve-fibers into the optic ganglia; NL.PRX., nuclei of proximal retinal pigment cells; PRX.RTN., proximal retinal pigment cells; RHB., rhabdome. X 125. 170 R. BKNMTT ANP A. I). MERRICK retinal pigment in fishes; the work of Spaeth (1913) on the melano- phores of fishes and of t'yeno (1922) on those of the frog, both of which showed distinct pigmentary changes accompanying oxygen want; and some early observations of the writers and others on the effect of overcrowding crayfish in an aquarium— all these led us to investigate the specific effect of oxygen deficiency on the action of the retinal pigment cells of the crayfish. We limited this investigation to observations on the proximal pigment cells, in which pigment streams dist, illy in the light and proximally in darkness through the cytoplasm of cells of virtually fixed length (cf. Fig. 1 and Plate 1). Photo- mechanical movements of the distal cells have been described in the shrimp Palaemonetes by \Yelsh (19306). When a crayfish is "light-adapted" (Fig. 1, L; Plate 1, L), anaes- thesia or death evoke no change in the position of the proximal pigment. When, however, the crayfish is "dark-adapted" (Fig. 1, D\ Plate 1, D), death, anaesthesia, or simply overcrowding in the aquarium induce distal migration of the proximal retinal pigment even though the animal is in total darkness throughout the experiment. GENERAL METHODS The crayfishes, Cambarus virilis Hagen and C. clarkii Girard, were killed in hot water to fix the position of the pigment, and were ex- amined: (1) after sectioning in paraffin, or (2) after macerating for 18 hours or more in Bela Haller's fluid. The latter method was simpler, and was equally reliable for the study of proximal cells, especially since the rhabdomes and accessory pigment cells usually adhered to the dissociated proximal pigment cells. Oxygen and nitrogen were taken from commercial cylinders. Carbon dioxide was taken first from a Kipp generator, later from a riMnnuTcial cylinder; no difference in results was detected. Oxygen determinations were made by the Winkler method, modified after Kemmerer, Bovard, and Boorman (1923). Free carbon dioxide was determined by the method of Seyler (1894). The pH was determined colorimetrically by the use of the ( 'lark and Lubs standard indicators bromthymol blue, bromcresol purple, and cresol red. The temperature of the water was taken with each sample, and the variation was found to be within 3° C. for any given experiment. The exact position of the proximal retinal pigment was recorded in every case, but for convenience in tabulation three general positions are used in this paper: (1) "Dark" -pigment not distal to the bases of the rhabdomes (Fig. 1, 1); Plate 1, D); (2) "intermediate" —pigment f-.teiidinc; along the rhabdomes but not distal to them; (3) "light" pigment extruding beyond the distal ends of the rhabdomes, i.e.. PLATE I Photographs of unstained sections of the compound eyes of Cambarus virilis, in the light-adapted (L) and dark-adapted (D) conditions. Only a part of each retina is shown, and only the proximal components of the ommatidia (cf. Fig. 1), viz. : acc., accessory pigment layer; b.mb., basement membrane; prx.rln., proximal retinal pigment cells, showing the nuclei at their distal ends, and extending proximally through the basement membrane; rhh., rhabdomes, visible in (D) as the crenulated areas among the proximal retinal cells, but obscured in (L) by the retinal pigment. X 125. 13 MIGRATION PROXIMAL RETINAL PIGMENT nearly or quite to the tips of the proximal pigment cells (Fig. 1, Plate 1. n Plate 1, L). EXPERIMENTS General I \ljcct of Excess Carbon Dioxide Preliminary experiments were performed in this laboratory by Air. L. M. Schmidt. He did not determine the oxygen content or carbon dioxide content of the water, but he found that exposure to an excess of carbon dioxide (1) inhibited proximal migration of the proximal retinal pigment when light-adapted crayfish were placed in the dark and (2) brought about distal migration of this pigment in dark-adapted animals which were kept in darkness. His method was TABLE I General effect of excess carbon dioxide on the position of the proximal retinal pigment —prelim inary experiments. Number of eyes examined Dark Intermediate Light Controls — killed at end of experiment after 6 hours in darkness 12 0 2* 0 0 IS Light-adapted animals, placed in darkness for 6 hours in dishes through which CO-; bubbled at rate of 2 liters per hour Controls — killed in darkness at beginning of exposure to CO2 7 0 6* 0 0 34 Dark-adapted animals, kept in darkness for 6 hours thereafter in dishes through which CO2 bubbled at rate of 2 liters per hour " The observed variation in pigment-position in the proximal cells of dark- adapted eyes was probably due to the fact that some of the controls were killed during the day, others at night. Bennitt (19-526) found evidence of diurnal variation in pigment position, even when the crayfish were in continuous darkness. to bubble carbon dioxide (2 liters per hour) for six hours through water containing a number of crayfish, keeping other animals in running water as controls. All his animals were alive and were moving their swimmerets at the end of the six-hour period. Schmidt's results appear in Table I. Relation between Position of Proximal Retinal Pigment, Carbon Dioxide Content, Oxygen Content, and Hydrogen Ion Concentration Carbon dioxide was bubbled through about five inches of water in a 12 x 18 inch aquarium at the rate of 3 liters per hour. Eight or ten dark-adapted crayfish in the aquarium were prevented from rising 172 R. BKNMTT AND A. I). MKRRICK to the surface by a layer of sheet cork. The apparatus was kept in darkness, and at intervals of one or two hours animals were killed and water samples were taken for determination of pH, oxygen content, and carbon dioxide content. The experiments involved 42 animals (84 eyes); 21 animals (the "dark controls") were killed at the beginning of the exposure to carbon dioxide; the other 21 were kept in water through which carbon dioxide passed, and were in darkness throughout. The results appear in Table II, in which the figures for oxygen content, carbon dioxide content, and pH represent averages for the five experiments. TABLE II Relation between position of the proximal pigment, oxygen content, carbon dioxide (onlent, ami pi I. Carbon dioxide bubbled through the water at the rate of 3 liters per hour. Animals in darkness throughout the experiment. (* — -of. foot-note, Table 1.) • Oxyui n COZ pH Number of eyr< examined Dark Intermediate- Light Dark controls cc. /liter 3.2 0.8 0.0 0.0 cc. /liter 1.2 45.0 66.0 124.0 7.6 6.0 5.3 5.3 34 0 0 0 6* 6 0 0 2 * 12 12 12 < niitinued in darkness; CO-> for 1 hrs. at rate of 3 liters per hr. Same conditions — 4 hrs. Same conditions — 6 hrs TABLE III Relation between position of the proximal pigment, oxygen content, carbon dioxide i ontenl, and /•>//. ( 'arbon dioxide bubbled through the water at the rate of 6 liters per hour. Animals in darkness. (* — cf. foot-note, Table I.) < Kyucn CO2 i> H Number of eyes examined Dark Intermediate Light 1 'ark controls cc./lihr 3.4 0.2 ' uater-sai 0.0 cc. /liter 35.4 137.5 11 pie lost) 277.1 7.7 5.4 4.4 2 2 4 0 4* 4 2 0 0 0 0 6 Continued in darkness; CO 2 for •in min. at rate of 6 liters per hr. Same com hi ions —60 min ^.iinr conditions ')() mm A- i lie experiment progressed, the carbon dioxide content in each • ase increased while (lie pi I and the oxygen content decreased. In an ;ii tempt to determine whether or not there was any quantitative MIGRATION PROXIMAL RETINAL PIGMENT 173 relation between the rapidity of pigment migration and the rapidity of change of these physical values, carbon dioxide was next bubbled through the water at twice the former rate (i.e., 6 liters per hour), and animals and water samples were removed every thirty minutes. The results appear in Table III. All the animals in this experiment were light-adapted at the end of one and one-half hours. When carbon dioxide had been administered only half as rapidly (i.e., 3 liters per hour), this condition was not attained until between three and four hours. The rapidity of migra- tion of the proximal pigment appeared to vary in general with the rapidity of administration of carbon dioxide, though whether or not carbon dioxide itself was the active agent in the migration remained to be seen. TABLE IV Relation between position of the proximal pigment, oxygen content, carbon dioxide content, and pH. Nitrogen bubbled through the water at the rate of 2-3 liters per hour. Animals in darkness. (* — cf. footnote, Table I.) Oxygen CO2 PH Number of eyes examined Dark Intermediate Light Dark controls (beginning) cc. /liter 2.1 0.6 0.3 0.0 cc. /liter 2.7 8.0 7.7 14.2 7.5 7.4 7.4 7.3 20 0 2 0 18 6* 8 2 0 12* 10* 20 26 12 6* Continued in darkness; N2 for 2 hrs. at rate of 2-3 liters per hr . Same conditions — 4 hrs. Same conditions — -6 hrs Dark controls (end) . . ... Comparing now the physical conditions, shown in Tables II and III, which obtained when distal migration was complete in all the eyes, it appeared that while the oxygen content dropped to zero in both cases, there was little uniformity in the values for either carbon dioxide content or pH. The next step was to determine whether the pig- mentary response was associated with (1) surplus carbon dioxide, (2) oxygen deficiency, or (3) increase in hydrogen ion concentration. Effect of Oxygen Deficiency with Only a Slight Increase in Carbon Dioxide Content Instead of carbon dioxide, nitrogen was bubbled through the water at the rate of 2 to 3 liters per hour. This permitted the free oxygen to be used up as in the preceding experiments without at the same time adding artificially a large amount of carbon dioxide to that already present. The procedure was as before except that "dark controls," 174 R. BENNI1T AND A. D. MERRICK kept in running water, were killed at both the beginning and the end of each exposure to nitrogen. The results of the six experiments are summarized in Table IV. Since complete distal migration occurred in all animals killed when the CO2-content was as 1<>\\ as 14.2 cc. per liter in the experiments with nitrogen, while it did not occur at much higher concentrations of carbon dioxide in previous experiments (cf. Tables II and III), it appeared that the concentration of carbon dioxide as such was not a factor affecting pigment migration in the proximal retinal cells except iiiH>far as it might contribute to other physical changes. Moreover, -ince distal migration occurred in all cases at a pH as high as 7.3 in the TABLE V Relation between position of the proximal pigment, oxygen content, carbon dioxide content, and pll. Water acidified with IK"!. The pH-range represents readings taken at the beginning and end of each experiment. Oxygen C02 pH nin Xumber of eyes examined Dark Int.Tinciliutc Light 1 )ark controls cc. /liter 1.7 (no deterr 0.2 3.4 cc. Iliter 1.5 linations) 5.6 44.0 7.6-7.6 5.2-5.6 5.0-6.0 3.9-4.8 12 4 8 2 0 2 4 8 0 0 2 16 ( 'nut iniii'd in darkness; acidi- fied water for 2 hours Same conditions — 4 hours. . . Same conditions — 2 hours. . . experiments with nitrogen, while failing to occur at a pH as low as 6.0 in the earlier experiments (cf. Tables II and III), it would seem that the hydrogen ion concentration also was not a directly contributing factor. However, additional experiments were undertaken to settle this point. l\tjeit of Altered Ilydiogen Ion Concentration Tap-water was acidified with H('l, and dark-adapted animals were left in it in darkness for varying periods. The "dark controls," as before, were kept in running water. The results of these experiments an- listed in Table Y. Thus hydrogen ion concentration appeared to become an active factor in promoting distal migration of the proximal retinal pigment • >n]\ at acidit ies as high as pll 3-4, and even under these conditions the ci'tei t \\a> by no means consistent, as is indicated by the 10 eyes which tailed to become light-adapted. With one exception, however, the pH ot the water in all previous experiments did not go below 5, and in all MIGRATION PROXIMAL RETINAL PIGMENT 175 these cases distal migration had proceeded to completion. Hydrogen ion concentration, within the experimental range produced by the addition of carbon dioxide, did not affect appreciably the position of the pigment. DISCUSSION It should be noted that the animals were neither dead nor even completely quiescent at the end of maximum exposures to carbon dioxide, nitrogen, or acidified water. In all cases the swimmerets were moving and in many cases the mouth-parts and other appendages as well. Check experiments showed that these animals could stand the most unfavorable conditions to which they were subjected in the experiments described in this paper, and would recover when replaced in running water. The only one of the three factors tested which showed any con- sistent relation to migration of the proximal retinal pigment was oxygen content. As this decreased, the pigment began to move toward the distal position, even in darkness, and this migration was invariably completed when the oxygen tension had dropped to zero. If we are correct in supposing that in aquatic Crustacea there is a fair degree of correlation between the oxygen content of the water and the metabolic activity of the animal (as expressed by its oxygen con- sumption) it seems reasonable to believe that the pigmentary changes here observed are associated in some way with varying metabolic conditions. Evidence that the supposition is correct is afforded by Amberson, Mayerson, and Scott (1925). In a large number of experiments on the lobster (Homarus) they found a correlation between oxygen tension in the water and oxygen consumption by the animal so close as to warrant the statement that "at every instant the oxygen-consumption is directly proportional to the oxygen-tension in the sea-water at that instant" (1925, p. 175). They found an equally close correlation of the same sort in the polychaet Nereis, and prelimi- nary experiments indicated a similar condition in the horse-shoe crab (Limulus), the blue crab (Callinectes} , and the shrimp (Palaemonetes). Palaemonetes was able to keep its oxygen consumption constant until the oxygen content of the surrounding water had dropped to 50 per cent of saturation, but wrhen less oxygen than this was present, its response was similar to that of the lobster. Since distal migration of the proximal retinal pigment, though influenced primarily by the presence of light, is promoted also by low temperature, oxygen deficiency, anaesthesia, and death, it begins to look as though it might be promoted by any general factor which tends to retard metabolic activity. This idea was first presented by W. H. 176 R. I5KNMTT AND A. l>. MKRRU'K Cole (1922, p. 408) with reference to the distal movement of melano- phore pigment in frog tadpoles — an activity analogous in many respects to distal movement of the proximal retinal pigment in crustaceans. The rhythmical activity of the distal and proximal pigment cells of the crustacean retina under constant conditions of light or darkness (Welsh, 1930^; Bennitt, 19326) is a condition which scarcely admits of interpretation except on the basis of daily changes within the animal's body. It is probable that such rhythmical activity as well may ultimately be explained on the basis of variations in metabolic activity. SUMMARY 1 . Overcrowding of crayfish in an aquarium often results in distal migration of the proximal retinal pigment while the animals remain in darkness. 2. Bubbling of carbon dioxide through the water surrounding the crayfish has the same effect. 3. When carbon dioxide is passed through the water at the rate of 3 liters per hour, distal migration of this pigment is complete at an oxygen content of 0.0 cc. per liter, a carbon dioxide content of 66.0 cc. per liter, and pH 5.3. 4. When carbon dioxide is passed through the water at t he rate of 6 liters per hour, distal migration is complete at an oxygen content of 0.0 cc. per liter, a carbon dioxide content of 277.1 cc. per liter, and pH 4.4. The speed of distal migration varies in general with the speed of administration of carbon dioxide. 5. When, instead of" carbon dioxide, nitrogen is passed through the water at the rate of 2-3 liters per hour, distal migration is complete at an oxygen content of 0.0 cc. per liter, a carbon dioxide content of 14.2 cc. per liter, and pH 7.3. 6. Acidification of the water with IK "I, so long as the oxygen supply remains normal, produces no discernible effect on distal migration, except in the pi I range 3.3-3.9, which is considerably below that involved in the preceding experiments. 7. The only factor shown to have a definite relation to distal migration in these experiments is oxygen deficiency. Distal migration is always complete when the oxygen content of the surrounding water reaches zero. 8. The various secondary factors promoting distal migration of the proximal retinal pigment — \\'/., low temperature, oxygen deficiency, anaesthesia, and death all tend to retard metabolic activity. The evidence indicates that distal migration of this pigment, while influ- enced primarily by the presence of light, may secondarily be induced by agents which reduce the metabolic rate. MIGRATION PROXIMAL RETINAL PIGMENT REFERENCES AMBI KSO\, \\". R., H. S. MAYERSON, AND \V. J. SCOTT, 1925. The Influence of Oxygen Tension upon Metabolic Rate in Invertebrates, four. Ccn. Physiol., 7: 171. AREY, L. B., 1916. The Movements in the Visual Cells and Retinal Pigment of the Lower Vertebrates. Jour. Comp. Neur., 26: 121. BENNITT, R., 1924. The Migration of the Retinal Pigment in Crustaceans. Jour. Exper. Zool., 40: 381. BENNITT, R., 1932a. Physiological Interrelationship in the Eyes of Decapod Crustacea. Physiol. Zool., 5: 49. BENNITT, R., 1932&. Diurnal Rhythm in the Proximal Pigment Cells of the Crayfish Retina. Physiol. Zool., 5: 65. COLE, W. H., 1922. The Transplantation of Skin in Frog Tadpoles, with Special Reference to the Adjustment of Grafts over the Eyes, and to the Local Specificity of the Integument. Jour. Exper. Zool., 35: 353. CONGDON, E. D., 1907. The Effect of Temperature on the Migration of the Retinal Pigment in Decapod Crustaceans. Jour. Exper. Zool., 4: 539. DEMOLL, R., 1911. Uber die Wanderung des Irispigments im Facettenauge. Zool. Jahrb., Abth. allg. Zool. u. Physiol. d. Tiere, 30: 169. DEMOLL, R., 1917. Die Sinnesorgane der Arthropoden — ihr Bau und ihre Funktion. Braunschweig. EXNER, S., 1891. Die Physiologie der facettirten Augen von Krebsen und Insecten. Leipzig und VVien. KEMMERER, G., J. F. BOVARD, AND W. R. BOORMAN, 1923. Northwestern Lakes of the United States: Biological and chemical studies with reference to possi- bilities in production of fish. Bull. U. S. Bur. Fish., 39: 51. PARKER, G. H., 1897. Photomechanical Changes in the Retinal Pigment Cells of Palaemonetes, and Their Relation to the Central Nervous System. Bull. Mus. Compar. Zool., 30: 273. SEYLER, C. A., 1894. Notes on Water Analysis. Chem. News, 70: 82-83, 104-105, 112-114, 140-141, 151-152 (vide Kemmerer, Bovard, and Boorman). SPAETH, R. A., 1913. The Physiology of the Chromatophores of Fishes. Jour. Exper. Zool., 15: 527. UYENO, K., 1922. Observations on the Melanophores of the Frog. Jour. Physiol., 56:348. WELSH, J. H., 1930«. Diurnal Rhythm of the Distal Pigment Cells in the Eyes of Certain Crustaceans. Proc. Nat. A cad. Sci., 16: 386. WELSH, J. H., 19306. The Mechanics of Migration of the Distal Pigment Cells in the Eyes of Palaemonetes. Jour. Exper. Zool., 56: 459. OSMOTIC PROPERTIES OF THE ERYTHROCYTE III. Tin-: APPLICABILITY OF OSMOTIC LAWS TO THE RATE OF HEMOLYM-* i\ HYPOTONIC SOLUTIONS OF NON-ELECTROLYTES M. II. JACOBS (From the Department of Physiology, rniversity of Pennsylvania, and the Marine Biological Laboratory, \\'oo(h Hole, Mtissaclmsetts) I Few subjects on the field of cellular physiology have received as much attention as that of osmotic hemolysis. Beginning with the observations of Hamburger (1886) — for a summary of the early litera- ture, see Hamburger (1902) — there have appeared literally hundreds of papers dealing with this process under all conceivable conditions of health and disease, and of age, sex, species, and previous treatment of the blood. As has been pointed out in the preceding paper of this series (Jacobs and Parpart, 1931), much of this work is of very doubtful \alue bec_ause of the neglect of certain essential experimental pre- cautions. Entirely apart from this defect, however, the work is disappointing in its almost complete failure to deal with the question of the rate at which such hemolytic processes occur. Unlike the published work upon other types of hemolysis (saponin, bile salts, specific sera, etc.), that upon osmotic hemolysis with few exceptions has had to do merely with the end state finally reached by the system. Information about this point, though possessing a limited practical value in medicine, is, on the whole, of comparatively little theoretical interest. So-called "fragility" studies — at least as ordinarily carried out — are concerned with the specific properties of the erythrocyte alone and they throw little light upon more fundamental problems of cell physiology. ( hi the other hand, a study of the rate of osmotic hemo- lysis, which is closely associated \\ ith the rate of entrance of water into the cell, has obvious applications to many important general problems in the fields of OMIIOSI^, cell permeability, etc. To all investigators who have tried in the usual way to measure the rate of osmotic hemnlysis the reason for the almost complete neglect of this Meld in the past is dear. Osmotic hemolysis, when it goes to completion, as, for example, in distilled water or in very strongly hypotoiiic solutions, is such a rapid process — requiring for its entire course perhaps only a few seconds — that the ordinary methods of 178 OSMOTIC HEMOLYSIS 179 studying hemolytic phenomena cannot be successfully employed with it. On the other hand, when it is sufficiently slow to permit ready measurement, especially when the end-point is some partial degree of hemolysis, the results obtained are likely to be so variable and irregular, and at first sight so generally inexplicable, that most persons who have tried to work under these conditions have soon abandoned their attempts. The author has pointed out elsewhere (Jacobs, 1927, 1928, 1931; Jacobs and Parpart, 1931) that while irregularities in the behavior of the erythrocyte can be minimized by a strict standard- ization of the experimental procedure, they are in part inherent in the nature of the material itself and are therefore unavoidable. The difficulty, in brief, is that factors such as temperature, pH, etc. which might be expected to affect the rate at which a given equilibrium condition is attained, have an unusually strong tendency in the erythrocyte to change at the same time the position of the equilibrium itself. Under these conditions the results are, in general, too compli- cated for ready analysis; and the experimenter is of necessity driven back to the other horn of the dilemma where the difficulty is with the rapidity of the process and therefore with the method rather than with the material. Fortunately, the simple method of the author (Jacobs, 1930) for the study of hemolysis proves to be adequate for a fairly accurate determi- nation of times of hemolysis greater than approximately 1.2 seconds and permits experiments on the rate of the process to be made under conditions where disturbing equilibrium factors are of negligible importance, namely, in distilled water and in very strongly hypotonic solutions. The method therefore opens to experimental study an important field which has heretofore been almost wholly neglected. This general field will be dealt with in the present and in several succeeding papers. A very fundamental question, which must first be decided before other work can be undertaken with profit, is how far the rate of hemo- lysis in hypotonic solutions may be considered to depend upon the rate of entrance of water into the erythrocyte in accordance with simple osmotic laws. In an earlier paper by the author (Jacobs, 1927) it was tentatively assumed that the erythrocyte behaves as a simple osmom- eter and gives up its hemoglobin to the surrounding solution when a certain critical hemolytic volume, Vh, is reached. On this assumption equations were derived for the calculation of permeability constants for water from data on the rate of osmotic hemolysis. It was, however, emphasized in another place (Jacobs, 1931) that osmotic hemolysis is in reality a fairly complicated process involving (a) the entrance into 180 M. II. JACOBS the cell of water, (b) the escape of hemoglobin, (r) the possible escape of salts and other osmotically active materials and (d) changes pro- duced in various other ways in what is commonly loosely spoken of as the osmotic reliance of the cell. Only where factors b, c and d can be shown to be ol negligible importance is it permissible to use the simple method of treating the subject previously employed; and a more critical examin.it ion of this point is therefore highly desirable. A further need for such an examination is created by the recent work of Ponder and Saslow (1931), who have given reasons for doubting the applicability of simple osmotic laws to the erythrocyte because of the leakage from the cell during the course of at least certain types of experiments of osmotically active materials (factor c mentioned above). It must be eidmitted that in cases where such leakage occurs to any considerable extent a simple treatment of the problem is impossible. However, in view of the fact that Ponder and Saslow dealt primarily with final equilibria, arrived at in the course of a considerable time, it is by no means certain that in hemolytic experiments whose duration is only a few seconds such leakage as they have described would be a disturbing factor, though it is not impossible that it might. The question can be settled only by experiment, preferably by a comparison of the observed rates of osmotic hemolysis with those deduced ac- cording to the theory that the erythrocyte behaves as an ideal osmom- eter. Such a comparison will now be made. II The equations given in the earlier paper (Jacobs, 1927) for relating the time of hemolysis to the concentration of the medium are somewhat inconvenient because they employ the initial volume, V0, and the hemolytic volume, F/,, of the cell. Though it has recently been shown by Ponder and Saslow (1931) that the idea of a hemolytic volume has actual experimental justification, there is, in general, a certain ambiguity in working with volumes. This is due to the fact that the volume that enters into osmotic equations is not the measured volume of the cell but rather that of the water which the cell contains. In some cases, e.g., the cells of the plant Tradescantia (Hofler, 1917), this distinction is unimportant, but in the erythrocyte, which is loaded with hemoglobin to an extent of over 30 per cent by weight, it un- doubtedly is — though, unfortunately, there is little agreement among different workers as to the actual magnitude of the true volume. For many purposes it is perhaps best to use the weight of the water in the cell as determined by chemical analysis (Van Slyke, \Yu and McLean, 1(>23, etc.), but since under ordinary experimental conditions the OSMOTIC HEMOLYSIS 181 weight of the water and what might be called the osmotically effective volume of the cell, i.e., that part of the total volume which takes part in osmotic changes, are approximately related in a very simple manner, it is perhaps permissible for more ready comparison with other published work in this field to retain for the present purposes the older type of osmotic equations involving volumes and concentrations. Assuming with Hill (1930) that the water within the cell is almost entirely "free," i.e., capable of taking part in osmotic equilibria, or at least that there are no marked changes during the course of the experiment in the degree of "binding" of water by cell constituents, we have the relation: cV = c0V0, where c0 and V0 are the initial osmolar concentration and osmotically effective volume of the cell, respectively, and c and V are any other corresponding pair of these variables. The hemolytic volume, F*, used in previous discussions may therefore be expressed in terms of constants and of the more convenient hemolytic concentration : Making certain necessary and probably well justified simplifying assumptions as to the nature of the diffusion of water across the membrane of the erythrocyte (see in this connection Northrop, 1927; Lucke, Hartline and McCutcheon, 1931; Jacobs and Stewart, 1932), it may be predicted that the rate at which it will enter the cell, i.e., the rate of increase of the cell volume, will at any given instant be pro- portional to the difference in the osmotic pressures, and therefore to that of the concentrations of the internal and external solutions, and to the extent of surface of the cell. kA(c-Q. (1) The external concentration, C, may be considered to be constant since the volume of the surrounding solution is very large as compared with that of the suspended cells (approximately 1000 : 1 in these experiments). Since V= V/-°-, c> we may write equation (1) in the form: dc 182 M. II. JACOBS Fortunately in the erythrocyte, because of its peculiar biconcave shape, a considerable degree of increase in volume is possible without any change in surface. No great error will result, therefore, if A be treated as if it were constant, and this simplifying assumption permits equation (2) to be integrated at once after separating the variables; that is: kAl -- - f0r0 I -~r J c\c - Q Remembering that when I ••= Q, c - r0, the initial isotonic concen- tration for blood, we finally obtain: ' r"7M££?_I' cC ' ' '" I kAl = —p^-ln - — 7=, H — TT- - • (3) C2 <•<•„ -- c0C C \c0 c I For the special case where the cells swell in distilled water, and C in equation (2) is therefore equal to zero, a simpler equation results, namely, As has already been stated, I'n, the initial osmotically effective volume of the cell, is not exactly known but is, in any case, a constant. I or our present purposes, therefore, V0 and A, which has also been treated as a constant, may be incorporated with the true permeability constant, k, to give a quantity, k' , whose constancy over a range of concentrations would equally well furnish a proof of the correctness of equations (3) and (4). Since we wish to use the equations only for the point at which hemolysis occurs, namely //,, o,, we substitute these particular values for I and c, respectively, and also for convenience represent the ratio r0/o, by R, giving finally: , / Oi , ' - C R • 1 , *'*==C*In^ AV — and -l). (6) If, therefore, the erythrocyte behaves as a simple osmometer, equations (5) and (6) should yield the same value of k' for all values of C including i lie value of y.cro when distilled water is employed. Ill To test the applicability of the theoretical equations derived in the preceding section to osmotic hemolvsis, experiments were performed upon erythrocytes by the method mentioned above (Jacobs, 1930). In the present paper only a single one of the earlier experiments with OSMOTIC HEMOLYSIS 183 saccharose solutions will he described in detail ; but it may be mentioned that essentially the same results have been obtained in a considerable number of other experiments both with this substance and with dextrose. Some of the later confirmatory experiments were performed by A. K. Parpart, whose careful assistance is gratefully acknowledged. The blood used for the experiment here described was that of the ox, defibrinated immediately after its collection. Because of certain abnormalities that seem to develop when erythrocytes stand for some time in contact with protein-free salt solutions (Kerr, 1929), the cells were not "washed" but were kept in the approximately normal surroundings furnished by their own serum up to the instant of ex- posure to the hemolytic solutions. It should be noted that previous washing in isotonic non-electrolyte solutions is also contraindicated by the tendency shown by erythrocytes to become agglutinated in such TABLE I Times of hemolysis of ox erythrocyte in water and in saccharose solutions. is assumed to have a value of 2.1. R Concentration Freezing point depression Observed time of hemolysis in seconds k' 0.00 0.00 1.40 2.10 0.10 0.0186 1.53 2.02 0.02 0.0373 1.60 2.05 0.04 0.0747 1.65 2.26 0.06 0.1123 1.73 2.49 0.08 0.1501 1.90 2.70 0.10 0.1880 2.15 3.00 0.12 0.2261 2.50 3.55 0.13 0.2452 2.75 4.12 0.14 0.2643 3.53 4.77 0.145 0.2739 7.33 3.88 solutions. Though the procedure that was of necessity followed resulted in the introduction into the non-electrolyte solutions of slight traces of electrolytes and of proteins, these were small since the dilution of the blood employed was approximately 500 : 1 and that of the serum therefore of the order of 1000 : 1, giving a final concentration of electrolytes in the vicinity of M/6000. With each solution four determinations of the time of hemolysis were made to the nearest tenth of a second. These figures have been averaged to give the times listed in the third column of Table I. As a rule, the individual observations in each group of four varied by only one or two-tenths of a second. Only with the highest concentrations employed, where the complicating factors previously mentioned are 184 M. II. JACOBS present, did the observations fail to show a high degree of reproduci- hility. The concentrations of the solutions employed are given in the first column of the table, but for purposes of calculation the freezing point depressions were used as being more nearly proportional to the osmotic pressures than the concentrations. These were calculated by the empirical equation: A '•• 1.86C -- 0.2C2, which tits very closely the data given in the "International Critical Tables" for saccharose over the range of concentrations employed. The critical concentration for 75 per cent hemolysis, which is in practice a convenient end-point to use, was directly determined as 0.148AI with a calculated freezing point depression of 0.280°. Since the freezing point depression for ox serum is in the vicinity of 0.58° (Hamburger, 1902) R, the ratio of r0 to ch may be taken as approxi- mately 2.1. From these data and from the observed rates of hemolysis, Table I has been prepared. An inspection of the figures in the first and third columns of this table, or, better, of the positions of the solid circles in Fig. 1, shows the relation between the observed times of hemolysis and the concentrations of the external solutions. It will be noted that an increase in concentration from zero (distilled water) to 0.12M or 0.13M has only a relatively slight retarding effect. The retardation thru increases at a rapid rate and at a concentration of 0.148M reaches infinity. This type of curve, characterized by the relative suddenness of its final rise, has always been obtained with non-electrolytes, but, as will be shown in a later paper, not with electrolytes. How far does it agree with the theoretical predictions made by the use of equations (5) and (6)? The answer to this question may be presented in two ways. In the first place, in the last column of Table I are found the calculated values of k' for the various concentrations employed. It will be noted that while there is good agreement between the values for water and for 0.01M and 0.02M saccharose, this agreement soon disappears and the last values of k' are about 100 per cent greater than the first ones. This lack of agreement is almost certainly not the result of fortuitous errors of observation, since the drift from smaller to larger values of k' i- a very regular one. Evidently the predictions made from the theoretical equations depart rather widely from the observed data. The same thing is shown still more clearly in Fig. 1 where the curve Libeled 2.1 represents the times at which hemolysis ought theoretically to occur if the first few values of k' applied throughout the concen- OSMOTIC HEMOLVSIS 185 tration range instead of increasing as they do with increasing concen- tration. It will again be noted that the agreement between theory and observation is very poor. More specifically, the observed values increase much too slowly until the highest concentrations are reached, and then they tend to increase very rapidly. It would appear, therefore, either that the rate of hemolysis is not governed in any very simple manner by osmotic laws or that some additional factor of importance has been overlooked in making the calculations. T 1.7 .2 04 .06 FIG. 1. Observed times for hemolysis of ox erythrocytes in distilled water and in saccharose solutions of different concentrations represented by the solid circles. Calculated times for different values of R represented by the curves. Ordinates represent times; abscissae, concentrations in mols per liter. IV A careful examination of the data suggests that the reason for the discrepancy between observation and theory lies not so much in the inapplicability of simple osmotic laws to the hemolytic process as in a serious error in obtaining experimentally the value of the critical hemolytic concentration, ch. This error, in turn, is responsible for one in the value of R which appears in equations (5) and (6). An in- spection of these equations, or even better, the actual substitution in them of several different values of R, all other figures remaining the same, shows that the effect of this constant on the calculated values of k' is very great and that even a small error in its determination must have serious effects. Now it might seem that of all the values entering into the calculations that of o,, and consequently also that of R, are the most reliable, since the concentration of the solution in which a given degree of hemolysis — for example, 75 per cent — is finally 14 186 M. II. JACOBS reached can he determined by direct observation with a very high degree of accuracy. But before using for purposes of calculation values of o, and of R as so determined account must be taken of a very suspicion- circumstance, namely, that the osmotic pressure of the solution of sucrose which ultimately gives 75 per cent hemolysis is very different from those of diluted serum or of NaCl having the same hrmolytic el't'ci i . For example, in four separate experiments to be described in a later paper, it was found that the unbuffered solutions of NaCl in which 75 per cent hemolysis was ultimately attained with a dilution ot blood of approximately 1 : 500 had concentrations of 0.097M, 0.093M, 0.104M and 0.098M, with an average value of 0.098M. The freezing point depression of such a solution is approximately 0.340°, giving a value of R of 1.7 instead of 2.1 as determined above. In other words, erythro- cytes in solutions of saccharose and, as may readily be shown, in solutions of other non-electrolytes also, have a higher resistance than in solutions of electrolytes or in diluted serum. This effect has been noted by many workers and has been variously explained. Leaving undetermined for the present the exact mechanism by which it is produced we may accept it as a known fact and consider some of its possible consequences. When an erythrocyte is placed in a non-electrolyte solution it possesses certain osmotic properties which are changed by exposure to its new surroundings. The critical hemolytic concentration as actually determined in such solutions is, therefore, obviously that of a cell whose properties have been modified and not that which the unaltered erythrocyte ought theoretically to show. But suppose, as seems likely, that this change is not an instantaneous one, but requires an appreciable time, say, 10 seconds, for its completion. Evidently under these conditions very rapid hemolytic processes requiring only two or three seconds would be finished before much change in the erythrocyte could occur; on the other hand, if the duration ot the hemolytic process were 15 or _'() seconds, it would be sufficiently slow to permit the change in this case an increase in the resistance of the erythrocyte — to take place. Indeed, a point would rather suddenly be reached where the increased resistance developed by the erythrocyte would be sulli( lent to prevent completely the hemolysis that would otherwise occur; and at this point there would be a sudden rise of the time-of-hemolysis curve to infinity. As has been noted abo\e, the curve expressing the observed relation between concentration and time <>l hemolysis shows as its most striking peculiarity a sudden rise as the highest concentrations are OSMOTIC HEMOLYSIS 187 approached. By far the simplest explanation of this peculiarity- indeed the only one that has so far occurred to the author — is that the increased resistance which is known to be brought about in non- electrolyte solutions does not develop instantly but requires for its appearance several seconds — almost certainly more than two or three. If this rather plausible view be accepted, then it is obviously erroneous to use the observed value of ck in calculations involving water and very dilute solutions in which hemolysis occurs in from 1.4 to 2 or 3 seconds, i.e., before the original osmotic properties of the erythrocyte have been greatly altered. Some other higher value of o» would evidently govern the behavior of the erythrocyte in such media. Let us assume as a first approximation that the true value of Ch is that determined by the use of NaCl rather than by sugar solutions. From the figures given above, the new value of R would be in the vicinity of 1.7 (i.e., 0.58 -h 0.340). Using this figure as preferable to the old one of 2.1 there have been calculated the new values of k' in column 2 of Table II. TABLE II Values of k' calculated from the data of Table I, assuming R = 1.7 and R = 1.4. Concentration W = 1.7 RW = 1.4 Concentration R<" = 1.7 RW = 1.4 0.00 1.16 0.59 0.08 1.34 0.63 0.01 1.11 0.51 0.10 1.39 0.64 0.02 1.12 0.56 0.12 1.46 0.63 0.04 1.20 0.59 0.13 1.48 0.62 0.06 1.29 0.62 0.14 1.32 0.53 It is immediately apparent on examination of this table (see also the curve labeled 1.7 in Fig. 1) that the agreement between observation and theory, as indicated by the relative constancy of k', is now con- siderably better than befoie, though there is still a slow drift in the constant, which can scarcely be accounted for by experimental errors. It is to be noted, however, that the critical concentration as inferred from NaCl experiments is itself probably too low. According to Ponder and Saslow (1931), in experiments whose duration is of the order of magnitude of the time used to determine this figure (1 hour), there is a change, interpreted by them as due to an escape of salts from the erythrocyte, which is sufficient to influence the cell-volume and which would undoubtedly render hemolysis by hypotonic solutions more difficult than otherwise. If this conclusion be accepted, then the observed critical concentration, ch, is still too low and the assumed value of R of 1.7 is too high. Though it is impossible at present to be certain what further 188 M. II. JACOBS correction is justified, it may be of interest to assume for the unaltered cell a value of R of 1.4. Calculations made by using this figure are given in column 3 of Table II. It will be noted that the value of k' is now almost constant, indicating an agreement of theory and obser- vation, up to a concentration of 0.13M. This agreement is even more strikingly shown in Fig. 1 where the curve labeled 1.4 has been calcu- lated for ilu- \alue of R by means of equations (5) and (6), starting with 1 .40 seconds as the time required for hemolysis in distilled water. An e\en better agreement could be obtained by taking a slightly lower value of R; but, in view of the simplifying assumptions used in deriving the equations, it is questionable whether the almost perfect lit that could be secured in this way has any very great significance. The important thing is that by assigning a not-improbable value to Ch (the theoretical hemolytic concentration for the unaltered cell) , the behavior of the erythrocyte over a wide range of concentrations shows a good agreement with simple osmotic laws, and its deviation from these laws at very high concentrations can be plausibly accounted for. V Up to this point the increased osmotic resistance of the erythrocyte in solutions of non-electrolytes has been accepted merely as an observed fact with no attempt at an explanation. Though for present purposes it is not strictly necessary that the cause of this peculiarity of the erythrocyte should be known, it may be noted that there are not lacking a number of more or less plausible explanations which because of their general theoretical interest may now be briefly considered. The first explanation is that non-electrolytes actually have a toughening and strengthening effect upon the cell membrane which renders the erythrocyte less susceptible to hemolysis (see for example Rhode, 1923). Although a solidifying effect of non-electrolytes upon certain gels and upon both plant and animal cells has frequently been observed 'for literature upon this subject see Hober and Memme- sheimer, 1923; and Hober, 1926), it seems very unlikely that this explanation is capable of accounting for such a remarkable increase in the osmotic reliance of the erythrocyte as is known to occur. In the first place, the membrane of this cell is so delicate that it probably offers little opposition under any conditions to volume changes (for some of the evidence see Jacobs, 1931). Even the much better developed membrane of the egg of Arbacia seems to be capable of resisting only very feebly osmotic volume changes (Lucke and McCutcheon, 1932; Harvey. 1<>31 ; Cole, 1932). It is almost incon- cei\abl<- ili.it the membrane of the erythrocyte could be so strengthened OSMOTIC IIEMOLYSIS 189 in the absence of electrolytes as to support an excess osmotic pressure of over an atmosphere (i.e., the difference in the osmotic pressures of solutions of NaCl and of saccharose, which just permit 75 per cent hemolysis to occur). Furthermore, according to Ponder and Saslow (1931) osmotic hemolysis does not necessarily involve any very appreciable stretching of the cell membrane. It seems unlikely, therefore, that a direct mechanical effect of this sort on the cell is chiefly involved. A second possibility is somewhat more plausible. It is that in a hypotonic non-electrolyte solution there is a sufficient leakage of electrolytes from the interior of the cell to lower the internal osmotic pressure and so to reduce the amount of swelling that would otherwise occur. This explanation of the increased osmotic resistance of the erythrocyte in non-electrolyte solutions has been accepted, among others, by Bang (1909) and by Ponder and Saslow (1931). The last named authors suppose "that leakage is greater in glucose than in NaCl and that this accounts for the critical volume being reached in a solution of glucose which is more hypotonic than one of NaCl." They cite in support of this view the direct chemical evidence obtained by Kerr (1929) that in solutions deficient in blood proteins there may be an escape of potassium from and an entrance of sodium into the cell. Bang (1909) also gives references to earlier work indicating a passage of electrolytes from the erythrocyte into non-electrolyte solutions, while Joel (1915) has studied this process and the influence upon it of narcotics, by an electrical conductivity method. That electrolytes may escape from the erythrocyte into non-electrolyte solutions may therefore be regarded as a well-established fact. It is very questionable, however, whether such an escape of electrolytes, which is probably associated with a loss of the normal permeability of the cell to cations, is capable of accounting for the very rapid rise of osmotic resistance that occurs in the present experiments. From the data shown graphically in Fig. 1 it would seem that a marked increase in resistance in non-electrolyte solutions must occur in less than five seconds, while even under the conditions of their experi- ments, Ponder and Saslow (1931) state that "the equilibrium volumes are attained within a minute and are maintained for hours." Since the diffusion of cations reported by Kerr (1929) is a process that seems to extend over hours, while the rise in conductivity studied by Joel went on gradually and steadily throughout experiments also lasting up to several hours, it would seem that some factor other than an outward leakage of salts (i.e., of both anions and cations) is involved in the case of very rapid changes. The factor that immediately suggests itself is a 190 M. II. JACOBS new ionic equilibrium of some sort, attained primarily by the movement of anions such as is known to occur readily in normal erythrocytes. How far the results of Ponder and Saslow with electrolyte solutions can be so explained cannot at present be stated with certainty; but, at all events, it seems likely that the extremely rapid increase in the osnii Mil- resistance of erythrocytes that takes place in non-electrolyte solutions is to be accounted for in this way. This view is supported by the work of Netter (1928), who ha> pointed out that theoretically ionic exchanges should by no means be absent between erythrocytes and a surrounding isotonic solution of a non-electrolyte, but that anions from the erythrocytes would tend to be exchanged for OH' ions from the aqueous solution in such a way as to make the interior of the cells more alkaline. This principle has been put to practical use by Bruch and Netter (1930) in obtaining various desired relations between external and internal pH values. Now it is known, especially from the work of Warburg (1922) and of Van Slyke, \Yu and McLean (1923), that a change in reaction within the erythro- cyte has important osmotic consequences. The osmotic pressure of a given amount of base bound by hemoglobin is considerably lower than that of the same amount of base bound by, for example, carbonic acid. If the compound of base with hemoglobin be represented as BnIIb, the osmotic pressure of this compound when completely dissociated would be to that of the same amount of base combined with carbonic acid as H + 1) : 2n. Anything, therefore, which causes a shift of base from hemoglobin to carbonic acid should increase the internal osmotic pressure of the cell and cause the latter to swell ; anything that causes a shift in the reverse direction should have the opposite effect. Since the new ionic equilibrium attained in non-electrolyte solutions is obviously of the latter nature, it should, without any escape of salts as such, result in a lowering of the internal osmotic pressure, and so raise the resistance of the cell to hemolysis by hypotonic solutions. If this view of the mechanism of the increase in the osmotic resistance of the erythroc\u> in non-electrolyte solutions be correct, then it ought to be possible even in solutions of electrolytes to produce the same characteristically sudden rise in the time-of-hemolysis: concentration curves described above by slightly increasing the alkalinity of the medium with a trace of NaOH or NH4OH. This is, in fad , t he case, as will be shown in a forthcoming paper by the author in coll. ilioiMtion with A. l\. I'arpart. l mentioned al»>\e, it seem> likely that while a leakage of salts of both anions and cations) is by no means excluded as a possible factor of importance in experiments of longer duration, the factor OSMOTIC HKMOLVSIS 191 cliiefly concerned in producing increased resistance under the conditions of these experiments is an entirely normal shift of anions alone, which because of certain peculiarities of hemoglobin, is secondarily responsible for a change in the internal osmotic pressure of the cell. VI It is a matter of some interest to determine the value of the true permeability constant, k, of equations (1), (3), and (4). This is a measure of the amount of water that would cross the membrane of the erythrocyte through unit area in unit time with unit difference in osmotic pressure between the cell and the surrounding solution. Such a constant would be useful for comparing the permeability to water of different cells, or of the same cell under different conditions (see Jacobs 1927, 1931). The true permeability constant, k, is found from k' of equations (5) and (6), by multiplying by V0, the initial osmotically effective volume of the cell, which may be measured in cubic micra, and dividing by A, the area of the cell, which is conveniently expressed in square micra. Unfortunately, both because of the small size of the erythrocyte and its peculiar shape, it is difficult to measure its surface with very great accuracy. This difficulty is well illustrated by the fact that the estimates made by different investigators for the erythrocytes of the same species of mammals may differ by as much as 50 per cent or more. Furthermore, even though the total volume of the cell can be fairly accurately determined by several methods (Ponder and Saslow, 1930) there has in the past been much uncertainty in calculating from it the osmotically effective volume, though it seems likely that this is in reality not very different from that of the total water contained in the cell (Hill, 1930). Finally, there is the serious difficulty, men- tioned above, that because of uncertainty as to the value of the theoretical critical hemolytic concentration for the unaltered erythro- cyte the value of k' itself is subject to a considerable error. It is evident, therefore, that the most that can be expected at present is to obtain the order of magnitude of the true permeability constant; but even this would be of considerable value. For the ox erythrocytes chiefly used in these experiments we may take as a first approximation a value of R of 1.7. This value is obtained from the observed hemolytic concentration of NaCl rather than from that of saccharose, since, as has been mentioned above, the normal osmotic properties of the erythrocyte are very quickly changed in non-electrolyte solutions. Indeed, even the value selected is probably somewhat too high, but in the absence of more complete information it is a convenient one to use. 192 M. II. JACOBS As to the necessary constants for the cell itself, the few published estimates are not in very good agreement. Probably the best available value for the volume of the ox erythrocyte is that given by Ponder and Saslow (1930) of 44 cubic micra. No estimates of surfaces are given in this paper, but in an earlier publication Ponder (1924) has given 37 cubic micra and 69 square micra as the values of the volume and surface, respectively, of the erythrocyte of the calf. If it be assumed that the shape of the somewhat larger cell is exactly the same as that 744x2/3 of the smaller one, then its surface would be 69 X I ^J or 77 square \ o7 / micra, and this value will here be used. In the absence of accurate chemical analyses of ox erythrocytes, it may tentatively, though perhaps somewhat questionably, be assumed that they contain the same percentage of water as those of man and that all of this water is "free." Taking an average of the figures given by Henderson (1928) for cc. of water in 1 liter of cells for three normal human individuals, and applying the same percentage to the ox erythrocyte wrhose total value is 44 cubic micra, we have for the initial effective osmotic volume, Fo, 0.69 " 44 or approximately 30 cubic micra. Remembering that the time for 75 per cent hemolysis in water is 1.4 seconds, we have all the data necessary to calculate from equation (6) the value of k. k ---- 0.08 XT^-X X[(1.7)2 -•!]== 0.036. i .4 i.io n The factor 0.08 has been introduced to change freezing point depres- sions in degrees centigrade into osmotic pressures in atmospheres. Expressed in words, the value of k so obtained means that with a difference in osmotic pressure of one atmosphere between the cell and its surroundings water should theoretically pass through each square micron of the cell surface at the rate of 0.042 cubic micra per second or of 2.2 cubic micra per minute. In an earlier paper (Jacobs, 1927) the value of k was estimated to be of the order of 3.0 for human erythrocytes when the unit of time was taken as one minute. Since the details of the calculation were not uiven at that time, it may be worth while to present here an additional typical set of figures, emphasi/ini; at the same time the fact that only the general order of magnitude of the results obtained from them is _nificant. In this particular case ch in terms of freezing point depressions was found to be 0.232, while r0 was taken as 0.56, giving a value of R <>l 2. I. The observed time of hemolysis was 2.4 seconds. I ollowinu I .mnions (1927) the volume of the human erythrocyte may OSMOTIC IIEMOLYSIS 193 be taken as 88 cubic micra (of which 0.69 X 88 or 61 cubic micra represents the true osmotic volume) and the surface as 145 square micra. We have, therefore: k = 0.08 X X X X [(2.4)2 _ t] = 0.060 or, if the unit of time be taken as the minute, 60 times this value or 3.6. In view of the large unavoidable errors in these calculations, due especially to uncertainty as to the exact value of R, it is questionable whether this apparently greater permeability to water of the human erythrocyte as compared with that of the ox is significant. In any case, the difference in the permeabilities of the two kinds of erythro- cytes to water is far less than is that to glycerol (Jacobs, 1927, 1931) or to erythritol and to certain ions (Mond and Gertz, 1929). SUMMARY 1. Equations are derived for predicting the relation between the time required for osmotic hemolysis and the concentration of a sur- rounding hypotonic medium. 2. It is shown that when allowance is made for certain known peculiarities of the erythrocyte the rate of hemolysis is, on the whole, in fairly good agreement with osmotic laws. 3. Reasons are given for believing that the increased osmotic resistance of the erythrocyte that develops within a few seconds in solutions of non-electrolytes is not caused by a leakage of salts from the cell but rather by a changed ionic equilibrium in which the normal impermeability of the cell to cations need not be lost. 4. Rough quantitative estimates are made of the permeability of the erythrocytes of the ox and of man to water. BIBLIOGRAPHY BANG, I., 1909. Biochem. Zeitschr., 16: 255. BRUCH, H., AND H. NETTER, 1930. Pfliiger's Arch., 225: 403. COLE, K. S., 1932. Jour. Cell. Compar. Physiol., 1: 1. EMMONS, W. F., 1927. Jour. Physiol., 64: 215. HAMBURGER, H. J., 1886. Arch. f. Physiol., p. 476. HAMBURGER, H. J., 1902. Osmotischer Druck und lonenlehre. Vol. 1, pp. 161-400. Wiesbaden. HARVEY, E. N., 1931. Biol. Bull., 61: 273. HENDERSON, L. J., 1928. Blood: A Study in General Physiology. New Haven. HILL, A. V., 1930. Proc. Roy. Soc., B., 106: 477. HOBER, R., 1926. Physikalische Chemie der Zelle und der Gewebe, 6th Ed. Leipzig. HOBER, R., AND A. MEMMESHEIMER, 1923. Pfliiger's Arch., 198: 564. HOFLER, K., 1917. Ber. dentsch. hot. Gesellsch., 35: 706. JACOBS, M. H., 1927. The Harvey Lectures, 22: 146. JACOBS, M. H., 1928. Am. Nat., 62: 289. 194 M. II. 1. \COUS JACOBS, M. H., 1930. Biol. Bull., 58: 104. JACOBS, M. H., 1931. Ergebn. d. Biol., 7: 1. JACOBS, M. H., AND A. K. PARPAKT, 1931. Biol. Bull., 60: 95. JACOBS, M. H., AND I >< >K< > i H v R. STEWART, 1932. Jour. Cell. Compar. Physiol., 1:71 JOEL, A., 1915. /' : //., 161: 5. KERR, S. E., 1<>2<>. Jour. Biol. Clicm., 85: 47. LUCRE, I'... \M> M. Me Cl rCHEON, 1932. Physiol. Rev., 12: 68. LUCRE, H.. 11. K. HAKIUM , \\i> M. McC'i n BEON, 1931. Jour. Gen. Physiol., 14: 105. MONO, K., AND H. i rERTZ, 1(>2<>. Pfl tiger's Arch., 221: 623. NETTER, H., 1928. Pfluger's Arch., 220: 107. \i)i (',. SASLOW, 1931. Jour. Physiol., 73: 267. RHODE, H., L922. Biochem. Zeitschr., 131: 560. \'AN- SLVKI . I >. !>., H. \\'r, AND }•' . eral surface (Fig. 1) they extend as five somewhat irregular bands along the lines o| the ambulacral grooves. As these bands approach the en of the animal's locomotion. Forward progress, burrowing, and righting movements of the sand-dollar are to be explained by the vigorous and well coordinated movements of these particular spines. The longest and most active spines of the anterior part of the oral side are the chief means to this end. The ambulacral feet are at best only a weak supplement to these movements. SUMMARY 1. The integumentary cilia in Echinarachnius cover the tips of the short spines and the sides of the long ones. On the oral surface they beat radially; on the aboral they beat from anterior to posterior. They play no essential part in the locomotion of the animal, but are probably concerned with feeding, with cleaning the outer surface, and with the respiratory currents. 2. The ambulacral feet form five complicated radial bands on the oral and the aboral sides of the test and a complete marginal fringe. Their tips are deep pink and provided with suckers. They are significant in locomotion to only a limited extent in that on the anterior edge of the test they pile up the sand on the aboral surface. .>. Spines cover the oral and aboral surfaces. They are of two types, long and short. They are best developed over the anterior portion of the oral surface where their distribution exhibits bilateral symmetry in relation to the axis of locomot ion. In this regjon waves of coordinated spine movement pass from the anterior edge of the test posteriorly. In these waves each spine makes a vigorous posterior thrust in a vertical plane and an unimpeded recovery in a plane more nearly lateral. Forward locomotion, burrowing, and righting are types of motion dependent primarily on these spine movements. REFERENCES GISLEX, T., 1924. I < hinoderm Studies. Zoo/. Bidrag r/>/>snla, 9: 1. MAcBaiDE, E. \V., 1906. I ( hinodermata. S. 1 . Harmer and A. E. Shipley, The Cambridge \;iiur.il Hi>i<>ry, 1:425. PARKER, G. H., 1927. Locomotion and Righting Movements in Echinoderms, Especially in I < hinarachnius. Am. Jour. Psychol., 39: 167. STROM BIDIUM CALKINSI, A NEW THIGMOTACTIC SPECIES E. 1 \\URE-FREMIET COLLEGE DE FRANCE, PARIS Many ciliated Infusoria belonging to different groups can settle in a more or less temporary manner by means of their ciliated apparatus; this property is connected with some structural particularity, for example in Infusoria of the family Ancistridae and also in Ancyst- ropodium Matipasi, an hypotrichous ciliate. Almost all the oligo- trichous Infusoria of the family Halteriidae are planktonic species; yet some of them present occasionally thigmotactic properties and can stick or slide on the surface of solid bodies. The species of the genus Tontonia use, in this case, their curious caudal process lined by some small cilia; on the other hand, Strombidium urceolare Stein uses, according to Maupas, three long cirri located on the left side of the peristome; these cirri are fringed at their free ends whereby the Infusoria are temporarily fixed. W. v. Buddenbrock (1922) has described under the name of Strombidium davellinae a small species found in Heligoland which can either swim freely or slide on the mantle's surface of Clavellina lepadi- formis with which this infusorian is an habitual commensal; it can also fix itself at the bottom of a dish. The fixing organ is here once more an apparatus of ciliated origin, made by four thin membranelles belonging to the left side of the adoral zone but different from the others in a considerable lengthening and in the fringed structure of their thinner ends. During the summer of 1929, I observed at Woods Hole another species of thigmotactic Strombidium able to fix itself temporarily by a ciliated apparatus: I will describe it under the name of Strombidium Calkinsi, sp. nov. Strombidium Calkinsi is a species very nearly related to S. sulcatum Clap, and Lachm. and to 5. lagenula, differing, however, from these by the presence of two long dorsal membranelles independent of the peristomal zone, the ends of which can stick on solid bodies. This infusorian measures about 40 n in length; its body is irregularly ovoid with a great antero-posterior axis; the basal region, hemispheric, is bounded by a transversal depression ; the ventral peristomal groove is lined on the right by a vertical lip extending itself on the anterior side by a so-called semicircular collar which turns dorsally toward the left. 201 202 E. FAURE-FREMIET The adoral sinistral zone takes its origin in the gullet at the bottom of the peristomal groove; this zone surrounds the collar with fourteen great membranelles. The peristomal groove, more developed than in S. sulcatum and less deep than in 5. kigenula, extends toward the middle of the ventral face. Just as in tlu> two previous species, the posterior pole is indicated neatly not only by a transversal furrow but also by a refractive outline; this is due probably to the presence of a cuticle thicker than on the other parts of the body. A conical bundle of intracytoplasmic radiating rods fills the basal part of the bodv and clearly delineates the annular furrow; the significance of this formation, which exists in almost all species of Stronihidinni, is much discussed. There is no Kit.. 1. Stn>mbi i- c.mstit uted by two dorsal membranelles nearly as long as the body, measuring .o to -10 ju in length and 7 to cS /i in breadth. They appeal- delicately -niated longitudinally and the coalescent cilia, which continue these membranelles, separate more or less from each other at the adhesive di>tal e\trrmit\ . STKOMIiimi'M CALKIXSI 203 The insertion-lines of the two membranelles are rather near one another and produce two wrinkles obliquely bent from left to right, on the body's anterior dorsal side. From this point, the two membranelles spread, generally forming an angle more or less obtuse so that the points of fixation on the support can be slightly separated from each other. The infusorian, when fixed, shows its ventral side up; the anterior dorsal membranelle is approximately perpendicular to the axis of the FIG. 2. Strombidium Calkin si in fixed condition seen from the left side. body and therefore to the solid support. The dorsal posterior membranelle stretches obliquely behind and often seems twisted. There is no evidence of a contractile property of the membranelles. In this fixed position the adoral zone draws around the collar half a circle perpendicular to the great axis of the infusorian's body. The vibrations of its strong membranelles produce a fluid current which flows in the ventral groove of the fixed individual. From time to <-\< :*S 204 F FAURE-FREMIET time, the reaction to the current shakes the infusorian, thus showing the elasticity of the two attaching dorsal membranelles. At times, the infu^ M i.m -.eem- to be walking on the solid surface like an hypotrichous int'usorian, the two membranelles moving one after the other, but with no regular rhythm. Mon- often tlu- Stroiuhidinni unfastens all at once and swims hastily. dra\vn forward by movement of the adoral zone. In thi- case the two dorsal membranelles bend along the body, showing a crumpled aspect. It is of consequence to point out that the two dorsal attaching membranelles of Stronibidiuni Calkinsi are absolutely independent of the ciliary adoral apparatus. The position of their insertion line- shows that they belong to a somatic ciliature, generally absent in the oli-otrichous Infusoria, but which can reappear more or less modified, either in the form of small, short, faintly mobile cilia, inserted on the longitudinal lines, or else as a more or less developed "ciliary residual field," the presence of which I have indicated in various species of Tintinnidae. Some true cultures have developed on the surface of the slides which were put in tanks tilled with running sea water. BIBLIOGRAPHY l!i MM NBROi K, \V. v., 1922. flier eine neue Strombidium — Art aus Helgoland (Sir. clavellinae). Arch. f. I'mtist., 45: 129. FACRE-FREMILT, I .., 1'K)5. I .a structure de 1'appareil iixateur chez les Yorticellidac. Arch.f. Protist., 6: 207. i-' i' I REMIET, I-!., 1'JOX. L'Ancystropodium .Maup.isi. Arch. f. Protist., 13: 121. F.u -RK-FRKMIKI . 10., 1910. I. a fixation chez les Infusoires cilies. Bull. Sclent. /•>. Bel?,., 44: 27. I \i KI'. FRKMII.I, I'... \'>\2. Htmles cytologiques sur c|iirl. (Miiiril.uiic.ii ,'t I'l'-tiKk- ni Infusoires cilirs. A nil. zn«l. c.vj>cr. ft ^CH., SIT. 2, 1: 427. STEIN, 1 . VON, W'7. Der Organismus der Infusionsiliii-rf. II, p. 162. STUDIES ON THE CHEMICAL NEEDS OF AMCKBA PROTEUS: A CULTURE METHOD WILLIAM F. IIAHXKRT1 (From the Mount Desert Island Biological Laboratory, Maine and the /.ooln^iciil Laboratory of the Johns Hopkins University) INTRODUCTION The chemical needs of plants have been fairly well worked out by the culture method, but knowledge regarding the needs of animals is very inadequate. This is largely due to the fact that plants can be grown in synthetic solutions containing only inorganic salts, while animals require in addition some organic material. For example, Amceba proteus feeds on Chilonwuns paramecium; Chilomonas in turn requires some organic nutrient. In the culture of protozoa, the organic nutrient has usually been added in the form of timothy hay or grain. These substances, however, contain a considerable amount of physiologically active salts, which diffuse out into the culture and alter it in an unknown way. Under these conditions, since the kind and concentration of chemical elements are not known and do not remain constant, it is difficult to ascertain the kind and number and relative amount of elements necessary for the maintenance and growth of animal protoplasm. By a method to be described below, variation in the salt composition of the medium is fairly accurately controlled, thereby making it possible to ascertain the relative importance of individual elements in rhizopod protoplasm. METHODS AND RESULTS Amceba proteus (Leidy) was used in all the observations made. It was derived from stock cultures made by adding a grain of rye to a mixture of half-spring-half-distilled water in ringer bowls with subse- quent inoculation with amoebae, and Chilomonas. Two series of experiments were performed in which variation in the salt composition of the medium was controlled. These experiments were made as follows: Five balanced salt solutions 2 were prepared as 1 The author is indebted to Professor S. < ). Mast for helpful suggestions and valuable criticisms, especially in the preparation of the manuscript, and to the Research Corporation for financial aid. 2 Kahlbaum (analysis grade) chemicals (except Merck's blue label Ca3(PO4)2) and water redistilled in pyrex glass were used in all solutions. According to recent investigations by Williams and Jacobs (1931), certain brands of C. P. sodium chloride contain a toxic impurity whose destructive effect overbalances any beneficial effect the sodium chloride itself may have. It should be noted that the Kahlbaum salt, which Williams and Jacobs found to be the least toxic of the five brands tested, was used in these experiments. 205 206 WILLIAM F. HAIIXERT indicated in Table I. The first, Chalkley solution,3 contained nine chemical elements in the form of salts, the rest contained fewer. Then 20 cc. of the solution under consideration was put into each of five 50 cc. pyrex glass beakers. Numerous amoebae were now removed from a slock culture and put into a pyrex glass beaker containing 25 cc. redistilled water and left a few minutes, after which all that were in good condition were transferred to another pyrex glass beaker con- taining ivdi-tilled water. This was repeated a third time. Then In : amoebae, washed free of culture fluid, were put into each beaker containing the different salt solutions. Specimens of Chilomonas were now concentrated by means of a centrifuge; they were then added to a lame quantity of redistilled water and again concentrated, after which 0.2 cc. d| the resulting dense culture of Chilomonus was added to each beaker. This organism served as food for the armrb.r. I hiring the process of washing in redistilled water, practically all of TABLK I Chemical Composition of the Solutions Tested as Culture Media for Aiiia'ba protcns Compound (i) (2) (3) (4) (5) NaCl gram 0.08 gram 0 OS zriim OOS Rram gram XaHCO3. . 0.004 0.004 0.004 KC1. . . 0.004 (1.004 0.004 0.004 < aCh 0.004 0.004 0.001 0.004 0004 ( all; PO4)j 0.002 0.002 ()()()_> 000' 000' MR PI - 0.002 0.002 0.002 0.002 ' .1 (PO4)o 0.002 H20 (cc.) 1000 1000 1000 1000 1000 the original culture fluid with its unknown chemical content was eliminated from both anio b.i- and Ciiiloniniins. No nutrient in the form of hay or grain was added to the salt solutions in the beakers. Therefore, since the composition of ('/iilt»)iont h series < >l experiments, all of the beakers were kept in diffuse light. Kach was co\ en-d with a glass plate to reduce evaporation. The temperature during the course o| the tests was fairly constant and was the same for all. < )bser\ at ions with reference to the number of ,11110 b. i- and their physiological condition were made with a binocular microscope. The condition of Chilomonas was also noted. These I hi-, -olui ion is a modification by Chalkley of one used by Drew in tissue culture work: ii IMS proven a rcli.itilc culture medium for Atna'ba prateus. CHEMICAL NEEDS OF AM(EBA PROTEUS 207 observations were made every day or every other day during the lirst half of the experiments; a final observation was made several days later, at which time the experiments were discontinued. The arn!" the three tested. The results obtained with spring water an- of interest only in comparison and need not be considered further. In the second experiment four solutions were tested in the same wax as those in the preceding experiment. These salt solutions are de- M-ribcd in Table I; i.e., (1) Chalkley solution, (3) Chalkley solution without potassium salt, (4) Chalkley solution without sodium salts, and (5) Chalkley solution with the sodium salts replaced by calcium tribasic phosphate. The hydrogen-ion concentration of (1) and (3) remained between pH 6.4 and 6.6 and that of (4) and t5) between pH 6.2 and 6.4 during the 17 days of (he experiment. The omission of .-odium bicarbonate in (4) and (5) is responsible for this difference in h\ drogen-ion concentration. TABLE III feet of Different Suit Solutions on Growth in Ania-ha firotcns :i ion Days after inoculation Average I.IM.M-MV per original individual 0 2 4 (i 17 Number of animals present 1 < halklcv si ilul ion 50 50 50 50 97 85 80 04 225 210 210 233 563 565 577 656 1956 1686 2404 2676 39.1 33.7 1S.1 53.5 s.inie less potassium salt 1 ^.ime less sodium sails .... (5) S.iinc \\iih sodium salts re- placed l>v Ca i|'(),i, The results obtained regarding the number of anneb.e produced are presented in Table III. The numbers have the same significance .1- in the preceding table. By referring to this table, it will be seen that there was little variation in the number of anm-b.e present in the different solutions during the first part of the experiment but th.it there was considerable \.niation during the last part. This variation on casual obserxation may not appear great, but when subjected io statistical analysis, it shows certain relations of sonic significance. Since the data were obtained on cultures (i.e., group- of ten amo-ba-) instead of on individuals, the mean progeny in live cultures of a solution was cho^-n as the relati\e value for that so- lution. The means with their probable errors and the difference between the means wii h their probable errors, computed from the same data as Table III, are gixen in Table IV. CHEMICAL NEEDS OF AM(KBA PROTEUS 209 It is considered reasonably certain that the mean of several measurements falls within three times its probable error; i.e., the chances are 22.5 to 1 that it does. Table IV shows that the means of cultures (1) and (3) and also those of (4) and (5) are separated by about twice the sum of their probable errors whereas those of (1) and (4) are separated by more than four times the sum of their probable errors. Computation of the difference between the means shows the same relations; the difference between the means of (1) and (3) is 2.73 times the probable error, that between (4) and (5) 1.73 times, and that between (1) and (4) 5.25 times. It will be noted also that when the other solutions are compared with Chalkley solution, each of the differences between the means is probably significant, i.e., in comparing (3) with (1), it is 2.73 times the probable error, (4) with (1) 5.25 times, and (5) with (1) 4.19 times. This indicates that the difference between (1) and (3) and also between (4) and (5) may or may not be TABLE IV Means and difference between the means of number of amceb(z present in the different solutions. Solutions arranged in ascending order of means. Based on same data as Table III. Mean amoebie Difference Mi-Ms with p.e. with p.e. p.e. (3) Chalkley less K-salt ... 337.2±15.55 (1) Chalkley 391 2±12.26 - 54.0±19.76 2.73 (4) Chalkley less Na-salts 480. 8 ± 7.03 - 89.6±17.06 5.25 /less Na-salts. . 535.2 ±30.70 - 54.4±31.47 1.73 (;>) Chalkley(plus ^.^ 144.0±34.33 4.19 significant of an actual difference between the sample distribution in these solutions but that the differences between (1) and (4) and also between (1) and (5) are significant of actual differences and cannot be due to random sampling alone. Comparison of the distribution in the different solutions by the X2 method (see Pearson, 1914) indicates also that for any two solutions the variation in the number of amoabae present cannot be due to random sampling alone. Consequently, since the cultures were set up with the same care, received the same treatment, and differed only in the salt content of the solutions, the observed variation in fission rate must be due, at least in part, to difference in the chemical composition of the solutions. The data presented in Table III indicate therefore that Chalkley solution without sodium is more favorable than that with sodium, and 210 WILLIAM F. HAHNEK T that Chulklev solution with tlie sodium salts replaced by calcium tribasic phosphate is still more favorable. It indicates also that the absence of pota->ium is detrimental because the am,e in the so- lution without it had the lowest fission rate of all. The difference in hydrogen-ion concentration of (4) and (5) as compared with (1) and 5 may be a factor in the more rapid fission rate in the former solutions, although the optimum is generally considered to be at pH 6.6-0.7. In all solutions the amu-ba- remained in good condition during the experiments; at each observation practically all were attached and moving in monopodal or bipodal form. Little change was observed in C/iilomomis during the first week, and they were still in fair condition at the close of the experiments, although less rounded and plump than at the beginning. Detailed studies on the structural and physiological changes in Chilomonas during starvation are in progress. SUMMARY 1. Ama'ba proteus grows and reproduces for several weeks in a balanced salt solution containing potassium chloride, calcium chloride, calcium phosphate, magnesium tribasic phosphate, and Chilomonas parameciuiu . 2. They also grow in other solutions, but not so well; e.g., in a solution containing sodium chloride and sodium bicarbonate in addition to the above salts. 3. The results obtained indicate that the presence of sodium is not only unnecessary but actually detrimental while that of magnesium and potassium is favorable if not essential for growth and reproduction in Amoeba proteus. 4. By observing the effect on fission rate and other physiological processes of omitting various elements or altering their concentration in the solution, it is possible to ascertain the relative importance of tin- elements to the rhizopod protoplasm and the number and kind and amount necessary for growth. Mllil.lOi .KAI'llY CLARK, \Y. M., 102s. Tin- Determination of I lyilro^rn Ions. Hall inmiv. DAVENPORT, C. I'. 1S')7. Kxprrinn-ntal Morphology. Vol. 1. \f\\ York. I>KI\\. V. H., 1928. Motes on the Cultivation of rumours in Vitro. Arch. f. exper. Zell HOI-KIN-^. 1 >. I 1928. I In- I Hi i i of ( rii. tin Physical and Chemical Factors on Locomotion and Other 1 lie I 'locesses in Amo-l>.i proteus. Jour. Morf>h. and Physiol., 45: ''7. MAST, S. O., 1928. I a< tors ln\ ol\ nl in ( 'lunges in Form in Amu-ha. Jour. Exper. Zool., 51: (>1. CHEMICAL NEEDS OF AMCEBA PROTEUS 211 MAST, S. O., 1931. Eli'ect of Salts, Hydrogen-ion Concentration, and Pure Water on Length of Life in Amoeba proteus. Physiol. Zool., 4: 58. PEARSON, KARL, 1914. Tables for Statisticians and Biometricians, p. 26. Cam- bridge University Press. WILLIAMS, M. M., AND M. H. JACOBS, 1931. On Certain Physiological Differences between Different Preparations of So-called "Chemically Pure" Sodium Chloride. Biol. Bull., 61: 485. THE FORMATION AND STRUCTURE OF THE GLOCHIDIAL CYST1 LESLIE B. ARKY .\N\IIIMICAL LABORATORY, NORTHWESTERN UNIVERSITY MEDICAL SCHOOL CONTENTS Introduction 212 Attachment of Glochidia 213 Hooldess Glochidia 213 Hooked Glochidia 214 Encyst ment 214 Generalities 214 The Method of Cyst Formation 215 The St ructure of Cysts 2 \ <> Relation of the ( ilochidium to its Host 2 1 S The Rupture and Repair of Cysts 2 1 s Summary 220 INTRODUCTION An important phase in the development of fresh-water mussels is the obligatory period of parasitism spent upon appropriate fish hosts. While superficially encysted on such hosts the tiny larval ' glochidium ' transforms into a free-living juvenile mussel, more complex in internal structure but without any corresponding increase in external size. The purpose of the present communication is to record the events incident to the formation of the glochidial cyst, to describe the structure of the cyst throughout parasitism, and to examine the morphological relations existing between parasite and host to subserve metabolic functions. A preliminary report was published some years ago (Arey, 192. -i). |)aia for the hookless group of glochidia have been drawn chiefly from an inteiiH\e study of infections of f.nnipailis In/coin on the gills of the large-mouth black bass (Mtcropterus xnlnwides), and of /,.\ilis tnioilontoidcs on the long-nosed gar (Lepisosteus osseiis). Similarly, the hooked series comprised stages of Hemilastena ambigua on the gills of the niodele \(»iiis humilis}. ( 'losely graded stage- of enrys.iment are easily procured by intro- i ontribiii ion \d. l<>^. Published by permission of the United States Com- missioner of Fishei ii^. Acknowledgment is due the stafT of the Kairport Biological •ion for many helpful courtesies extended during the prosecution of this inquiry. 212 r.LOCHIDIAL CYST 213 ducing the host into a small aquarium into which ripe glochidia have been placed. Attachment follows quickly, and samples of the gills or fins bearing glochidia can then be removed at intervals as desired. Such samples of L. litteola, A. corpulenta and //. ambigua were fixed promptly in Zenker's fluid. The L. anodontoides stages were preserved in Bouin. All the material was sectioned serially in paraffin at 6 ^ and stained with hematoxylin and eosin. ATTACHMENT OF GLOCHIDIA The tiny bivalved glochidium (0.3 mm. or less in size) is incapable of locomotion when liberated from the maternal gill. Chance alone brings it in contact with suitable hosts. If a fin or gill filament becomes inserted momentarily between the valves so that the chimney- like hair cells are touched, the glochidium snaps shut vigorously, pinching the intercepted tissue.2 The more delicate, hookless group of glochidia attaches to the soft gill filaments, but the sturdy, hooked glochidia can pierce the fins as well. The process of attachment may be observed on excised gill filaments or fins placed with glochidia in a watch glass under the microscope. Hookless Glochidia. — The sharp edges of the valves cut cleanly through the gill epithelium, affording surprisingly little evidence of hemorrhage or seeping from the incision. The location of the parasite on the gill filament governs the character of the bite. Those that attach to the blade-like edge of the filament cut through the epithelium and, usually, considerable connective-tissue stroma as well.3 This is characteristic of most well-attached larvae, for these enclose a liberal amount of the deeper tissues. After the epithelium is passed the valves continue to close, cleaving and compressing the underlying stroma. The softer tissues are cut; the tougher gill substance, especially that containing blood vessels, is merely pinched. The walls of the blood vessels and other resistant constituents are con- stricted at the level of the compression, and expand like an hourglass on either side (Figs. 1 and 2). It appears that the valve rim cuts both epithelium and soft stroma with ease, but when it encounters the tougher elements, the rim buckles inward until it lies flat against the interior of the valve proper.4 2 Arey (1921). This publication contains a full discussion of the factors involved in closure. 3 Attachment along the edges of the filament is most favorable for easy and successful encystment. Many glochidia embed deeply in the firm gill substance, sometimes even half below the surface. 4 The mechanisms involved in the operation of both flange and hook are described in full in a separate contribution (Arey, 1924). 214 LESLIE B. AREV This serves the very practical purpose of furnishing a broad zone of contact, while at tin- >,ime time the glochidium is prevented from cutting itself entirely free (Figs. 2 to 5). Attachment to the gill lamellae is essentially similar, but as thin- walled, vascular laminae are encountered in this instance, the chief factor is compression rather than incision. In a typical case the pinched lamellae converge inward toward the approximated valvular rims. I looked Glochidia.— The events during the attachment of hooked hidia are comparable to those already described for the hookless type. Gill infections are practically identical, but fin parasites may lie wholly within the epithelium. The hooks flex much as do the flanges in the other group, but their effect is more local.4 They pierce the host tissue like tongs, and then are inturned; the tissue is thereby held firmly, while the spines which beset the outside of the hook lock it still further. ENCYSTMEXT Generalities. — The process of attachment is completed almost instantaneously. Both incision and compression are accomplished in less than a second. As a result, the ventral edges of the valves sink somewhat below the surface level of the host tissue (Fig. 1). There next ensues a period during which the glochidium is overgrown by the contiguous cellular tissue of the host. Successive stages of this are shown in Figs. 2 to 5. The covering-in process is rapid. In summer the black bass completes its gill cysts in about 3^ hours; yet I have observed fully formed cysts as soon as 1\ hours after attachment, and well advanced stages at one hour. Excised filaments in watch glasses may encyst glochidia even quicker than under normal conditions; two hours has been found sufficient to complete the process. The response in the gar-pike is slower than in the black bass, but encystment has been observed in three hours Lower temperatures retard the re- action proportionately. Glochidia which attach to gill lamellae do not form cysts as readily as those on the thicker edges of the filament. This is doubtless due to the amount of material available, as will be explained presently. On the same gill the lamellar cysts may demand twice the time taken by iliose along the filament's edge. Lamellae adjacent to the glochidium may unite by fusion to form the basal part of the cyst, which is then roofed over in the usual way. At first, cysts tend to be somewhat thick, irregular and unsym- metrical (Fig. 5). Within two or three days they usually become thinner, smooth contoured and even (Fig. 6). My observations are in GLOCHIDIAL CYST 215 complete agreement with Young (1911) on this point; it is strange that Schierholz (1889), Faussek (1901) and Harms (1907) have all described the cyst as originally thin and only gradually gaining thickness. When glochidia acquire a weak attachment and clasp but a small shred of host tissue, encystment is commonly unsuccessful and the glochidium is lost. The Method of Cyst Formation. — It is natural to assume that direct proliferation of the cells of the host tissue provides the material for the cyst that encloses the glochidium. Indeed, this assertion is presented as the correct explanation of encystment by several observers (Young, 1911; Lefevre and Curtis, 1912; and still earlier workers). That such an explanation is both inadequate and contrary to fact has been the topic of another publication by the present writer (Arey, 1932&). The a priori argument against encystment through cell multiplication rests on several facts: (1) The cyst may be composed of several thousand cells; (2) the time required for the formation of a cyst under favorable conditions is only three to four hours; (3) the mitotic cycle is relatively slow and consumes several hours in cold-blooded verte- brates. Actual observation of encystment stages does not show the presence of more than the ordinary number of random mitoses seen in control, uninfected tissue. For example, in 78 cysts of Lampsilis luteola on the black bass, representing stages between 30 minutes and 9 hours after attachment, a total of only 20 positive mitotic figures and 14 doubtful ones were found as the result of a thorough census under the highest magnification. Again, in 17 cysts of Hemilastena ambigua on the gills of Necturus only one mitotic figure occurred during the period of encystment.5 These results definitely disprove the theory of encyst- ment through the proliferation of new cells. Turning now to the real factor underlying encystment, the natural alternative method is actually encountered. This process is one of cell migration, whereby neighboring host cells assemble and actively push forward over the invader until the wound is closed and the glochidium is covered in.6 After encystment is complete there may be a compen- satory period of cell division in the vicinity of the cyst to replace the cells lost during the cellular emigration leading to cyst formation. For the details of this process, and its relation to wound healing in general, the reader is referred to the complete publication already mentioned (Arey, 1932a). 6 My material did not include stages beyond cysts three-fourths completed. 6 In gill infections on fishes the cyst wall is composed both of epithelium and connective tissue (Figs. 5 to 8). Goblet cells or pigment cells are frequently carried along into the cyst. (Figs. 5 to 8). 216 LESLIE B. AREY THE STRUCTURE OF CYSTS The cyst is repeatedly designated as 'epithelial' by authors who have written on these matters. This, however, expresses only a half truth. Fin parasites, to be sure, may lie entirely within the stratified epithelium, and the same is true for some of the Ilemilastena cysts on Nectnrns gills. But even in these cases there is commonly attachment to fin rays or connective-tissue stroma which necessitates a more or less extensive defect in the epithelial covering at its base. Parasites on the gills of fishes usually bite deep into the stroma. Not only does connective tissue adjoin the glochidium here but it is carried up into the roof of the cyst as well, so that the larva in reality lies embedded in stroma (Figs. 2 to 4).7 Often the epithelium forms a mere external arching canopy. The demarcation between epithelium and cellular connective tissue is commonly very indistinct, and the latter is easily mistaken for the former.8 Doubtless this circumstance accounts for the existing confusion and erroneous statements con- cerning the composition of the cyst wall, for in some locations the interpretation is indeed puzzling and the two do appear to blend. Yet EXPLANATION OF PLATE Abbreviations a.m., adductor muscle of glochidium g.c., goblet cell (-./., connective tissue of cyst .?./., gill filament C.ILI., cyst wall ;•/.. .ulochidimn ep., epithelium of cysl /'./., host tissue /., flange of valve /.»/., larval mantle FIG. 1. Glochidium of Lampsilis Ititcola just attached to a gill filament of the black bass. Photo. X 150. FK,. 2. An early stage in the encystment of L. luteola (30 minutes after attach- ment . Photo. X 3(Mi. I IG. 3. A half-formed cyst enclosing L. luteola (\\ hours after attachment). I'):.. to. X 300. FK,. 4. A cyst nearly completed aboul L.lnlfoli (1\ hours after attachment). Photo. / 300, FK.. 5. ! he K.iiiplcte encystment of L. lulcoln <1\ hours after attachment). Phot... / 300. IK,. (.. I IK- \\.-ill nf a /,. /n/i'i>ln <•!' i lie \\all of a large L. lntral. (-ill filamriii <>l i In- l.lark kiss. Notches indicate the former location ..nulled L. liili-iil, i ^lochidia. Photo. ',< ln. 7 1 >i-i-p-lyiii;< iiiclaimphorr- ha\r l.rrii loiind in the cyst wall. In some cysts, i-spccially those assoi i.iicd with iinmunity, rosinophils also invade the stroma (Arey, 193 \Ialloi-\'.s connective tissue stain does not differentiate these tissues in fishes. CYST 217 ••».-. «>»* V3 - i y ep. dc C,w. cw. *m <»!*•' cw 218 LESLIE B. AREY in favorable preparations the demarcation is clear (Figs. 6 to 8) ; in the Lampsilis anodontoides series on the gar this differentiation was especially evident. Corroborative proof lies in the fact that delicate blood vessels may course through the cellular stroma, and in immune animals eosinophils \vander freely through it (Arey, 19326). Correlative to these findings, the propriety of the term 'cyst' may l.r <|iic-tioned as an exact designation for all glochidial investments. If by 'cyst ' is meant a distinct envelope which demarks the larva from the adjacent tissues, then such does not exist. In the fin parasites the •chidium simply lies buried in the epithelium, or partly in the connective-tissue stroma. In the gill parasites the position is primarily in the stroma, with a roof-like canopy of epithelium outside. Vet the term is so thoroughly established and convenient as to make its replacement unwise. The original irregularities of the cyst (Fig. 5) smooth over, and after a few days the roof tends to appear stretched and compact (Figs. 6, 7 and 9). Usually the distinction between epithelium and connective tissue then becomes plainer (Figs. 6 to 8). The goblet and pigment cells carried up in gill infections persist there (Figs. 6 and 7). RELATION OF THE GLOCHIDIUM TO ITS HOST Since the glochidium cannot metamorphose except on appropriate hosts it might be thought that special nutritive relations are established between parasite and host, and that this results in recognizable morphological changes or adaptations in the enveloping tissue. As a matter of fact, this possibility is not realized (Figs. 6 to 8). The soft, and for the most part highly vascular, tissues in which the parasite is embedded are apparently adequate for handling whatever interchanges are necessary without anv special elaborations. The host tissue ingested at the time of attachment, together with the degenerating larval adductor muscle, are important sources of nutriment during transformation (Arey, 1932r), so that there is no metabolic 'strain.' The adjoining host tissues do not become unusually vascularized (Figs. 7 and 8) except in the Proptcra glochidial type which is peculiar in that it undergoes marked increase in size during a postmetamorphic period of retention. In some specimens of Pwpu-ni laevissima which liad increased in luilk some 40 times, the cysts were found to be very large and thick, and capillaries were present that presumably repre- sented secondary iiua.sive growths. TIN. Ki I'M KI \M> Ki.r.uR OF CYSTS When the cyst first forms, its \\all is regionally variable in thickness and usually bears irregular outgrowths (Fig. 5). After a day or two it GLOCHIDTAL CYST 1 \ <> becomes smooth and quite symmetrical. The tissue over the top gradually assumes a compact and stretched appearance, and the wall as a whole is thinner (Fig. 6). This reduction and thinning is much more spectacular in the bulky cysts which characterize the brief attachment of glochidia on immune hosts or non-hosts. In a contribution (Arey, 19326) specifically describing these conditions it will be shown that the thinning is apparently due to the removal of cells back into the filament, rather than to a loss by desquamation or otherwise. After the first days of encystment there are no especially significant changes in the cyst until the time when the glochidia are shed. Young (1911) has described a characteristic loosening of the cyst tissue and a concomitant infiltration of lymph after about one week of parasitism. A mild degree of cellular separation, in which intercellular bridges become prominent, occurs also in some of my series. Nevertheless, this is by no means a regular phenomenon, while sometimes it is observed relatively distant from an encysted glochidium as well. To what extent such alterations are artefacts and what their proper interpretation may be must remain unanswered at present. Miss Young suggested a causal relation to the premature sloughing of partly transformed glochidia. This may be true, but if so it is not a charac- teristic method by which these parasites terminate a normal period of encystment. At the end of the parasitic period the glochidium becomes free of the host.9 It has not increased in external size, but internally the meta- morphosis is marked. Liberation is partly the result of the young mussel's own activity, for at intervals prior to detachment the valves may be observed to move and the foot to be pushed about, pressing the cyst wall. This is demonstrable when at this time filaments are removed and kept in watch glasses under a microscope; incidentally, there is reason to suspect that emergence is accelerated by such in vitro procedure. The cyst is eventually ruptured, but sections do not show that this is made perceptibly easier by any sudden terminal thinning or weakening of the wall. Portions of the old cyst-covering may be carried away and adhere for a time to the freed glochidium. Ap- parently a certain amount of gross sloughing aids the shedding process, for when infections are made on immune fish, transformation of the glochidium fails and the passive glochidium is liberated while still encysted (Arey, 19326). The freeing of the transformed glochidium leaves a defect in the filament which is rapidly filled in (Fig. 9), probably by the same sort of 9 Those at the tips of gill filaments are often retained longest (Fig. 9). This is conceivably due to their less favorable position for receiving nutrition or oxygen. 220 i I>LIE B. AREY cell mobilization that characterized encystment. Sections are not particularly informative on this point, and even the sites of the cysts are not easily detectable in microscopic preparations. In some in- stances the ca\it\ of the remnant of the former cyst is temporarily filled with a coai^ulable exudate. SUMMARY Mechanical stimulation of the larval glochidium induces non- selective, automatic closure upon the impinging gill or tin. The valves largely cleave the soft tissues encountered, but merely clasp such tough elements as blood vessels and tin rays which lie deeper. As a result, part of the glochidium is buried in host substance. The glochidium is then covered by host tissue which advances from all sides, primarily for the purpose of closing the wound. Encystment is not the result of cell proliferation. On the contrary, it is accom- plished by a mass movement of cells from the adjoining regions, advancing by their own activities and directed over the exposed valves by thigmotaxis. A compensatory period of mitosis may appear -ubsequent to encystment, apparently to replace cells lost to the cyst by emigration. Fin cysts are largely epithelial in structure. Glochidia which attach to gill filaments lie embedded in cellular connective tissue, roofed over with an epithelial canopy. Shortly after encystment is completed the cyst becomes thinner, smoother and more symmetrical. Thereafter, and even until the time of rupture, there are no further significant morphological changes in the cyst. Special adaptations of the host tissues to care for the wants of the metamorphosing parasite are not developed. The glochidium is liberated partly through its own efforts, ap- parently aided somewhat by sloughing. Repair of the resulting defect in the host tissue is rapid and probably follows the general method utilized at encystment. This would involve an early redistribution of existing cellular elements, followed later bv the formation of new cells to restore the tissue balance. LITKKATI RE CITED Aki N , I.. M., 1(>J 1 . An \ \peri menu I Study On < .loch id ia and t he Factors Underlying I n. \Mmeiit. Jour. /'.'.Y/'IT. /.mil., 33: 4d.v Aki <, , I . I'.., 1('2.v [he I cirm.iiiiin and Structure of llie < .locliidial Cyst. Anal. AY, .,26:. •! 50. I . I',.. \<>1\. < .lodiidi.d Cuticulae, Teeth, and the Mechanics of Attachment. Jour. Mort'li. ,uid /'/;V.N /--/.. 39: S_'.v ! . I'.., I932a. Certain Basil Pt UK i i iles of Wound Healing. .1 nat. Rec. 51: r.mriliniAL CYST 221 AREY, L. B., 1932/;. A Microscopical Study of Glochidial Immunity. Jour. Morph. and Physiol., in press. AREY, L. B., 1932c. The Nutrition of Glochidia during Metamorphosis. A micro- scopical study of the sources and manner of utilization of nutritive sub- stances. Jour. Mar ph. and Physiol., vol. 53. FAUSSEK, V., 1901. Ueber den Parasitismus der Anodonta-Larven. Verhandl. d. V internal. Zool.-Congr., (Berlin) S. 761-766. HARMS, W., 1907. Zur Biologic und Kntwicklungsgeschichte der Flussperlmuschel (Margaritana margaritifera Dupuy). Zool. Anz., 31: 814. LEFEVRE, G., AND CURTIS, W. C., 1912. Studies on the Reproduction and Artificial Propagation of Fresh-water Mussels. Bull. U. S. Bur. Fish., 30: 105. SCHIERHOLZ, C., 1889. Uber Entwicklung der Unioniden. Denkschr. d. k. Akad. Wissensch. (Wien), math.-naturw. Kl., 55: 183. YOUNG, D., 1911. The Implantation of the Glochidium on the Fish. Univ Missouri Bull., Sci. Ser.f 2: No. 1. Vol. LXII, No. 3 June, 1932 THE BIOLOGICAL BULLETIN PUBLISHED BY THE MARINE BIOLOGICAL LABORATORY A STATISTICAL TEST OF THE SPECIES CONCEPT IN LITTORINA JOHN COLMAN MUSEUM OF COMPARATIVE ZOOLOGY, HARVARD UNIVERSITY "One may believe that if larger series were more often utilised in taxonomic work the current bewilderment over variation would give way to a renewed respect for a certain uniformity that exists thruout such groups of individuals." — Kinsey. With certain exceptions, no two individuals of a species are ever genetically identical; hence it is not so much the uniformity as the character and range of variation in a species that are diagnostic. Conversely, the fact that two related animals differ does not necessarily mean that they belong to separate species unless it can be shown, after the examination of sufficient numbers collected over a wide area, that there is not a series of overlapping intergrades between the two differing forms. It was, for example, the range in variation in the number of vertebrae in conger eels that enabled Johannes Schmidt (1931) to separate the American species, Conger oceanicus, from the European C. mdgaris. The ranges of the larvae of the two species overlap in part geographically, but not anatomically, the number of vertebras in C. oceanicus being from 140 to 149, average 144.63, and in C. vulgaris from 154 to 163, average 158.16. This lack of overlap in the numbers of their vertebrae clearly justifies their segregation. Kinsey (1930) has shown, too, that the highly variable Gall Wasp, Cynips erinaceus, is one species, though extreme forms of its gall have been previously assigned to separate species. In any part of its wide range a comprehensive collection over a square mile will very closely resemble a similar comprehensive collection at any other place in the insect's range. The variation in the species is roughly constant throughout its entire geographic range. These are two extreme cases; the first of two distinct, the second of one homogeneous, species. If, however, the two conger eels had 15 223 224 JOHX COLMAN numbers of vertebra- ranging from 140 to 152 and from 150 to 163 respectively, it would IK- impossible from the vertebrse alone to assign to either group those possessing 150, 151 or 152 vertebra*. To get round this difficulty, one man would consider them mere races, another sub-species, and a third separate species with a certain amount of hybrid!/, ition, the choice depending on the taxonomic upbringing and prejudices of the worker. Again, if in such a case as Cynips rrindceiis the amount of variation were in >t constant, a comprehensive collection from one part of its range would be different from a similar collection from another part. If the variations in these two collections did not overlap, they might In- considered separate species unless they wrere merely the ends of a continuous series. This last is the case with Littorina obtustitti. There has been but little uniformity in the treatment of its varying forms, and the resulting taxonomic muddle must be first of all cleared up. The establishment of the genus Littorina has always been ascribed to the elder Ferussac, the reference in the Index Animalium (Sherborn, 1927) being to page xxxiv of the Tableaux systematiques generaux de rembranchement des Mollusques (Ferussac, 1822). This refers, however, only to where Littorina is listed as the fifth sub-genus of the genus "PALUDINE, Paludina, FERUSS. (Fluv. et marin.)," with no definition or description. The definition, such as it is, is given on page xi, where Ferussac says: "Quant aux paludines marines qui constituent le genre Trochus d'Adanson, comme nous n'avons pu adopter cette denomination a cause des troclnis de Linne (en general formes d'especes reellement congeneres, ce qui nous a empeche de les appeler tnrho avec Adanson, en y rapportant tous les turbo de Linne qui doivent s'en rapprocher), nous en formons un sous-genre sous le nom de littorhif." (No Latin is used here, but an objection on that ic would, I think, be oversteeped in pedantry, since only the final e needs altering.) Two pages before, at the bottom of page ix, Fcrussac gives a list "dresse d'apres 1'edition de Grnelin," referring to the thirteenth edition of the Systema Natune, edited by J. F. < .melin (Linna-us, 1 7SS). This list, compiled from the Linnrean genera Troihus and '1'urlio, which Ferussac wished to combine, includes five -pecies referred to as " I'lil-mluin Marine." These are oblnsatus, nrritoitlcs, littorcus, ninricntiis, and (ifcr, constituting his sub-genus Littorina. " I.e ijenre Trochus d'Adanson," mentioned above, consists of four -pe< ies described and illustrated in the Histoire Naturelle du Senegal i Adanson, 1757). Their names, on a binomial system of Adanson's SPECIES CONCEPT IN LITTORINA own devising, antedate the tenth edition of the Systema Naturae (Linnaeus, 1758) by one year, and therefore do not stand. These species are beautifully illustrated in Adanson's work, and undoubtedly belong to the genus Littorina. Menke, 1828, was responsible for the spellings Litorina and litorea, although both Lin minis and Fcrussac used two /'s. Menke said that the alteration made for correct Latin, which is true, but it also led to a small confusion in nomenclature which has lasted till this day. In truth, "optima nomina quae nihil significant." The first full definition or description of the genus Littorina is given in the second edition of Lamarck's Histoire Naturelle des Animaux sans Vertebres, Vol. XI, pp. 201 et seq. (Lamarck, 1843). Of the British Littorinas, L. littorea and L. neritoides are both described under the genus Turbo in the tenth (1758) edition of Linnaeus' Systema Naturae, and they have retained their original specific names and authority, except in the work of a few authors who have fortunately caused no lasting confusion, save for the fashion, introduced by Menke (1828), of spelling with one /. The nomenclature of the mollusks generally known by naturalists as L. obtusata and L. rudis, however, has for a long time been a field for error and dispute. Littorina rudis is assigned by the Marine Biological Association (1931) to Maton, quoting Jeffreys (1865). Johnson (1915) gives L. rudis (Donovan), quoting Gould (1870) and Donovan (1804) (whose date is erroneously given in Gould as 1800, the date of publica- tion of the second volume; the first volume was actually published last, in 1804, after the fifth in 1803). Kuester (1856) calls the un- fortunate animal L. rudis (Montagu), quoting Montagu (1803) and Maton and Rackett (1807), in spite of the fact that the latter authori- ties cite Maton (1797) as the originator of the name and description; Jeffreys (1865) also gives Maton. Forbes and Hanley (1853), however, mention Donovan (1804), though Maton and Rackett (1807) give Maton (1797), Montagu (1803) and Donovan (1804) in order of priority. Montagu disturbs this order by quoting from the then unpublished volume of Donovan, having presumably seen the manu- script or proofs, but Donovan nevertheless gives the credit of the name rudis to Maton (1797). Menke (1830), the first to put Turbo rudis into the genus Littorina, cites Montagu (1803) as author of the species. Under the Law of Priority, the name rudis must be referred to Maton (1797), since there is no doubt of the identity of the form he described. 226 JOHN COLMAN However, nidis is not the correct specific cognomen of the snail which usually passes under that name. In 1792 Olivi published a description and rough figures of a shell near Venice which he called Turbo saxatilis. Jeffreys (1865) states that this form is identical with Littorina ncritoides (L.), but Dautzenburg and Fischer (1912) present on two plates sixty-two exquisitely colored and enlarged figures of L. saxatilis and L. rudis, which show that L. ntdis is specifically indistinguishable from L. saxatilis, of which names the latter has the priority by five years (1792 and 1797).1 In the case of the snail variously known as Littorina obtusata, L. litloralis and L. palliata, the difficulties begin with the tenth edition of the Systema, where two very similar shells are described under different genera, to wit, Turbo obtusatus (Vol. I, p. 761) and Xcrita littoralis (Vol. I, p. 777). The early British conchologists Montagu (1803), Donovan (1804) and Maton and Rackett (1807) all accepted Nerita littoralis as representing the common form of the English Channel and British Coasts. Here Montagu introduces a minor confusion by claiming that Nerita littoralis L. is the same as Turbo ncritoides L. He is alone in this opinion, and it is difficult to see what led him to form it. In 1822 Thomas Say described Turbo pli\h is not to be confused with L. sn-xntilis (Johnston), which is a SO-r.dlrd \.iric-ty IKMII llnuirk, KntJ.md i Jeffreys, IS'o . 2Tlu"M- liciini. names are regarded in this paper as «mly of subgencric rank. They ;m- extremely valuable in ilu- tun-i- delimitation ul tin- i^i'iius Littorina as a \\liolr, but ronsidi M d as IM-IHT.L tlicy only bewilder (lie poor naturalist and field worker. \< ry full lists of synoniyiiiir- are ^ivm for the obtusata-littoralis-palliata group 1 i.uit/riiburi; .UK! Fischer (1915 .md lor vixatilis-rudis l>y Dautzenburg and hei ' 1912). SPECIES CONCEPT IN LITTORINA 227 doubtful, all of which leads us to the impasse of a difference of opinion among experts. This confusion is due solely to the fact that inferences about varieties have been drawn from individual specimens, and the only way to unravel this tangle is to examine a large enough number of animals until the answer to the problem ceases to be a matter of opinion. Specimens of L. obtusata were therefore obtained from eleven different localities, on both sides of the Atlantic Ocean. The minimum number aimed at was one hundred from each locality, but sufficient TABLE I Localities, number of individuals and catalogue numbers of eleven lots of L. obtusata. M.C.Z. = Museum of Comparative Zoology, Harvard University. B.S.N.H. = Museum of Boston Society of Natural History. Locality No. M.C.Z. B.S.N.H. a Bergen, Norway 100 47508 b Cattewater, Plymouth, England 100 76980 c Church Reef near Plymouth, England 100 76978 d Near Westerly Rhode Island 100 13980 e . South Cohasset, Massachusetts 100 26655 f Briar Neck, Gloucester, Massachusetts 86 24787 0 Rye Beach, New Hampshire 75 26806 1, Broad Cove, Georges River, Cushing, Maine .... 38 67775 i Port Clyde Knox County, Maine 74 67773 j Tenants Harbour Knox County, Maine . . 60 67774 k Isle au Haut, Maine 100 13972 Total 933 specimens were not always forthcoming. In only one case, however, was the number less than sixty. Table I gives the localities of the lots, the number of individuals, and the catalogue numbers in either the Museum of Comparative Zoology, Harvard University, or the Museum of the Boston Society of Natural History. The positions of the localities are shown in the maps, Figs. 1 and 2. The lot a in Table I were kindly sent to me by Professor Brinkmann of the Bergens Museum, lots b and c by the Marine Biological Associ- ation at Plymouth, England, lots i andj by Mr. N. W. Lermond, and lots e, f and g were lent by Mr. C. W. Johnson of the Boston Society of Natural History. The rest, d, It, and k, were already in the col- lection of Harvard University. Each shell was measured with callipers to the nearest tenth of a millimeter along three dimensions; the length (a) of the final whorl 228 60' JOHN COLMAN 40" 20° 60' 40° FIG. 1. Map showing the localities on the t\vo sides of the North Atlantic. The shaded area is that shown on a larger scale in Fig. 2. The letters refer to Table I. 7 I* 70" 69' 68* 44' 43' 42° 4 I' FIG. 2. .M,i]> the luc.ilitics on the New England Coast. The letters refer to Table I. SPECIES CONCEPT IN LITTORINA 229 a FIG. 3. Shell from Church Reef, near Plymouth, England, showing the dimensions that were measured. (usually the overall length of the shell), the breadth (b) of the final whorl at the base of the penultimate whorl, and the distance (c) from the top of the spire to the most distant point on the lip. The ratios between these dimensions, -r , -and -r , were calculated, and they form the basis of comparison between shells from different localities. Figure 3 shows exactly which dimensions were measured on each shell. TABLE II Shells from the two sides of the Atlantic: numbers of shells of different sizes. (See Fig. 4.) Localities Bergen Cattewater Church Reef Westerly Number ot Shells 100 100 100 100 Millimeters a b c a b c a b c a b c 5.75 to 6.25 .... 2 8 25 37 19 8 1 2 25 30 39 4 1 5 11 21 35 19 7 1 7 17 21 26 14 9 4 2 1 9 28 34 13 13 2 1 1 7 21 30 21 12 4 2 2 1 2 17 22 28 20 7 2 1 5 30 31 29 5 1 1 8 15 17 33 14 9 1 1 7 29 28 19 9 8 1 16 43 29 10 1 1 6 33 27 18 9 5 1 6.25 to 6.75 6 75 to 7.25 . . 7 25 to 7.75 7 75 to 8 25 8.25 to 8.75 .... 8.75 to 9.25 .... 9.25 to 9.75 9.75 to 10.25 10 25 to 10.75 10.75 to 11.25 11.25 to 11.75 11.75 to 12.25 12.25 to 12.75 . . . 12.75 to 13.25 . 13.25 to 13.75 . . 13 75 to 14.25 14 25 to 14.75 14.75 to 15.25 15.25 to 15.75 15.75 to 16.25 .... 16.25 to 16.75. . . . 16.75 to 17.25.. . 230 JOHN COLMAN TABLE III Shells from the two sides of t lie Atlantic: numbers of shells of different proportions. (See Fig. 5.) Local Bergen Cattewater Church Reef Westerly Nuinln.T of Shrll.- 100 100 100 100 !'• portions a a c a a <• a a c a a c b c b b c I, 6 c b 6 c b 0 99 to 1 03 1 7 38 27 21 5 1 4 20 52 20 4 1 7 29 33 18 Ki 2 1 15 45 26 12 1 1 4 17 51 21 5 1 1 3 18 45 22 7 1 2 1 3 6 26 31 23 7 4 3 35 47 12 3 1 2 5 21 22 34 7 7 1 1 13 24 36 17 / 1 1 1 15 53 27 4 5 12 32 28 17 5 1 1 n; t<> 1 (17 1 (17 to 1 11 1 11 to 1 15 . 1 IS to 1.19 1 19 to 1.23 . . L.23 to 1.27 1 ~>7 to 1.31 1 31 to 1 35 1 35 to 1.39 1.39 to 1.43 . 1.43 to 1.47 1 17 to 1 51 1.51 to 1.55 . . . 1 55 to 1.59 . . . 1.59 to 1.63 ... 1.63 to 1.67 a a c b c b 1.518 1.079 1.407 It represents a shell from Church Reef, Wembury Bay, and the measurements in millimeters and the proportions calculated therefrom are as follows: Length (a) Breadth (b) Spire to Lip (f) 16.4 10.8 15.2 SlIKLI.S ] ROM Till'. TWO SIDES OF THE ATLANTIC OCEAX \Ylien shells from Bergen, the Cattewater, Church Reef, and Westerly are compared, they arc found to be so similar that it is impossible id separate them. In average size of individuals the four lots vary considerably, tho-i- from the Cattewater being the smallest and those from Church Reel, only six miles away, the largest. Among the Cattewater shells, however, there were nine very large individuals illustrated by No. 6 in Fig. 6. These presinnablv represent a previous generation or OUp, .md they h,i\e n»i been included in the Cattewater shells in Tables II and III and in Figs. 4 and 5. Table III and Fig. 5, SPECIES CONCEPT IN LITTORINA 231 giving the numbers of shells at the different proportions, show the amount of overlap between the four lots. Bergen and Cattewater shells are very similar, and the curves for Church Reef and Westerly are well-nigh identical. The Cattewater lot are somewhat more globular than the others, but they are probably younger, if the giant shells like No. 6 are really of the mature adult size. These nine large Cattewater shells have apparently lived an exceptionally long time, for they are the largest Littorina obtusata in the extensive collection at Harvard, with the exception of an individual specimen from Cornwall. Id I CO U- O 6 40 . 30 . 20 . 10 . 30 - 20 - 10 . 30 . 20 - 10 - 40 - 30 . 20 . 10 . BERGEN ..©' CATTEWATER n- -t* WESTERLY 10 1 1 12 13 14 15 16 17 SIZE IN MM. FIG. 4. Shells from the two sides of the Atlantic: numbers of shells of different sizes (See Table II). - = a, • • • == b, - = c. In Fig. 6 are shown camera lucida drawings of the obverse and reverse aspects of three shells from each locality, the relative height of the spire to the rest of the shell increasing from left to right. The dimensions and proportions of the same twelve shells are given in Table IV, where the ratio - decreases in each group with the increase of the spire component, c. From these figures and tables it will be seen that the amount of variation is much the same in each locality and that it would be quite impossible to sort a mixture of shells from the four places, either on 232 JOHN COLMAN or after measurement. There is then no doubt of the con- specificity of Littorina obtnsata from Bergen, from the Plymouth district and from Rhode Island. CO 50 . 40 - 30 - 20 . 10 - 50 40 30 20 ,0 30 . 6 2° - ^ ,0. 50 . 40 - 30 . 20 _ 10 . 1.07 • ' • ' 1_ 1.23 1.39 111 1.55 11 — i - 1 BERGEN CATTEWATER CHURCH REEF WESTERLY PROPORTIONS l;Ki. 5. Shells from the two sides of the Atlantic: numbers of shells of different a it proportions (See Table III). = 7 , • • • = - , -- 0 C c , 0 SHELLS IK<»M MM-: \i.\v ENGLAND COAST It has now been shown that the same species occurs on both sides of the Noiih Atlantic. It remains to be proved that the same species extends up the American Coast, where shells called " L. palliata" or SPECIES CONCEPT IN LITTORINA 233 FIG. 6. Shells from the two sides of the Atlantic: obverse and reverse aspects of shells whose dimensions are given in Table IV. TABLE IV Shells from the two sides of the Atlantic: dimensions and proportions of shells illustrated in Fig. 6. Locality Length (a) Breadth (« Spire to Lip (c) a b a c c b 1 . Bergen . mm. 14.1 mm. 10.5 mm. 12.8 1.343 1.102 1.219 2. Bergen 14.1 9.9 13.0 1.424 1.085 1.313 3. Bergen 13.8 9.7 13.0 1.422 1.062 1.340 4. Cattewater 11.5 8.3 9.9 1.386 1.161 1.192 5. Cattewater. . 12.5 9.3 11.7 1.344 1.069 1.258 6. Cattewater . 19.1 13.0 18.9 1.469 1.010 1.454 7. Church Reef... . 8. Church Reef.... 9. Church Reef.... 10. Westerly 14.9 15.5 13.7 12.4 9.5 10.5 9.1 8.3 12.9 14.6 13.5 11.3 1.569 1.476 1.505 1.494 1.155 1.061 1.015 1.097 1.358 1.391 1.484 1.361 11. Westerly 13.2 9.0 12.3 1.467 1.074 1.367 12. Westerly 11.2 7.9 10.5 1.418 1.067 1.329 234 JOHN COLMAN " L. obtnsata pallia ta" occur from Long Island Sound to Newfound- land. Between shells from Westerly, R. I., and Isle au Haut, Me., there is a considerable difference, namely that between the varieties littoral-is and f>C in nomenclatural complexity. SPECIES CONCEPT IN LITTORINA 235 OO bio ai OJ c/5 =0 H iJ 03 < s s -Jg *« a "-^ (^ Jj v «O CM OO OO OO i— i ^H ro oo ^f w " c 8 - Tf Ol O C^l fNI r<5 10 ^— i "To M a 0 0 t- O O i— < )H Xl HH - W oo « * ;* « o rs .1 eJS o o -o o o oo 10 VT-I vrH CSI TH c OJ H o l-H l-H 1— 1 ro-.es cu - IO I—1 v^ ^2;t~^~- ?! 2 -o CN ^ OO O 1-- fO CN CN 1-1 o OH 0 *" T-H *O CD t^* VO ^O *O u O i— i CN CN s - i— I CO CN * a ^^^ ^ - s 2 ^ -* ^ & - CM ro .0 ^ « 22^^^^^^ £ « i^ 'fs ll vO so -o ^H CN CS ra o ^CM^VO T-I CN i-l 4-1 VO ^-i OJ CN ^— i ^^ c*3 rf O <^ § JS (*O "^ ^0 ^^ f*O 0s! I-H ro <^ ^— i o Q CO ^O O\ IO O 00 -* Xs VO ^0 (*3 e>i I-H J-S o o -O t-i sO <-O ON O i-i "1 « t— ON OO ON Ov OO CN CN i-H "3 CO 1-1 V ££££££££££ IO *O ^O ^O L^ i/^ IO ^O t/3 1 iO^O\Ol>.t^OOOOONONO O TI —< tN 01 "0 ^0 Tji u o O = oooooooooo oooooooo ^ cQKcQI^cQScQKcQIC l_O <-O *O LO *-O *O ^O IO lOiOOxOl^t^OOOOOOv 00— (^c^rsv-oro 236 JOHN COLMAN c/. - 0) o / - _ C- -5 < "C ^1 a a s •'' O "o/> R w >-> I -a fv, . ^^ ^ ^ ~ ^. rr ~ 3 rt CM — ~ 1 ^ l •- • — "T ^ "7 t: 1 -i esicnovoo*-.'*,'*^ J: — . ^ J -a ^— i f^ J-i^ ^fo *?f* ^\ f^ ^s) ^-+ r~ .— B.S - ~ s C — o cc 10 o H a i -c - "--^ g 3 - o 7 . v- i -a ^— rq cc ic t^ -t O »c *— '— -f a i ^ - a l •« ^, ^ ^ oo oo - - > »,- rN c> ^i i- >c c^j ^H l'| X f> a i u fO O iO *^ a i -= I-H ro i --.»—• O U ~ "x u I -O CC 0 ^ ro ,- <*> - «l ;c- a i v> " g^2^ ~ i a i -a — "". — " "! X -f •— 1 X ,,. CS C CC C *S C « 1 ^» •— ""- "". i - -f — ir, r-| •3 ! -a -^^^-'--- 2 - L, X 1 "o •C1 C" ^ ~ " *— • *-"• * — ' ^ l r 1 rf~f f^ r'j *^« -^* 10 tO 'O ^O ^D l^* I"** ' "- 0 J = ooooooooooooooooooooooo 3 X 'xc^g^^SSSSSSSSSSSSSSE SPECIES CONCEPT IN LITTORINA 237 As a final demonstration of the essential unity of Littorina obtusata over this considerable geographic range, embracing, as it does, both southern Norway and New England, the data of all the 933 shells from the eleven localities may be added together. This is done in to LJ X to o 6 50 40 30 20 10 20 10 20 10 20 10 30 20 10 20 10 30 20 10 40 30 20 10. ISLE AU H A u T O o TENANTS HARBOUR ~i — — I — — r PORT CLYDE Q T 1 V" BROAD COVE -o— -ft RYE BEACH O .©•" G---E COM ASSE T WESTERLY .9 6 7 8 9 10 11 12. 13 14 SIZE IN MM. FIG. 8. New England shells: numbers of shells of different sizes (See Table V). — = a, • •• = b, --- = c. Table VIII, and the result shown in Fig. 11. Although the chief distinction between so-called varieties of L. obtusata is usually the variation in the relative height of the spire, the curve •; in Fig. 11, t/ JOHN COLMAN giving the ratio of Length to Distance from Spire to Lip, is an almost perfect frequency curve, because the series from blunt to elevated 40 30 20 10 . 20 . 10 - 30 - 20 - 10 , 10 . I/) 3° ' J 20 . I '0 - u. 0 30. o 20 • Z 10 . JO . 30 . 20 . 10 . 50 . 30 . 20 . 10 . FlG. <). Xc\V 1 Table VI). 0 spires is com in ohtiixtitti , accordi l 1 1. o ISLE AU H AU T •6 & J Vv n 00 a 0— O D^fl- ^ p — o TENANTS •• , /3k \^ HARBOUR o JQ^ORT CLYDE / o- -°-/ txBROAD COVE o yR RYE BEACH 0 o ff y \ p BRIAR NECK 0 P- O >o-9 i i i , i - - COHASSET / ° WESTE RLY P" /\ — i — i — n EL ,^" rf^i T-i — r^ 1 1 1^ ' I1"1 — 0-91 in 1-31 ,5, ,.7I PROPORTIONS Cn^land shells: numbers of shells of ditti rent jiroportions (See a c = ~c' ' =b' nous, with no sudden jumps. The "average" L. 11; to these figures, will have the proportions: i a c ) ~C I 47 1.07 1.37 SPECIES CONCEPT IN LITTORINA 239 31 28 ^^ 22 16 13 10 32 29 26 23 20 17 14 1 1 33 30 27 24 o 18 O 15 12 FIG. 10. New England shells: obverse and reverse aspects of shells whose dimensions are given in Table VII. In conclusion, I wish to thank Mr. \V. J. Clench, Curator of Mollusks in Harvard University, for his advice, criticism and kindness. 16 240 JOHN COLMAN TABLE \"II New England sliflls: dimensions and proportions of shells illustrated in Fig. 10. Locality Length Breadth W Spire to Lip (f) a 6 a c c b 10 \Yesterlv mm. 12.4 mm. 8.3 mm. 11.3 1.494 1.097 1 361 1 1 Westerly 13.2 9.0 12.3 1.467 1.074 1.367 1 ~> Westerly .... 11.2 7.9 10.5 1.418 1.067 1.329 1 i ^initli ( oh asset 10.4 7.0 9.8 1.486 1.059 1.400 14 ^outh Cohasset 10.3 7.0 9.9 1.471 1.040 1.414 IS ^i >uth ( "ohasset 9.3 6.0 9.0 1.560 1.033 1.500 16 Hi "i.ir Neck 11.8 8.0 10.8 1.475 1.092 1.350 1 7 Briar Neck 13.0 9.0 12.4 1.444 1.048 1.378 is Hriar \eck 14.2 9.4 14.0 1.511 1.014 1.489 19. Rye Beach 12.8 8.7 11.7 1.471 1.094 1.345 >0 Rye Beach 11.7 8.1 11.1 1.444 1.064 1.371 21. Rye Beach 13.0 8.5 12.4 1.530 1.04S 1.459 22. Broad Cove .... 12.9 8.9 11.9 1.450 1.084 1.337 23. Broad Cove 11.7 8.0 11.7 1.462 1.000 1.462 24. Broad Cove 11.5 7.7 11.9 1.494 0.966 1.545 25. Port Clyde 12.4 8.5 11.5 1.459 1.078 1.353 26. Port Clyde 11.9 7.9 11.2 1 .506 1.063 1.418 27. Port Clyde 12.4 8.7 12.6 1.425 0.984 1.448 28. Tenants Harbour . . 10.0 7.0 9.7 1.429 1.031 1.386 29. Tenants 1 larbour . . . . L0.2 6.9 10.2 1.479 1.000 1.479 30. Tenants Harbour . . . 9.3 6.2 9.7 1.500 0.959 1.564 31. Isle au I laut 10.5 6.8 9.8 1.544 1.072 1 .44 1 32. Isle au Haut 10.5 6.8 10.3 1.544 1.020 1.515 33. Isle au I laut 11.5 7.1 11.7 1.619 0.983 1.647 CO 200 _ 100 _ o a_ b o , o * s* O 0-91 1^ 1^ I 11 1-31 I 1-51 I 1-71 PROPORT IONS FIG. 11. Nine Imti'lnil .m, 1788. S\-,tema Naturae. 13th Edition, by J. F. Gmelin. Lipsiae. MARINE BIOLOGICAL ASSOCIATION, 1931. Plymouth Marine Fauna. Plymouth, England. MA-ION, \\II.I.I\M GEORGE, 1797. Observation-, relative chiefly to the Natural History, l'i< t un-sque Scenery and Antiquities of the Western Counties of I upland, made iii lite years 17'H and 1 7°6. 2 Vols; J. Eaton, Salisbury, England. Vol. I, ]>. 277. M\ION, \\'HI.I\\I (,idi«,i, \ND 'lix>\i\-, l\ \( KI IT, 1S04. An Historical Account o| | esta< eoloi;ical \\'riters. Truiis. Linn. Soc. London, 7: 119. MA-ION, WII.I.MM < rEORGE, \NH THOMAS RA< KKIT, 1807. A Descriptive Catalogue of the British 'I'e^iace.i. Trans. Linn. Soc. London, 8: 17. Mi NKI;, CARL THKODOK, 1S2S. Synopsis Methodica Molluscorum. Pyrmont. Mi NKI , C\KI. TIIKODOR, 1830. Syno]>sis Methodica Molluscorum. 2nd Edition. SPECIES CONCEPT IN LITTORINA 243 MONTAGU, GEORGE, 1803. Testacea Britannica, or Natural History of British Shells, Marine, Land and Freshwater; including the most minute: syste- matically arranged and Embellished with Figures. London. OLIVI, ABATE GIUSEPPE, 1792. Zoologica Adriatica. Bassano. SAY, THOMAS, 1822. An Account of some of the Marine Shells of the United States. Jour. Acad. Nat. _SY/. Phila., 2: 221-248. SCHMIDT, JOHANNES, 1931. Eels and Conger Eels of the North Atlantic. Nature, 128: 602. SHERBORN, CHARLES DAVIES, 1927. Index Animalium. Vol. 2, 1801-1850. Lon- don. WINCKWORTH, R., 1922. Nomenclature of British Littorinidse. Proc. Mala. Soc. London, 15: 95. INHERITED irARIATION ARISING DURING VEGETATIVE RK !'!« )DUCTION IN PARAMECIUM AURELIA 1 DANIEL RAFFEL2 (I:rom the Zoological Laboratory of the Johns Hopkins I'nircrsity and the Marine Biological Laboratory, Woods Hole, Mass.) CONTENTS Introduction 244 Materials and Methods 245 Description of the altered and unaltered lines of the clone 245 1. Origin 245 2. Comparison of the unaltered and altered lines 246 a. Size 246 b. Form 246 c. Fission rates 247 d. Resistance 24S Experiments 248 1. Regularity in the production of altered lines 248 2. Effect of conjugation in diverse lines 251 Discussion and Conclusions 254 Summary 256 INTRODUCTION Jennings (1(K)8), and Jennings and Hargitt (1910) have shown that populations of Paramecium which are produced from a single individual are uniform in size, form and fission rate. Other investigators have since confirmed these findings in other genera of Protozoa. Increased variation and the production of different stocks have been found to be produced by conjugation by Jennings (1913), RafTel (1930), Jennings, Raffel, Lynch, and Sonneborn (1932). Other investigators, Middleton (1918), Mast (1917), and Jollos (1921), have found inherited variation arising during vegetative reproduction which they attributed to the act inn of environmental agents. However, most investigators agree that when a clone of Puniiiicciitin is cultivated under uniform conditions it is usually decidedly uniform. An excellent review of the literature :< -i-ently been given by Jennings (1()29). 1 The present contribution is one of a scries <>f studies on the genetics of con- •atiun and reproduction in I'mto/oa, in progress by H. S. Jennings, his associ- •xl students. The author wishes to record his indebtedness to Professor Jennings for his interest and advice throughout the course of the work. - Fellow of the National Research Council. 244 INHERITED VARIATION IN PARAMECIUM 245 This paper presents the results of a study of a clone which sporadi- cally produced branches, or lines, which differed markedly from the original type. These lines, which will be described in detail later, dif- fered in many respects from the unaltered lines and these differences persisted for long periods under identical conditions. None of the altered lines was ever observed to revert to the original type although some were cultivated for nearly 100 generations. MATERIALS AND METHODS The organisms used in this investigation were members of a clone of Paramcciuui aurclia descended by fission from a single ex-conjugant (known as 128a) obtained July 15, 1930, in the course of another investigation. The methods of cultivation employed were the same as those used in the author's (Raff el, 1930) earlier work except that it was found necessary to add a small quantity of the bacteria to the fresh culture fluid each day. The culture medium consisted of a fluid containing KNO3, 0.5 gram; K,HPO4, 0.06 gram; MgSO4.7H2O, 0.02 gram; FeCl3, 0.001 gram ; H2O, 1 liter. To this solution cultures of an alga, Stichococcus bacillaris, and a bacterium, usually Achromobacter candi- cans, were added. The paramecia were cultivated in this suspension under bacteriological conditions. DESCRIPTION OF THE ALTERED AND UNALTERED LINES OF THE CLONE 1. Origin. The clone 128a differed markedly from the other clones isolated from a group of individuals undergoing conjugation July 15, 1930. It was found to have a characteristic form and size quite diverse from the form and size of the other clones which were studied. It also differed from most of the other clones in its slow rate of reproduction. For these reasons it was selected as one of the few clones to which was devoted further intensive study. Later study has shown that it also differed from most of the other clones in its tendency to give off branches quite different from the main clone and in its ability to withstand ad- verse environmental conditions. On August 4, 1930, twenty-two days after this clone was isolated, two of the twenty-four lines which were being cultivated became radi- cally altered in ways to be described below. Three other lines became altered in a similar manner during the next three days. These five lines were discarded on August 15. During the five days following August 29, seven more lines became altered in a similar manner. These were 246 DANIEL RAFFEL cultivated until September 18, when they were discarded. No other altered lines appeared until early in November, but from that time until June of the following year, they were produced very frequently. 2. Comparison of /he unaltered and altered lines. Tin' typical individuals of this clone, hereafter referred to as the unaltered lim •>. differed from the altered lines in the following ways: (a) Size. One of the most striking differences between the two i; mil ps was the difference in size. This difference is well illustrated by Fig. 1. which shows typical specimens of each group under identical FIG. 1. Camera lucida dra\vin.us of typical individuals of (a) unaltered and ( /' ) altered lines. conditions. As can be seen from this figure the individuals of the altered lines were more than twice as long as those of the unaltered lines. This marked si/.c difference was a matter of continuous daily observa- tion. (b) Form. The two branches of clone 128<; differed markedly in form. The unaltered individuals of this clone were thick in the central region and tapered to points at each end. They were quite pale and u-ually had a dark food vacuole in the anterior tip. The altered indi- viduals were, on the other hand, of the typical P. aurclia form, i.e., somewhat pointed anteriorly rounded posteriorly with their greatest breadth about two-thirds from the anterior end. Usually the}- had many food vacuoles distributed chiefly in the posterior part of the or- ganism. These differences in form as well as the size differences are shown clearly in Fig. 1. This figure shows typical adult individuals of INHERITED VARIATION IN PARAMECIUM 247 both kinds living under identical conditions and drawn at the same time. The differences in form and also the differences in size persisted even during the periods of depression when the fission rates of the two groups approached one another. At these times the two groups could be readily distinguished by their appearance. (c) Fission Rate. The altered lines differed from the unaltered lines greatly in their rates of reproduction. The clone originally was a very slowly dividing one with a mean fission rate much below one divi- 3. Or July Au^. Sept. Oct. Nov. Dec. Jan Feb. March April May FIG. 2. Average daily fissions of unaltered (solid line) and of altered (dotted lines) lines plotted in periods of approximately five days each. sion per day. The altered lines, on the other hand, divided very vig- orously and some of them averaged well over two fissions per day for periods exceeding two weeks. This difference in fission rates is illus- trated in Fig. 2. The solid line in Fig. 2 shows the mean daily fission rates for the original clone from July 15, 1930 to May 17, 1931, plotted in approximately five-day periods. The dotted lines represent the av- erage daily fission rates of six different groups of altered lines which 248 DANIEL RAFFEL were studied. These are also plotted for the most part in five-day periods. From this figure it is evident that the altered lines differ markedly from tin- unaltered lines in their rates of reproduction. (d) Resistance. The altered lines of clone 128a were much less resistant to unfavorable- conditions than were the unaltered lines. This can be readily observed from Fig. 2 which shows that in nearly every period of depression the altered lines suffered a decrease in fission rate earlier than the unaltered lines, and that the extent of the depression was greater in the altered lines. For example, group (c) Fig. 2 declined in fission rate from 1.26 to 0.43 fissions per day between January 10 and January 20, while the unaltered lines showed an actual increase from 0.88 to 0.93 fissions per day during the same period. Similar though not so great differences can be found by comparing groups (rf), (c) and (/) with the unaltered lines which were cultivated simultane- ously with them. \Ye have seen then that the clone 128a produced branches during vegetative reproduction which differ markedly in size, form, fission rate, and resistance from the unaltered lines. The differences in size and form persist even during periods of depression during which the fission rates of the two groups were similar. These differences are as great as any found by Jennings, Raff el, Lynch, and Sonneborn (1932) in their extensive study of diversities between biotype-s produced by conjugation. The altered lines persisted for long periods and none has ever been ob- served to revert to the original condition although many thousands of individuals have been examined. EXPERIMENTS 1. Regularity in the production of altered lines. The remaining parts of this papiT di-al with experiments designed to discover whether any regularity was apparent in the production of these altered lim^, and whether after conjugation these unlike branches of the same clone would produce similar or dissimilar populations. In order to learn more about the occurrence- of the- altered individuals and to • •rtain whether then- is any regularity in their production, the follow- ing experiments were performed: From November 3rd to 19th the progeny of one individual of the clone was expanded to form 33') lines of descent whose relationships were known. During this period some of the lines became altered. These, because of their faster rate of reproduction, gave rise to the greater part of this population. After the desired number had been obtained, the organism-, wen- cultivated under identical conditions for INHERITED VARIATION IN PARAMECIUM 249 ten days. During this period the lines which were altered at the be- ginning passed through from 1-21 generations (all but one of the 226 such lines which lived until November 29 passed through from 11-21 generations), while the unaltered lines passed through from 4—9 genera- tions. As is shown in Fig. 2 (c), the mean daily fission rate for the altered lines is 1.67 as compared with 0.83 fissions per day for the un- altered lines. On November 29 all except one of each of the altered and unaltered lines were discarded. Then each of these was expanded to a population of 168 individuals. These two populations wrere then cultivated for ten days from December 11—21, during which time they showed great di- versity in their size, forms and fission rates (sec Table I). Although TABLE I Distribution of total fissions December 12-21, 1930, of populations derived from (a) a single individual which manifested the altered set of characteristics on November 18, and (b) from an individual of the original type. Fissions 0 1 ? s 4 S 6 7 8 <) 10 11 P 1S 14 IS 16 17 18 Mean Altered lines 1 1 1 3 3 7 7 7 -» o Q 1? 1? ?4 n 10 4 13.1 Unaltered lines . . 0 1 5 10 1? 17 18 8 6 5 5 5 6 5 4 3 2 7.2 the altered lines were much depressed at this time because of unfavor- able environmental conditions or endomixis, the population derived from the altered individual produced no individuals which were like the unaltered lines. Figure 3 is a genealogical chart showing the relationships of the lines descended from the unaltered individual, with their numbers of fissions during the ten days following December 11. These lines were all of the original type on December 11; but, as is shown, many of them, repre- sented by the filled-in circles on Fig. 3, became altered subsequently. From the data given in Fig. 3, an attempt was made to ascertain whether or not there is any regularity or rule in the production of altered stocks. That is, if a given line is altered, are the lines which are sep- arated from it by fewqst fissions more likely to be altered than lines separated by more fissions or not ; or is any other rule to be found which would describe the production of the altered lines? To investigate this question the coefficients of correlation were obtained for the total num- ber of fissions between December 12 and December 21, of lines repre- sented on Fig. 3 which were separated on December 11 by one, two and three fissions. The lines which were separated by but one fission at this time showed a correlation of 0.56 ± 0.06; those separated by two fis- sions a correlation of 0.23 ± 0.12; while those separated by three fis- 250 D AX I EL RAFF EL ' " _; "3 C g o P O ~ u rt c: j= •• -^ E « l°gg V " O ^ *o g - " .> § °-S •o » _ X X 2 X 1 1 s 7 X 8 6 J *O ts. IE C/3 U L __2 *rt o u o '• "u 'o «-" a 0 -i 12 -*-* — _o ^ 13 12 7 J2 "o IN - H -o 1> •4-> o 1 X ^~| o 4 O *^ o u • ^ ej • — . ' 16 L 14 ' 0 -*-• rt "o Niivfinbc V r i v i- rt •a o u •^ 4-1 _ -- _ *5 •*-• / ; - 0 X -; - -— B & X • • X r2 'So ct \veri P u Q •a o -»-» . B "c^ — '.-5 rt "r 6 o — = E — -;; X o 2 41 w • C/l 6 o -v • — u ~ . V y X ''-:_ i ~ '-i 0 ,\z rt _ U 1 - _ J= ^. INHERITED VARIATION IN PARAMECIUM 251 sions gave a coefficient of correlation of 0.48 ± 0.06. Thus there is apparently no relation between the number of fissions by which lines are separated and the fate of the lines. The decrease in the coefficients of correlation of the lines separated by two fissions as compared to those which are separated by only one appears to be meaningless when we find that those separated by three fissions do not differ significantly from those separated by only one fission. A careful study of the records, however, shows that when two lines are derived from a common an- cestor, one or two generations before one of them becomes altered the other usually becomes altered in a similar manner at nearly the same time. This occurred frequently throughout the entire six months that this subject has been intensively investigated. The experiments and observations which were made with the object of discovering any regularity or rule in the production of altered lines did not give any evidence of any such regularity. They did show, how- ever, that if a line becomes altered, lines which were derived from the same individual not more than two generations previously generally be- came altered at the same time or at nearly the same time. 2. Effect of conjugation in the diverse lines. In another paper Jennings, Raffel, Lynch, and Sonneborn have shown that when two diverse clones conjugated they produced two very diverse populations. In the two clones intensively studied, although each population showed great variation in fission rates, the mean fission rate of each was not very different from the fission rate of the clone which produced it. In the particular cases which they studied the mean fission rates of the two populations differed in the same direction and to approximately the same extent from the mean fission rates of the clones from which they came. In the present investigation we have two races which differ more than the two clones studied by Jennings, Raffel, Lynch, and Sonneborn. These two races are, however, not different clones but are branches of the same clone ; that is, they are descended by vegetative reproduction from a common ancestor. Will these two very diverse races also produce by conjugation populations which are on the whole diverse, or will they, unlike diverse clones, produce populations which are similar? In order to answer this question typical lines of the two races of this clone were chosen and expanded in isolation drop cul- tures until large numbers of both types were available. Then conjuga- tion was induced in both branches of the clone and 105 pairs of con- jugants and 96 non-con jugants were taken in each group. In the case of the slowly reproducing unaltered lines care was taken to choose among the non-con jugants some split pairs and others which were be- 252 DANIEL RAFFEL ginning to conjugate. These four sets of organisms were then culti- vated side by side fur ten da\ s after the last ex-conjugants were isolated. The results of this experiment are shown in Figs. 4 and 5. Figure 4 shows the course of the mean daily fission rate of the two groups of non-con j ugants ; and Fig. 5 .shows the same for the ex-conjugants. As can be seen. the. non-con j ugants of the unaltered lines continued to re- produce at the rate of approximately 0.9 fissions per day, while the ex- conjugants from this group reproduced much more rapidly, reaching at the end of the experiment a rate of 1.6 fissions per day. This pro- gressive increase in fission rate was due entirely to the dying of the many abnormal lines which were produced by conjugation. Xinety- three non-viable lines were produced. Only 10 of 210, or 4.8 per cent, of the ex-conjugant lines resembled the parent race in fission rate. 2.0 1.5 i.o 0.5 2.0 1.5 1.0 0.5 J_L 8 10 12 14 16 18 20 FlG. 4 I I I I I I I I I I I I i i 6 8 10 12 1416 18 20 FIG. 5 I' u;. 4. Average fissions on successive days of non-conjugant lines of the un- altered (solid line) and altered (dotted line) groups. l-'ic,. 5. Average fissions nn successive- days of the ex-conjugant lines of the unaltered (solid line) and the altered (dotted line) groups. Among the % lines of the non-conj ugants, 29 became altered and 24 lines died out during this period. The results obtained from the altered lines were quite different. The non-con jugants at the- beginning were very healthy and vigorous, reproducing at the rate of two fissions per day the second day after conjugation and nearly as rapidly the next day. From then on they declined rapidiv in vigor — the fission rate falling to between 0.3 and 0.5 fission per day. accompanied by a great increase in mortality. The ex- conjugants, which were also very vigorous at first, with a fission rate of 1.4 for the first day, declined to approximately one fission per day INHERITED VARIATION IN PARAMECIUM 253 in spite of the death of the 90 non-viable lines which had been produced by conjugation. Because of the depressed conditions of the organisms during this experiment it seemed advisable to repeat it under more favorable con- ditions. Great difficulty was experienced in trying to obtain the con- jugants from the unaltered lines as these were constantly producing organisms of the altered type, which in a short time would greatly out- number the type which the culture originally contained. Finally, in order to know which type of organisms was conjugating in the cultures of the original group, split pairs were taken at the same time as the conjugants. Thirty-three per cent of all of the organisms obtained from split pairs were of the original type. However, no pair among the split pairs was composed of two organisms of this type. It is very probable that this was due in a great measure to the subsequent altera- TABLE II Distribution of fissions during the first five days after conjugation of ex-conju- gants derived from (a) altered lines and (b) unaltered lines of clone 128a. Fissions 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 No. of lines Mean Altered lines 4 0 1 1 9 1 1 Q 10 14 17 10 ?,4 ,1S 19 1 151 10.5 Unaltered lines. . . . 0 4 2 3 4 11 (. 6 8 8 19 18 38 46 24 19 216 10.9 tion of one member of the pairs, as altered lines wrere being produced rapidly at this time. The results of the conjugations in the two branches were as follows: The unaltered lines gave rise to a population very different from them- selves. The mean fission rate of 2.2 fissions per day is much greater than the 0.62 fission per day of the non-con jugants which lived through the same period. The ex-con jugants were somewhat less viable than the non-con jugants of this group; 22.3 per cent of the former died dur- ing this period as compared to 18.9 per cent of the non-con jugants. Only 11 of the 278 ex-con jugant lines or 4 per cent had the shape and size of the unaltered lines. On the other hand, the altered lines gave a quite different result. The ex-conjugant lines were much less viable than the non-con jugants; 34.7 per cent of the lines derived from ex-con jugants died, while there was no mortality in the non-con jugants. The mean fission rate of the ex-con jugants was 2.1 fissions per day as compared with 2.3 for the non-con jugants. No single individual among the ex-conjugant lines obtained from the altered group had the size or form of the unaltered lines. A comparison of the two groups of ex-conjugants with respect to their fission rates is given in Table II. This table shows the great 254 DAXIHI. RAFFEL similarity between the fission rates of the two populations which have the same mode and means which differ very slightly. These results differ markedly from those obtained by Jennings, Raf- fel, Lynch, and Sonm-born, on the effects of conjugation in their two diverse clones. The two populations produced after conjugation by the two diverse branches of the same clone, in the investigation presented here, are nearK identical with respect to their fission rates. The mean fission rates of both populations are nearly equal to the mean fission rate of the altered population. In the case of the altered lines this result is in accordance with the results obtained by earlier workers. Jennings (1913) and Raffel (1930) found that conjugation within a clone results in a population showing much variation, with a mean fission rate slightly lower than that of the parent clone, and an increase in mortality. In the case of the unaltered lines of this clone the effect of conjugation is to produce a population quite different from the non-conjugant popula- tion. The mean fission rate is increased to more than three times the mean fission rate of the non-con jugants and the size and shape of 96 per cent of the ex-con jugants is altered. Thus the unaltered lines of clone 128a are affected by conjugation in a way quite unlike most of the clones which have previously been described. It is obvious that the two very diverse branches of clone 128a are affected quite differently by conjugation. Instead of giving rise to two diverse populations as did the two diverse clones investigated by Jen- nings, Raffel, Lynch, and Sonncborn, they produced, after conjugation, populations nearly identical in fission rates and appearance. This indi- cates certainly that the inherited diversity which appeared during vege- tative reproduction in this clone differs in some way from the diversities between the clones studied by Jennings. Raffel. Lynch, and Sonneborn. DISCUSSION AND CONCLUSIONS The inherited variations which arose in this clone (hiring vegetative reprodnetion differ markedly from those which are produced by con- jugation. The altered lines which were produced were similar to one another in all respects and they were quite different from the original type, while conjugation produces from a single clone a large number of clones which differ much or slightly from one another so that a more or les^ continuous series is pn>dured. The differences between the two branches of this clone are to a great degree eliminated by conjugation while diver-e clones produced bv conjugation usually yield after subse- quent conjugations populations which are similar to themselves in their lission rates and other characteristics. This suggests that the basis of INHERITED VARIATION IN PARAMECIUM heritable variation found in this clone is different from that of the vari- ation produced by conjugation. The increase of variation after conjugation and the cytological de- tails of that process have led to the conclusion that conjugation involves biparental inheritance and that the increased variation produced by con- jugation is brought about by a recombination of genetic factors. The inherited variation reported in this paper, however, does not seem to arise from such recombinations. It differs from the variations produced in that manner as set forth above ; and in order to ascribe its origin to such recombinations it would be necessary to postulate some such process for this particular clone only. Therefore, it seems that recombinations must be dismissed as a possible explanation of the origin of this varia- tion. That the genetic constitution of the original clone must contain a number of heterozygous pairs of genes is evident from the results of conjugation between members of the unaltered branch. In order that a particular clone should produce, when inbred, a population only 4 per cent of which resembles the parent, it is necessary that the clone should be heterozygous for four or five genes. Therefore, we can assume that a heterozygous condition of four or five pairs of genes is necessary for the production of the unaltered type of this clone. A mutation then of any of eight or ten genes would produce an altered line in this clone. It is possible, however, that the observed altered lines could only be pro- duced by the mutation of any of the four or five " mutant " genes to the " normal " condition (as a mutation of the " normal " gene might be lethal when homozygous). The evidence leads to the conclusion that the production of altered lines in this clone is probably caused by the mutation of one of the members of the four or five pairs of heterozygous genes. This conclusion explains all of the phenomena observed. There is no regularity in the production of the altered line — no regularity would be expected of gene mutations. The altered lines never produce the original type after conjugation — as it is not heterozygous for all five pairs, it cannot produce progeny which are heterozygous for all five pairs. The only question which this conclusion raises is the frequency of gene mutation in Parameciuin. This is to be the subject of a subse- quent paper, but it might be said here that this explanation of the ques- tions raised by this investigation throws some light on other questions. Jennings, Rafifel, Lynch, and Sonneborn found that clones differ in their uniformity. Others have found variation between clones in mortality, etc. If gene mutations are comparatively frequent in Parainccium, it 17 256 DANIEL RAFFEL can account for tin.- cases of inherited variation reported by other work- ers as well as the unexplained mortality that occurs in all isolation cul- ture work. It. furthermore, can explain the differences in mortality which are observed in dilTerent clones. The only alternative to this ex- planation is that the original clone contained a detachable translocation which was lost in the production of the altered branches. The former explanation seems preferable because it explains the numerical relation- ships found in this investigation as well as the other phenomena which have been mentioned; and also because much evidence has subsequently been obtained which indicates frequent mutations in Paramecium. A more detailed treatment of these questions will be presented in the near future. SUMMARY The clone of Paramecium aurclia studied in this investigation pro- duces branches which differ from the original clone in many respects. They are larger, have a different form, reproduce at a greater rate, and are less resistant to unfavorable conditions than the original type. These branches consistently manifest their diverse characteristics and none has ever been known to revert to the original type in any of the cases which were studied. Xo rule or regularity was found in the pro- duction of the altered lines except that when a line became1 altered, lines which were separated from it by only one or two fissions usually became altered at the same time or nearly the same time. Unlike diverse clones, these durti<>n in I'\-«\»/M:I, in progress by H. S. Jennings, his associ- :id students. The authors \\Mi to record their indebtedness to Professor 1'imiiiL's fnr his inttn-t and advice throughout the course of the work. This investigation \\a-~ aided by a grant from the Bache Fund. 258 RACIAL DIFFERENCES IN CONJUGATION 259 as did conjugation in the less depressed lines. In other words, there was a positive correlation between the mean fission rates of the non- conjugants and the mean fission rates of the derived ex-con jugants. MacDougall (1931) has shown that morphologically different races of Chilodon uncinatus yield similar morphologically different groups of ex-conjugants. Jennings et al. (1932) have shown that in Parame- ciuni anrclia two ex-conjugant clones which differ markedly in mean fission rate yield groups of ex-conjugants which differ similarly in mean fission rate. Raffel (1932), on the other hand, working on a single ex-conjugant clone of the same stock of Paramecium aurelia employed by Jennings et al., found that two sub-clones with considerably diverse fission rates which arose from this single clone during vegetative re- production, yielded groups of ex-conjugants that were not significantly different. (It needs scarcely to be recalled that in the work of Jennings ct al., and of Raffel, great differences were found within each group of ex-conjugants; the values compared are merely mean values. See the original papers for details.) These results of Woodruff and Spencer, MacDougall, and Jennings et al. strongly support the idea that the effects of conjugation may differ in different lines of descent within the same species. Only in the work of MacDougall, however, has there been repeated demonstration in the same races, of the same differences in the effects of conjugation. In her work, the differences studied were morphological. There is on re- cord no thorough study of characteristic racial differences in the effects of conjugation on fission rate, variability or mortality. These physio- logical characteristics are the ones most intimately related to current theories of conjugation. Studies on characteristics such as these are required in order to determine whether the differences in results, and in theories constructed from these results, as set forth by different in- vestigators, are due partly or entirely, not to specific differences, but merely to racial differences in the stocks they studied. Might conjuga- tion result in rejuvenescence in some races of a species and not in others? Might conjugation result in great increase in variation in some races of a species and in little or no increase in variation in others? Might conjugation result in great mortality in some races and in little or no mortality in others? If conjugation in Protozoa is fundamentally similar to zygote forma- tion in higher organisms, then, as suggested by Jennings (1929), con- jugation should result in similar recombinations of chromosomes and genes. The present state of knowledge of the genetics of the Protozoa, unlike the state of knowledge of the genetics of higher organisms, does not justify the assumption of genes; but the works of Jennings (1911, 260 T. M. SOXXKBOKX AXD R. S. LYXCH 1913), Jennings and Lashley (1913, 1913a), Pascher (1916), Raffel (1930), and Jennings <•/ al. (1932) demonstrate that conjugation in Protozoa is a process resulting in l>i|>nrental inheritance and in the formation of many new hereditary types. If the nuclear recombina- tions in the conjugation of Protozoa further resemble the nuclear phe- nomena in gamete and zygote formation of higher organisms, racial dif- ferences in the effects of conjugation might he predicted. Races which differed in the number of heterozygous " gene " pairs (or whatever the corresponding determiners in Protozoa may be called) would differ in the amount of variation produced by conjugation. Races which differed in the number of heterozygous lethals would differ in the mortality pro- duced by conjugation. Races which differed in the genes determining fission rate would yield conjugants which differed in these respects; thus some races might be rejuvenated by new. favorable, chromosomal re- combinations, other races depressed by new, unfavorable chromosomal recombinations. The importance of the issue for current theories of conjugation, the seeming likelihood of obtaining a result throwing a light on the possible analogy between conjugation in Protozoa and zygote formation in higher organisms, and the fact that encouraging results had already been ob- tained both in our own work and in the work of others, induced us to study intensively this matter of characteristic physiological differences between races, in the effects of conjugation. Plan of Investigation The present investigation begins at the point reached by the results of the investigation of Pascher and of Jennings and his collaborators, mentioned above, and goes on to ask the following questions: 1. When the diverse clones produced by conjugation within a single clone are allowed to conjugate, each within itself, do they all produce .similar sets of new ex-conjugaht clones, or do the results of conjugation differ in the diverse parent clones? That is, does conjugation produce the same result in slowlv multiplying clones as in those that multiply rapidly? Does conjugation produce similar results in clones differing in other rcxpecN : 2. If different results are produced, is there a correlation between the characteristics of the parent clones and those of derived ex-conjngant clones? For example, dot's a slowly multiplying clone produce a set of ' "iijngants that multiply on the average less rapidly than the set pro- duced by a rapidly multiplying clone? 3. Do successive conjugations among the members of a single clone yield similar sets of ex-con jugant clones in each case? Or do different conjugations within the same clone yield diverse RACIAL DIFFERENCES IN CONJUGATION 261 4. Does the mean fission rate of the ex-conjugant clones always hear the same relation to the fission rate of the respective parent clone? Or may conjugation in some clones result in an increased fission rate, in other clones in a decreased fission rate? To answer these questions, the following were studied (compare the diagram of Fig. 1) : A. Comparison of the effects of successive conjugations in the same clone. In the clone employed (designated 247a), part of the individuals were allowed to conjugate, while others were not; the two sets were then compared. Later, another set were allowed to conjugate, and these again compared with the non-con jugants. This was repeated for a third time. B. Comparison of the effects of simultaneous conjugations in six different ex-conjugant clones, which had similar vegetative character- istics. C. Comparison of the effects of repeated simultaneous conjugations in two of the ex-conjugant clones that had shown diverse results in the first comparison (B). In all, four successive simultaneous conjugations of the two stocks were compared. D. Comparison of the effects of conjugation between the above two clones, with the effects of conjugation within each of these clones. This part of the work is but briefly reported here, as the matter is of such novelty and importance as to warrant treatment in a later separate pub- lication, in which will be given also the results of other crosses that have been made. In giving the results of these four sets of experiments, it will be observed that the data are mainly for the early effects of conjugation : for those manifested within the first two or three weeks after conjuga- tion. During this early period occur many of the most characteristic effects of conjugation, such as the dying out of weak races. As to the relation between the earlier and later effects of conjugation, the present paper makes no assumptions. A comprehensive study of conjugation requires a knowledge of both its earlier and its later effects. It is planned to have later contributions in this series of investigations deal explicitly with the later results of conjugation, as compared with the earlier ones. II. MATERIALS AND METHODS Throughout the course of the investigation, the paramecia were cul- tured in isolation, one paramecium to one drop of culture fluid. The culture drops were placed on double-concavity, hollow-ground slides. These slides were kept together in groups of twelve on a glass plate 262 T. M. SONNEBORN AXD R. S. LYNCH raised by glass supports above the bottom of a 10-inch inverted Petri dish sealed with distilled \V;IUT. These chambers of isolation cultures were examined daily, at which time one animal from each culture drop was transferred t» a fresh drop of culture fluid on another slide. At different times during the course of the work three different methods of preparing the culture fluid were employed; these may be referred t<> as methods A, />'. and C. Method A is the method de- scribed by Jennings ct ai. (1932). This consists of boiling 30 flakes of ( >uaker Rolled \\"hitc Oats in 100 cc. spring water for 3 minutes; al- lowing this to stand five minutes; filtering into 250 cc. Erlenmeyer tlask ; letting tilter and flask stand for 24 hours; then inoculating filtrate ( now rich in bacteria from air and glassware) with the alga Stichococcus hucillaris (grown on agar slants). Method B is the method described b\ Raff el (1930). except that precautions designed to exclude foreign bacteria were not employed. Raffel's synthetic fluid was inoculated, just before using, with the bacterium Achromobacter candicans (grown on agar slants) and the alga Stichococcus bacillaris (also grown on agar slants). Culture method C is merely a slight modification of method A. It differs from method A in two particulars: (1) The first part of the filtrate, which comes through the filter rapidly in the first five minutes, is rejected; (2) The 24-hour-old infusion is boiled and cooled just be- fore inoculating with SticJwcoccits. These modifications of method A were found necessary at certain times to keep within satisfactory limits tin- quantity of bacteria in the culture fluid. The paramecia used in the present study were all descendants of a single ex-conjugant. known as 247. consisting of 100 lines of descent from 50 l'a! onjngants, and the /: group beginning December 9, consisting RACIAL DIFFERENCES IN CONJUGATION 263 of 194 lines of descent from 97 pairs of conjugants, are represented in a similar way. From this E group were selected for further examination lines of descent from six of the ex-conjugants. These as shown in Fig. 1, were ex-con jugant clones 71400. E4\a, E46b, ESOb, ESla, and E85b. On January 12-14, conjugation was simultaneously induced within each of these six ex-conjugant clones. These conjugations, known as F, G, L, 247a n f 48 pairs "OaxElOo 48 POTS 48 pairs / / / 20 pairs r/ F.41a / 48 pairs / Q L| F46b / ' / 48 pairs / E46b * E46b / *( 20 pairs E40axE81a / E80b /48 pairs E80b«E80b of «' J~ E81Q j 48 pairs / '44 pare / '96 pairs / EfflaxESla /E8lQ*E8to JEBIaxEBlQ 21 pairs E81a«E81a E8,Q M( / ' 48 poirs 5 8 pairs 247oV247a 5O pairs 247ax247a E85b 97pairs 247Qx247o October 1 1930 October 16 247a 1931 Februory 9-12 Februaryl5-18 Marchll-16 FIG. 1. Pedigree chart of the clones used in this investigation; all clones were derived, as shown, from the clone 247a. The horizontal lines represent single ex- conjugant clones (descendants of a single ex-conjugant) multiplying vegetatively without the intervention of conjugation. An oblique line branching off from a line representing a clone indicates that a part of this clone was allowed to conjugate. Each group of conjugants is designated by a single capital letter to which an oblique line leads. The brackets following this letter enclose a set of horizontal lines representing some of the new ex-conjugant clones. Below these lines are given, in parentheses, the number of pairs of conjugants and the clone-names of the conjugating individuals. K, J, and H, respectively, each consisted of 48 pairs of conjugants. Two of the six clones, namely £40a and ES la, were kept for further comparisons. In these two clones, simultaneous conjugations were in- duced on February 9-12, February 15-18, and March 11-16. These simultaneous conjugations are designated P and M, Pl and Ml} and .5" and T, respectively. At the same time as the conjugations 6" and T were taking place within the races E4Qa and ESla, conjugation X was also taking place. This was a cross-conjugation between representatives of £40a and £Sla. In the course of the experiments, sixteen different sets of conjugants and their descendants were studied. These are listed, together with 264 T. M. SONNEBORN AXD R. S. LYXCH Culture method BQ cq - •^ - - - - ~ ^ ^ T- T- O ^ Cj .'I'i'i,-; in.i i • >.ij;i.>|> in JHUVJ .uiui'J.iilui.>x o CM CM 0 0 o "•. "~. C- >^l ~l -* o1 o' .; — rrj ° CM CM ^ 10 C ~, ^, ~, / a ^ a Q\ _ .^ -jj ol 1 O *— I cs _>, "rt en 3 4-1 .0 0 " O "T" o O O O O O 0' O — — — — — X "2 in 4-1 3 JO o ~ "?0 o o o o °5 0 10 10 10 10 | ^ 7 7 7 7 o1 o1 o o o1 o1 o1 o1 "> - « 2 -1 C C O CN t^ ?N .r: .-r .t: .— .r: — H ~ ~ ~ ~ ~ »*5 "5 f*5 "^O O CN •— • — CM CM CM rv) <^i -o ~ 3 - ''- .- = .= t- C — o o -* -HO CN O O O O O O CN ON CN CN CN CN OO O CM O O •* O OO CN O CN Vj OO OC i — i O 00 CN ^^ *^i r^) Tjl CO •* *J <_<*- ' e 3 ijlf - 00 g.2 •= 5 3 "~ ii 0 OO O r~- i/5 10 CN CO CC CC CC X CC v^l ^* ^« -^« -^1 -^1 •& 00 O cc O -H o -* •* CN -*1 CM CM CM 10 rf ON 0 0 O CM -H CM •o'S ^1-7 S > c^ IO ^O IO »O >O *O r^i fN rs cs <^i fN 1 1 1 I I 1 ^1 rx| ^] ^ 'r ^] Tf OO OO ^1 -H CM CM . CM | 1 | 1 -H 1 ^ 1 ^ | -H 10 10 "• CN'— IT— l»— 1 U U U £ T 0 • • • • .: o. .- a .• a — .2 O tj ti2 | o o - c = = = c c a cs cs « ri fs ^^^^ i=< rt< S < 4> O O O <— i — . — '~ U. ^- U. •< ^ r', C S -^ ^ ^ -t -t -t N 2^7 2 2 -•5 '. - 0 M "' T ^02 0 I I I I 1 ^1 fS <^l ^1 ^ CM *T 1 1 JL -H J, CJs ^ £ 12 ^' - - 3.2 ^o "". . . . . u 4J CN ^J u (j *"" (j Q> O 0 - C C- C C E C C ft ^^ rt rt C3 C3 rt .2 ji .2 j^ ~ a a CJ U O O LI L i L i U, U. U. U. ^ r', 2 — — - = <=.S« £&?H ^T Si a u q ^ ^ CJ ^ ^ ^ ^ ^ i, ^ C y: ^ ^ a a a c — ' - ' — 3 -^ ^ Clou \vliii-li omul H.mt ]>.iirs u. ir ilfrivi- a « •= i - i - i- -c -f -r ^1 CM r-i XX X a a •= i - i ~ i ~ -c — — '1 fN '1 a 3 -c> o -o -o o — 10 — o o -t -t cc vi :/: -t ^ <: ^ ^ <: ^ X XXXXX •= •= -c a -c. -o O — 10 — C O -f -f X) OO OO Tt« ^ ^ <; tt) tq !ii a a a 3 3 ^ OC »t CO -t -f OO OO tt) to} K]^^ til faj XXXXX X X a a e a a a a — ' C — O 0 ~ O OO •**• C« •* -f VJ ~* bq KJ Sj K] ft] UJ kl xxx ,20^ * -t oo Sb|bl rt rt n3 4-J 4-I -^J 0 0 O H H H 'J? 13 2^ ~ -t; < ^ H ^ 5 - RACIAL DIFFERENCES IN CONJUGATION 265 useful information about them, in Table I. There are given, in column 1, the names of the clones of the two conjugating individuals ; in column 2, the designation given to each particular group of ex-conjugants ; in column 3, the dates on which conjugation occurred; in column 4, the dates between which the ex-conjugants were studied; in column 5, the number of pairs of conjugants in each group ; in column 6, the number of lines of descent from each ex-conjugant ; in column 7, the total num- ber of lines in each group; in column 8, the temperature range during each experiment ; in column 9, the culture method employed. Figure 1 and Table I are intended to give the general plan of the investigation and many of the pertinent details. At the beginning of each of the subsequent sections of the paper it may be helpful to refer both to Fig. 1 and to Table I. III. THREE CONJUGATIONS WITHIN THE CLONE 247a Most of the necessary information concerning the origin of clone 247a and the details of the three conjugations within it have been given in the preceding sections. (See especially Fig. 1 and Table I.) With descendants of the 58 pairs of conjugants obtained on October 1-2, the twenty-four co-existing non-conjugant lines were compared during two successive five-day periods (December 18-22 and December 23-27). Table II gives a detailed picture of the distribution of total numbers of fissions obtained in these experiments, and shows some relations that will be more precisely formulated by statistical methods in other tables. It is at once apparent that the range of fission rates of the ex-conjugants in every period exceeds that of the non-con jugants. Less obvious is the fact that the values for the ex-conjugants are less concentrated within the modal range than are the values for the non-con jugants. Calculations based on Table II show that on the average, only 73.3 pel cent of the ex-conjugant values for five-day periods are concentrated in the three most frequent classes of values, whereas 93.1 per cent of the non-conjugant values are within their three most frequent classes. Table II also shows clearly that the numbers of fissions in five days vary greatly from period to period in both groups. Taking the modal values as an illustration, the ex-conjugant mode drops from 11 fissions in the first period to 5 in the second period of the first experiment. The non-conjugant mode dropped from 10 to 6 during the same periods. These two periods are typical of the sort of changes that were observed in many experiments. At the time of such shifts of mode in one group, similar changes were observed in the reproductive rates of all the races being cultivated, so that it appeared practically certain that the change was due to some change in the conditions of cultivation. 266 T. M. SOXXEP.OKX AXD R. S. LYNCH i ^: C2 § -=: .8 £ • H "& ^ "2 X O h 'C ^: o __. a. ^ >. •s x = ~ 'C ^: 3 ^"° c^ tn <-> 5 R -2 " •*-~ *— . 51 o R J = S = .1.2 -^ r OJ 1-sJ Z. -<•. ,0 - tN . T_— t~^» to ~" CN ^ - i/". 1 ~- 2^ "-. OO 2 O\ CN t^-. 10 "~, O "i OO O PC » OO PC CN l/-. IO ^-i f*i 3C --.0 en ^^ •— i r) O •/. 00 T}. -H O CN CN PO f^ "--•<* - - -* 00 ^ ^ | o OO O> -f CO CN PC T-( CN 1/5 ^- O PC ro O "0 ^^ CN o r*3 t 2 •^ I-~ to "> „ OV- V. •r-t ^ M ~ ^~ ~ cc - CN- ~ - T— t 0 «, re CN ti Q e « ^* ^* * t ci J? c: !^ ^ '-? b; ^ o "o o "° o'o 1*0 S^o g^ c U x he m 4J C 4- C ^* t/) irv c/) "-/ j • QJD , > feoS bbS c 1 ^ rt c ^ c rt c ^ c ) « 3 g, 3 rt W 3 ^ S 3 So 3° 3 '^ 3 '^ 3 '? 3 '^ .-, c .2, c 1? la " c V = 1? 1? V c V c o V o V V = V c X - X O X 0 x 0 u; x u: x X O X O ^ X U 2 0 o o o S- ° oo" S CN o CN O 4 OO O* c IO •— • ^J */ i •> 1 5 : j S T" 4J 0 0 ^ 10 ro t— ^^ ro Tt^ r^* o ^— < o ^o *O '^ ^T* *-"* OO so o t— so 00 Os CN *— « r*» ^ oo r^j Tf oo t— ro TJ< i— I OO t— c.2 O O »-i CN C^) (*O ^"I ^1 O -H ^H O O ^ ^H ^H d d 'S.{3 -H-H -H-H -H-H 41-H -H-H -H41 -H-H •H-H -H-H u ^ -H oo CN so ^^ C^l C*4 ^H Tf t— CN SO OO 00 -i- i— t— ON o •* OJ ro Os 'O \O f^ t^* *™^ ^t< i— i— i Os so so OO sO sO 00 U "•f so CN *O •^ ro ^O GO *— i *— i -*• i - — t- <— ' O t- CN If} •-i ro CN co r^ to — i c^l ^H ro T— i --i IN CN 1— t— OO sO i>» *O r^ ^-H co I-H r^i oo CSI i-l Os 'O so rs O 0 0 I-H ^-4 C"1) T— t l-H O >— i ^— i O i— I CS 1— 1 T— 1 i— i i— i S-B O 0 O O o o o o O O O O O O 0 0 0 0 •si -H-H -H-H -H-H 4i-H JJ _M M _M Tl Tl Tl Tl -H-H -H-H •H-H 0) > •" 0 *— < *O OO *O Os t~^- Os ~^ ^H t— (M f> O "t1 Os 10 r^. oo in -3 10 o 10 ^* OO O t— OO ~ O -H t- r<0 OO GO O ro \o T-H O 1— 1 *— 1 fS CN '-HO tS T- 1 f> O CS I-H rsi «-H -* -H E » ^— I ON *— ' CN JH.O 0- SO Os t— i-H tN Ol jq^ c.o O O O O O O O O o o o o o d d d d d ~- CA •Sic -H-H -H-H -H-H -H-H -H-H -H-H •H-H -H-H -H-H *J tX be 3 >-• bO 00 3 a o 1 0 | 0 a v a v •'-' '?-" -"-• 'r-1 So 5 o CO r \ Cu f ) be V oo V 4-> '£T -U '^ So So CD r ) CD f) bO T bO V 1? 00 V = '§ be V O 3 C 3 C 3 C 3 C 3 C 3 C 3 C 3 C 3 C '§ ° '§ = •;? o •=• o C c ^ c 0 0 C '=c 'eg 0 £ 0 ' •*— t Q § c If '§ = o a u a o a o a o a o a CJ Q >< *^" ^ *^ ^ t— ^ t— i t — i t^ x '^ y *~~ i r— U (N W «N •d IH cu G o d o. o' o '"I o o Os Os CO •o _o CJ C. d CO o' Os O Os >» ^H K^ »— i ^— i ^ Os * i-H ri ni •a *— » .? co" £ CN~ 1— ' o t-C a « *-H C^Q CNJ C ^f ^l CN i i I 1 CJ 1 s 1 0 i i ** Os OO ro H 1 ** OO 10 -H CN CN »-H CS es J-j J-j • CJ CJ | > j CJ U (J 0 0 0 U 0 0 (U as fN CN w E T. M. SOXXEBOKX AXD R. S. LYXCH It is a remarkable fact that in spite of the great shifts of mode from period to period, the modes of the ex-conjugants of 247a and the modes of the non-con jugants of this clone were always very similar within any one period. Such a persistent similarity in modes appears in Table II. During the first experiment, the modes of ex-conjugants (using ex- con jugant LM'""]1 O and nun con jugants are 11 and 10 for the first pe- riod, 5 and () f<>r the second period ; during the second experiment (using e-x-conjumant j^roup D) the modes are 10 and 9-10 for the first period, 4 and 4-5 for the second period; during the third experiment (using ex-ci injurant i^roup /: ) the modes are 11 and 10 for the first period, 11 TABLE IV Ratio of ex-conjugant groups (C, D, and E) to non-conjugant groups of their parent clone (247a) in mean total number of fissions, standard deviations, and coefficients of 'inn. These ratios are calculated from the data in Table III. Ten-day periods Exper- iment Ex-conjimunt gn nips com- pared with groups of nim-conju- gants of 247a Ratio of ex-conjugant group to synchronous group of non-conjugants of 247a in mean total number of fissions in standard ii ion in coeffi- i-ii'iit MI variation Oct. 5-14 . . 1 2 3 C D E 1.0 1.0 1.1 1.3 2.8 4.1 1.3 2.7 3.9 Oct. 24-Nov. 2 Dec. 18-27 . and 9 for the second period. This close agreement of modal values of non-conjugant and ex-conjugant groups within the same period, in spite of great fluctuation in mode from period to period, must be due to some innate' similarity between these1 two kinds of groups, both derived originally from the same ancestors. Tables 1 1 1 and IV demonstrate statistically the relations which Table II showed on inspection. In Table- III the1 mean fission rate with its standard deviation and its coefficient of variation are given for both cx- conjuganl and non-conjugant Croups for all the periods of the experi- ments. In Table I V. the rat ios of each of these statistical values for the i-oiijugant group to the corresponding values for the non-conjugant »up are mi veil. Again appears in Table III the ^reat fluctuation in means from period to period. The ex-conjugant mean drops from 10. 'o to I. Si in the two periods of the first experiment and from 8.34 to l.X'» iii )])<• two periods of the second experiment. Also the means of the two Croups in every period ai;ain appear remarkably similar: 10.65 and 10.18; 4.81 and 5.68; 8.34 and 8.86; 4.89 and 4.r>2*; 9.83 and . ''.13 and ''.I''. In three' of the- six periods the cx-conjugant mean RACIAL DIFFERENCES IN CONJUGATION 269 exceeds the non-conjugant mean l>v an average of 0.51 fission; in the other three periods the ex-conjugant mean is less than the non-conjugant mean by an average of 0.48 fission. Thus the algebraic mean difference between the five-day means for the two groups is practically negligible. This similarity between the means of ex-conjugants and non-conjugants of the same race is expressed another way in Table IV where the ratio of the former to the latter is given. The ratios for the three experi- ments (total of ten days each) are 1.0, 1.0 and 1.1. The uniformity of this ratio of ex-conjugant mean to non-conjugant mean in all three ex- periments is the more remarkable because of the large differences in the absolute values of the means for the three experiments (Table III). In the first experiment the values are 15.70 and 15.78, in the second experiment 13.87 and 13.74; in the third experiment, 19.26 and 18.32. It is therefore quite clear that in the clone 247a the mean value of the fission rate is not changed by conjugation ; nevertheless it should be remembered that some of the individual values, as well as the distribu- tion of values, among the ex-conjugant groups are very different from those of the non-conjugant groups. So far as mean fission rate is con- cerned, there is in these experiments neither rejuvenescence nor depres- sion as a result of conjugation. For the purposes of the present paper, this is the most important result of these three experiments. But it should be remembered that so far as the separate lines are concerned, many ex-conjugants are very diverse from the non-conjugants. In the first experiment one ex-con- jugant line reproduced more slowly than any non-conjugant line (13 others reproduced more slowly than all non-conjugants but one), 5 more rapidly, and 84 at the same rates as the non-conjugants; in the second experiment, 10 ex-conjugant lines reproduced more slowly than any non-conjugant line, 11 more rapidly, and 31 at the same rates as the non-conjugants ; in the third experiment, 22 ex-conjugant lines re- produced more slowly than any non-conjugant lines, 95 more rapidly, and 50 at the same rates as the non-conjugants. Thus, in all three ex- periments, conjugation produced some rejuvenated lines, some depressed lines, and many lines that were unaltered in fission rate. It is only when all types of lines are averaged together that no effect of conjuga- tion on fission rate appears. In this average absence of effect, all three experiments are in complete agreement. The effect of conjugation on variability of fission rate has already been indicated. Table III shows that both the absolute variability, as measured by the standard deviation, and relative variability, as measured by the coefficient of variation, are greater for the ex-conjugants than for the non-conjugants, in every period of all three experiments. Conjuga- 270 T. M. SOXXKBORX AXD R. S. LYXCH tion has unmistakably increased the variation, both relative and ab- solute, of the fission rate. This result is in agreement with the results of Jennings (1913). KatiVl (1930), and Jennings ct al. (1932). In Table IV an- given the ratios of the ex-conjugant standard deviations and coefficients of variation to the corresponding values for the non- con jugants. Because of the close agreement of the means of corre- sponding groups of ex-conjugants and non-con jugants, little difference appear.- between the ratios of the standard deviations and the ratios of the coefficients of variation in any one experiment. The two ratios differ very much, however, from experiment to experiment. For the first experiment both ratios are 1.3; for the second experiment, the ratios are 2.8 and 2.7, respectively; for the third experiment, 4.1 and 3.9, respectively. Thus conjugation has increased both relative and absolute variation in fission rate in all three experiments. In each case it has given rise to lines with higher fission rate, and lines with lower fission rate, as well as to lines with unchanged fission rate. As to the relative mortality of conjugants and non-con jugants, no constant result was found in these three experiments. In the first ex- periment the mortality was 22.4 per cent for the ex-conjugant lines, 20.8 per cent for the non-conjugant lines; in the second experiment, 44.1 per cent for the ex-conjugants, 31.2 per cent for non-con j ugants ; in the third experiment, 10.2 per cent for ex-conjugants, 20.8 per cent for non-conjugants. Thus the mortality is nearly alike in the two groups in the first experiment, it is greater among the ex-conjugants in the second experiment, and is greater among the non-conjugants in the third experiment. The three experiments on the effects of conjugation in the race 247a have thus shown, as far as they go, that conjugation does not always pro- duce the same effect on mortality in this race, but that it does always produce the same effect on mean fission rate, on absolute variability, and nil relative variability. The mean fission rate is regularly unchanged by conjugation, so that conjugation neither rejuvenates nor depresses the population as a whole. However, both absolute variability and relative variability of fission rate are regularly increased by conjugation. This results, in spite of absence of the effect on the mean, in the regular pro- duction of both depre-^ed and rejuvenated ex-con jugants. IV. SIM n/iANKors CoNj i '..VMONS WITHIN Six EX-CONJUGANT CLONKS 1 )i KIYKD FROM TIII-: CLONK 247 272 T. M. SOXXKI'.ORX AXD R. S. LYNCH 5 £ -e B ca < 8 - -z-' : a~T3 ^ o 3'S . -- ^i— •f. - S ui R-g i - ~ 3 '-_- = ~~. a (N - - r ~ •r- t^ o •<* •f. •— • IO CN CM -s^ ~ t*. *"• o\ r^. CM "•- rr tN r-) r^ CN r«3 CN OO -H CN C — — ir. -1 - 1 >n '•'-. — vC O CN — •^ rvj «— i c*5 CO «— i iv. tr -'-. rs CN — CN !•- ^- IO tN c*5 0 >- — CN 10 ro i^O CN es i — i O -^ 00 ro -r ~ — '- "— 0) ro ro re r^i -f ' f — CN — 01 "- ^— i CN ' — ' — ro T-I CN CN CN t«! oo 00 ro >-( rq «-i CN Vi ^i LT, ^-i CN •^ — C -H Yi ~i i - - ^ -i re )- L A C ~ , CN - - rO C^l ~ = 3 O> vH - •^-i *-H ro — ro •— >— i— « rl — j O H X - rc - O .— ^ t^ - ^— « ^~ rC »-« ^~ vO — ^ ^ ~ O CN CN — U-, »— • — CN CN CN CN ¥—1 •* »H -H CN ro [•) - - 04 ^— CN r}< — -* ro CN O CN — »— i U] en en •£ en ti <-.•-*-. 5 C C3 £ - rt « -: -J- b£ 3 be 3 M <« t/i ti en ti S rt ^ rt 5 £f « " bij 3 (3fl 3 (/) U) <" = <"£ ^_, — ±j W« : rt c rt « £f a y be 3 bo 3 cn i ™ C « C c rt -- r3 rt 5f « Sf bfl 3 M 3 «) (« cn 'P tn ^ *-• 5 <-> 5 r- rt C ™ cd " = --f- bo 3 bo.3 tn rt ^ 0)5 3 5/0 3 < .E/c'.E/E1 .X = X = = '?.= '^ .=,'=" 3 '? 5. c .S. e 3 '? 3 '? ' 3 ? g ? 0 z - o V V c V c / - •/, c _ x _: / l?§8 V = V c X - X - u z w z §?§8 V c V c X 0 x C U Z W Z c o c o o V o V V = V = •s. c / - W 2 W 2 §v§8 V c V c X O x O ^ X. W £ 'e?o'eO o V § V V 3 V C X O X O W X. _^ X. - ' - O i/-. - 1 ^1 ^ ~ — r-j C C rt rs 1 — i — , O "•- '-I CN A JL -^ CN = C n rt o >o ^1 CN 1 I O ~^ — rvj c c n ca 1 — , •— > o >o CN CN ^ JL — CN C C c; rt •— i >— i o >o CN CN >i J. — r«.| C C C3 rt O 'O CN CN I — CN C C cs rt 1 — . • — 1 ( . = 2 _ a c -r K| a 5 -a O 3 -a O OO si a ^- Kl -s IO OO f^ RACIAL DIFFERENCES IN CONJUGATION 273 w P3 H M • *« ^ •*^ K a 5X0 § a "o en -4—1 C a> >*i O) . o -c o a 2 *~^ •— 4) II K ti a^ « rX" '< " 5X0 « 0 fSJ O OO ON ON Tf OO O t— oo CN ON CN O SO ON '-H 10 t^- OC oo i^- >O *— i IM O t^. 10 t— o CN so v~t ON ^^ *~t "O CN VI a 'o fe^ &^ £N &*• &x&x- SO '•f ON CN OO O Tj< OO ON OC ON SO CN ON ON CN ON t~*» sO O OO ro O 00 o U o o x W O O O O Tf -H t~~» l~-» *-H CN OO f} OO ON OO sO CO Wi O .-H O ^H O O JJ JJ JJ JJ JJ JJ Tl Tl Tl Tl Tl Tl ON ^ ro ^t1 so T— < sO "^ c-o rf ro OO OO *O t~- *rf ^O so ^^ OO iO CN OO CO >— i T-H t— i CN SO «— i OO so i^} ON CN OO *— i *— i CN CN c o U 0 O O O O 0 II _M JJ M II JJ Tl Tl Tl Tl Tl n O O O O O O -rf CN fO i~C 00 t^ o •d _ o ^ ^ ^ 0 «-H 1-H 1-H 1-4 CO »H •a •a r^» ^f so PO o fo O »— i >— ' *— < c*3 CN 00 Tj< CN t~- -H CN a o U 0 O O O O 0 41 -H41 -H -H 41 -f SO 00 O OC O ^t SO OO -rf ON O C' 0 O O O O O CN O 00 OO O so *O ^— ' ON SO ON i— c CN CN CN O IO t^ SO o o o o o o JJ JJ JJ JJ JJ JJ Tl Tl Tl Tl Tl Tl ro O ON O ro •— i ^— i u"j CN O OO t~- U Tf so sO -<*i ro iO ro CO CN ro OO ro .0 g 5 2 CN CN ON CN ro CN OO t-- ^f ON t-- T— < ^^ ^— 1 CN W) »-H 4J 1 o O O O O O O 4i-H41-H4i-H \o c> oo cc ~o ^— « CN O T— i co rt< ro O O 0 0 0 O JJ JJ JJ JJ JJ JJ Tl Tl Tl Tl Tl Tl O SO O SO ON <-O r-~ so t-« CN ro t-~ ^*O ^O ^^D v^2 1"^^ *-O ^2 ^2 ^2 ^2 "^ ^ C : L 'O '•^ *O cO '~™) *-O ^ ^ ^ oo oo oo tq ^ K) bq te] fe) C3 C!3 "i^!l "-^l \t3 »^^ O ~ sO O ^ 10 •^ "* Tj< 00 00 00 LiJ fc] I-ij bJ IjJ IjJ : £ (1 1— 1 ON o" CN 0 c V— 1 ON 2 CM C C8 274 T. M. SOXXKI'.nRX AXD R. S. LYXCH E4\a and 7T81a arc 9.23 ± 0.45 fur the first period and 8.27 ± 0.62 for the second period. The- differences between ESOb and £81 a are 8.92 ± 0.45 for tin- first period and 7.87 ± 0.64 for the second period. The difference^ between / ll are 0.31 ± 0.27 for the first period and 0.40 ± 0.30 for the second period. Thus the differences between £41 and 718 la are clearly significant . the differences ranging from twelve to twenty times their probable errors. On the other hand, the differences between 7:41a and /:Si)/> art' only 1.1 to 1.3 greater than their probable errors and therefore cannot In- considered as certainly significant. The uniformity of these differences between 714 la and £80/>. however, makes them somewhat nmre significant than their probable errors would indicate. It is clear, in any ca>e. that there is a well-defined difference in fission rate between 7;'Sb/ and all the other groups. The differences between £81a and the TABLE VII Differences in standard deviation uniang the three grades of ex-conjugant groups. Grade 1 := group £40a; grade 2 == groups £41a, £466, and £806; grade 3 == group ESla. Grades Compared Groups Compared Period 1 Period 2 Difference Difference Difference Difference P.E. P.E. 3-2.... /•:sia-£466 1.10 ± 0.34 3.2 1.58 ± 0.43 3.7 3-2.... E81a-£41a 1.32 ± 0.33 4.0 1.16 ±0.43 2.7 3-2. . . . /•.xia-£806 1.58 ± 0.33 4.S 0.70 ± O.I! 1.6 2-1.... E41a-E40a 1.22 ± 0.16 7.6 0.92 ± 0.16 5.8 2-1.... E46h-KWu 1.44 ±0.18 8.0 0.50 ± 0.14 3.6 2-1.... £806-£40,; 0.96 ±0.14 6.9 [.38 ± 0.19 7.3 3-1.... £81a-E40a 2.54 ± 0.31 8.2 2.08 ± 0.42 5.0 nearest it in each period are 7.85 ± 0.53 (E85b—E&la) in the first ]>eriod and 7.87 rb 0.64 ( E80b-E8 1 a ) in the second period. In absolute variability, 'fable VI shows that the six ex-conjugant Croups fall into three consistently distinct grades: one with standard iation near 1.5 fissions, one between 2.0 and 3.0 fissions, and one bet \\een 3.7 and 4.0 fissions, for the two live-day periods. The group /:!<),/ bad standard deviations of 1.44 and l.f>0 fissions; these were lower than those of any other group. The groups E41h, and /:S()/' all fell in an intermediate -rade with standard deviations varying from 2.< ) to 3.0 fissions and without consistent differences among the three groups in this i;rade. The -rotip 7:8 la, with standard deviations RACIAL DIFFERENCES IN CONJUGATION 275 of 3.98 and 3.68 fissions had a variation distinctly higher than any other group. Not only are the differences between these three grades of standard deviations consistent, but they are also statistically significant, as ap- pears in Table VII. All the differences except those in the second period between E4\a and £8 la and between £80^ and £81a are more than three times their probable error. Therefore the differences be- tween the three groups £40(7, E46b, and £8 la are clearly significant and establish the existence of three grades of standard deviations among the six races. The data on relative variability of fission rate, as measured by the TABLE VIII Differences in coefficients of variation among the three grades of ex-conjugant groups. Grade 1 == group £40a; grade 2 == groups £41a, £466, and £806; grade 3 = group ESla. Grades Compared Groups Compared Period 1 Period 2 Difference Difference Difference Difference P.E. P.E. 3-2.... £81a-£806 38.76 ± 5.05 7.7 45.92 ±10.77 4.3 3-2 £81a-£466 35.57 ± 5.08 7.0 53.03 ±10.72 5.0 3-2 £Sla-£41a 37.41 ± 5.05 7.4 49.92 ±10. 74 4.7 2-1.... £80Z>-£40a 5.75 ± 0.94 6.1 10.75 ± 1.46 7.4 2-1.... £466-£40a 8.94 ± 1.01 8.9 3.64 ± 1.08 3.4 2-1.... £41o-£40a 7.10 ± 0.96 7.4 6.75 ± 1.21 5.6 3-1 £81a-£40a 44.51 ± 5.00 8.9 56.67 ±10. 71 5.3 coefficient of variation are also given in Table VI. Differences in rela- tive variability among the ex-conjugant groups closely parallel the dif- ferences in absolute variability. The group £40a is again the least vari- able with coefficients of 8.87 per cent and 11.69 per cent; £41a, E46b, and E80b again fall into the intermediate grade, with coefficients from 14.62 per cent to 22.44 per cent ; and £81a is again the most variable, with coefficients of 53.38 per cent and 68.36 per cent. As appears in Table VIII, the differences between these three grades are clearly sta- tistically significant in every case and in both periods. The data on mortality, as measured by the ratio of the number of lines that died out during a given period to the number of lines that were alive at the beginning of the period can be expressed as percent- ages. These percentages, in the cases of the groups of ex-con jugants. 276 T. M. SOXXKi'.OKX AX!) R. S. LYXCH are the percentages of line-, or ex-conjugant clones (since only one line of each ex-conjugant clone was in progress), that died out during a given period. The percentages of mortality among the six groups of ex-conjuganl clones, 'luring the \vliole experiment of January 16-25, fell into three distinctly different grades. The gmups 7:400b and £85/' had a moderate mortality (35.7 per cent, 24.2 per cent, and 37.5 per cent, n spectivi i- i ; and tin- group 718 h/ had a very high mortality () in which both the non-con jugants and the ex-con jugants of the one consistently multiply more rapidly than the corresponding o-roups of the other. All of the parent non-conjugant clones except /:'S1,/ \\-ere so similar in mean fis- sion rate that it is surprising that even one pair of clones (E4la and £80/' i duild hi' selected with consistent differences from period to period. These results are similar to the results of Calkins (1919 and 1920), \V<",dniff and >pencer (1924). and Jennings ct <;/. (1932), in that di\er>ities in mean fission rate between different lines of descent TAI'.LH IX Comparison of six clones in the ratios of ex-conjugants to non-conjugants in mean total number of fissions, standard deviation, and coefficient of variation Period Clones £40a E41a £466 £806 £8 la £856 Mean total number of fissions . . Standard deviation Jan. 16-20 Jan. 21-25 Jan. 16-20 Jan. 21-25 Jan. 16-20 Jan. 21-25 1.14 1.04 1.32 1.04 1.16 1.00 1.00 1.01 2.74 1.91 2.76 1.88 0.98 1.1 2 2.29 1.15 2.33 1.03 1.12 1.02 1.51 1.95 1.34 1.91 0.54 0.61 2.27 1.03 U8 1.69 0.98 1.00 4.32 1.14 1 11 1.14 Coefficient of variation within the same species are found to per>i>t through conjugation, re- appearing as similar diversities between the derived sets of ex-con- jugants. In 'falile IX art' s^iveii the ratios of ex-conjugant mean fission rates, .standard dexiatimi-. and o >eflicients of variation to the corresponding value-, t'.tr non-conjugants. The ratios of the mean fission rates fall into three classes: those less than one. those equal to one, and those greater than one. The ratios for the elone /iXb/ are 0.54 for the first period, and O.nl for the second period. The corresponding ratios for the clone /. Ib/ are 1.00 and 1.01 ; and for the clone ESOb they are 1.12 and 1.02. Thus, the mean li»ion rate of the clone /:X1<7 has been markedly decreased by conjugation; tin- mean fission rate of the clone / Hi/ has been unchanged by conjugation; and the mean fission rate of the clone /:SO/' has been \-ery slightly increased by conjugation. With >'t to tin- effect of conjugation on mean fission rate, l:.\\a is a clone similar to 247, /. as described above on pa^e 268. This simultaneous comparison of tin effects (.f conjugation in six clones demonstrates RACIAL DIFFERENCES IN CONJUGATION 279 that conjugation docs not have the same effect on mean fission rate in all clones : in some clones, conjugation raised very slightly the mean fission rate (rejuvenescence) ; in other clones, it lowered the mean fis- sion rate (depression) ; and in other clones it left the mean fission rate unaltered. It appears here again that ihe effect of conjugation on mean fission rate depends on tlie nature of tJic clone which conjugates. The ratios of the standard deviations in the ex-con jugants to those in the non-con jugants (Table IX) show a number of consistent differ- ences, but most of them are slight. In £40a the ratio is lower than in £85/>. FA\a, E46b, ESQb : in ES\a it is lower than in ES5b and E4la; and in E46b it is lower than in E4\a. The greatest consistent differ- ence in ratio is between E40a and E4 la. The ratios for the former are 1.32 and 1.04, for the two different periods; for the latter the corre- sponding ratios are 2.74 and 1.91. Thus conjugation increased ab- solute variability of fission rate about twice as much in the clone E4la as in the clone E40a. TABLE X Comparison of clone E40a -with clone E4la in differences between non-conjugants and ex-conjugants in standard deviation and coefficient of variation of total number of fissions. Jan. 16-20 Jan. 21-25 Difference between non-conjugants and ex-conjugants in Standard Deviations Clone E40a 0.33 ±0.19 0.06 ±0.27 Difference between non-conjugants and ex-conjugants in Coefficients of Variation Clone E41a Clone £40a 1.69 ±0.21 1.22 ± 1.38% 1.20 ±0.26 0.01 ±2.09% Clone £81a 10.19 ± 1.18% 8.64± 1.99% The ratios of the coefficients of variation of the ex-conjugants to those of the non-conjugants (also in Table IX) show differences simi- lar to the differences in the ratios of the standard deviations. E4Qa is consistently lowest, with ratios of 1.16 and 1.00; E4la is among the highest, with ratios of 2.76 and 1.88. Thus the relative variability of £41a has been about twice as much increased by conjugation as the relative variability of £40(7. These results are another instance of the diversity of the effects of conjugation in different clones. In the clone £40(7, conjugation has not significantly increased either absolute or relative variability. On the other hand, in the clone £41a, conjugation has increased both abso- lute and relative variability by 100 per cent to 200 per cent. LIiRAR 2Sn T. M. SOXXRBORX AND R. S. LYXCH a - r - OO - 3 - c — — 10 CN — •= u 2 -- ^-H •o > * rO ^-> *— » *O lO ?C CN <— • N - 1C * \O ^^ *~< ^-H «-H r^ \O r^l «— t - B _o 'to VI - rO ^H '•^ rs ^^ o-] i^» -^ 0 O CN ^H ^^ r^ ^H ^* ** CJ £ 3 tf. - — o -f oo ^j 22 3 S ^ 2 99 ^ -i ^ J i . 1 U- — U. r^ S tN CN u u rt rt 'C i X U 0 H T-> <^| ro -f RACIAL DIFFERENCES IN CONJUGATION 281 13 <» S X w 2 CO «: CO a o ^CN O i_ w> ro CN 1-4 M 3 2 *— i rs ^-H I-H CNT^I ^— i ^ 'c O O 2 ro ^-H r^ ^H CN *-H ^f a o 2 -1 "^ -"^ _ ^ ^ CO T3 O - -H ^ t-> Q 6 0 -.^- - - cs i C* ^ ^,0 - -- _fl * ^-H 10 •rH ^H CS CO _O *co - ^H *-l CN •rt * "8 0 ^^CO aj 1 « CN 1-1 i-( CN 3 C 13 « *-H CN rf) o H *> *-< CN c<0 „ T-H CN -H «— 1 T— H o ^ — 1 T— ( C /) u 3 O' — iO* — ! O'"1 O ^^ O ^— i O* — ' O '— i O ' — i tiJtiJt^Kl KJK) hJKJtiJKl tt} ft] b} bj ft] KJ I a 2 SS 2 SS 77? ^^ i 2S 2S» r- r* O O _Q !lj i— tn *~r! *TT *T! CU TO Cy _j QJ ^ W ul LX! Cxi S S ^ c « . T. M. SOXXKRORX AXD R. S. LYXCH As appears in Table X. tin- diversities between ex-conjugants and non-con jugants in 7:40 and ESSb. On the other hand, the non- conjugants of clone E8\a, like the non-con jugants of clone E85b, had a mortality of 37.5 per cent, but the ex-conjugants of clone ESla had a mortality of 81.3 per cent. Thus, as with mean fission rate and varia- bility, in some clones (c.;j.. /:S1(/), conjugation greatly alters mortality, and in other clones leaves mortality unchanged. The evidence thus far shows clearly that diversities in the effect of conjugation in different clone's are due to constitutional diversities be- tween the clones. It si-ems desirable to discover whether such diverse effects in different clones are constant. Will the same diversities recur in repeated conjugations? This matter was tested in two of the clones that were found to give diverse results of conjugation; it is fully clis- ;ed in the next section. V. REPEATED COM r. \RISONS OF THK KFFF.CTS OF CONJ rc.vnoxs IN THE Two CI.ONF.S /:40(/ AND /:S1(; Conjunctions within the Clones I: I Oil and /•>'/<; In order to determine whether the diverse effects of conjugation in different clones occur regularly, four sets of conjugants were obtained and studied in clones /•.}(),/ and E81o. These eight sets of conjugants were obtained two at a time, one from each of the two clones. In every case, of course, conjugation in the two clones occurred m lines that had not conjugated since the /•.' conjugation. RACIAL DIFFERENCES IN CONJUGATION 283 Results of the four sets of conjugations were studied during the in- tervals: January 16-25, February 12-14, February 19-28, and March 18-April 1. The experiments, as usual, were designed to compare mor- tality and fission rates, and variability of fission rates. As will appear in detail below, mortality among the ex-conjugants from the clone E8la was so high that, although comparison of rates of mortality could be made for each of the four experiments, comparisons of fission rate, variability of fission rate, and the ratios of the data of ex-conjugants to the data of non-con jugants in these respects could not be made for four of the eight periods of the experiments. During the remaining four periods (January 16-20, January 21-25, February 12-14, February 19-23), however, there were 41, 18, 30, and 51 ex-conjugant lines of 0-2 3-5 6-8 9-11 12-14 15-17 18-20 21-23 FIG. 2. Comparison of group of ex-conjugants from clone E4Qa (solid curve) with a group of ex-conjugants from £81o (broken curve), during the five-day pe- riod, January 16-20. The percentage of lines of each group (ordinate) is plotted against the number of fissions (abscissa). ESI a to compare with 87, 84, 84, and 46 ex-conjugant lines of £"40a, so that statistical comparisons of fission rate and variability could be made. The basic data for all eight periods of these four experiments are given in Tallies XI and XIII. Table XI gives the basic data on fission rates, from which all the data on means, standard deviations and co- efficients of variation are derived. Table XIII gives the data on mor- tality. The most striking fact shown in Table XI is the difference between the distribution of values for the ex-conjugants of E8la and the dis- tribution of values for the ex-conjugants of E4Qa. The former are 284 T. M. SOXXEBORN AXD R. S. LYXCH obviously massed considerably to the left of (that is, are lower in value than) the latter. This difference in distribution is brought out graphi- cally by Fi.u- 2, 3, 4. and 5; these arc- graphs on which are plotted the percentage "f ex-conjugants attaining each number of fissions during a single five-day period (three-day period in Fig. 4). Figures 2 and 3 give tin- curves h>r the two periods of the first experiment; Fig. 4 gives the curves fur the >lu>rt second experiment; Fig. 5, the curves for the first period of the third experiment. In all periods of the fourth experi- ment too few ex-con jugants of E8\a survived to warrant plotting curves. 64 56 48 40 32 24 16 8 0 I 0-2 3-5 6-8 9-11 12-14 15-17 18-20 l-i<;. 3. Comparison of group of ex-conjugants from clone /T40a (solid curve) with a group of ex-conjugants from /:S1<; (broken curve), during the five-day pe- riod, January 21 J5. The percentage of lines of each group (ordinate) is plotted against the number of fissions (abscissa). In the first experiment (Figs. 2 and 3), the curves for the two groups of ex-conjugants are totally different. The solid lines, repre- :in- the Croups of ex-conjugants from 7:40t period (Fig. 2) and be- tween 12 and 1 1 fusions in the second period ( Fig. 3). On the othcr hand. the broken line-.. repri'M'iitniL; the groui>s of ex-conjugants from /:'S1 bring out the much greater ran^e and variation, but much ater frequency of lower li»i. 4 and 5), the curves for the two groups have nearly the K \CI \L DIFFERENCES IN CONJUGATION 285 same range, but the curves for E40a are peaked near their upper limits and the curves of ESla near their lower limits. In the second experi- ment (Fig. 4) the peak of the E40a curve is between 6 and 8; the peak of the £8 la curve between 0 and 2. In the first period of the third experiment (Fig. 5), the peak of the E40a curve is at 12-14 fissions; 56 r 0-2 3-5 68 9-11 12-14 FIG. 4. Comparison of group of ex-conj ugants from clone E40a (solid curve) with a group of ex-conjugants from £81a (broken curve), during the three-day pe- riod, February 12-14. The percentage of lines of each group (ordinate) is plotted against the number of fissions (abscissa). 40 32 24 16 8 0 0-2 3-5 6-8 9-11 12-1415-17 FIG. 5. Comparison of group of ex-conjugants from clone £40(7 (solid curve) with a group of ex-conjugants from E8\a (broken curve), during the five-day pe- riod, February 19-23. The percentage of lines of each group (ordinate) is plotted against the number of fissions (abscissa). the peak of the ESla curve at 3-5 fissions. These curves bring out the much more frequent occurrence of low fission rates among the E81a ex- conjugants as compared with the E40a ex-conjugants ; and also the very similar range and absolute variation. The relative variability, neverthe- less, is much greater among the £8 la groups of ex-conjugants on ac- count of their much lower means. Both of these sets of curves (Figs. 4 and 5) agree with the former sets (Figs. 2 and 3) in that the main portions of the curves for the E40a groups lie to the right of the main T. M. SOXXEBOKN AND R. S. LYNCH portions of the corresponding /:~S1 i""r all groups the mean fission rates, their standard deviations, and their coefficients of variation for all periods yielding sta- tistically significant data. In every period the mean fission rate of the ex-conjugants of 7:40 per cent; 92.9 per cent with 47.5 per cent. For the four experiments the average mor- tality among the groups of ex-c< >n jngants from clone ESla was 81.3 RACIAL DIFFERENCES IN CONJUGATION 287 X a H E 3 03 — O rt . o O ON O '* O CO ON a O O O 0 o o o o e^ ON O -H X •H-H-H-H -H-H-H-H -H -H -H -fl VJ •—* rO NO ^? IO OO *— ' t^- NO T— i eg eg •g 1^ CO >O IO t^. IT) NO ID I— >O -H >O rt tfl ro CO CO eg i — i ro O ' — i eg o t— e-i 3 r* O cl ^^0^0^ o NO t^» NO 'O co NO ^H eg o o ^ -H <^ e-4 CO *— ' Tf -H eg •^-i ro t~o O eg co 3 O O O O O O O O ^H e^ •T-H e-i o -H-H-H-H -H -H41 -H -H -H-H 4i IO to *O t^~ ON "-fi NO NO ir> O 10 eg eg .—i eg ir} O >O NO l>- '^^ t^«. i-^ ^*J 2 2 °" Zl ~ ~" ° ^ f^» ^^ t^^ lO OO O^ O^ i^~' eg O\ r— O ^ 5 2 ?3 ON N^ ON ^ 0 O O O O o o o o •rt 00 Kl 41^ -H -H •ti-H-H -ft -H-H-H-H NO ON rO :O OC CO O ^ CO NO CO CO c/) ^ ro CO O ^\ \O ' — i '"^ rO fO ro O d f~- 10 r^i NO c*o ^^ f^*1 fi rO OC t— eg M if) ^O i^^ ^5 3 's O 0 >vO >^O vO >^O £ W O c^ NO eg I-H ^H ^- 1 ro t^- co ••-H eg O o '-H eg O eg oo TJ« Tt1 NO eg r •H-H-H-H ^-H4i-H -H-H-H-H vo O *— ' ^— ' •^ O CN OO t^- ON fO eg CN t^- lO ^f rfn so ^— ' ^H OO NO 'f IO \O **O ^O ^5 *— ( i— i i— i ^H ^H CN ro CO *-( cc O f~i LO ^^ **O CN r-1 '-H CN C*-4 C"1-) ^^ ^J \v t<> an examination of the relations between the values obtaineil f»r the ex-conjugants and the values obtained for the non- conjugants of the two races, the pertinent data will be found in Tables XIII and XIV. the former giving the mortality data, the latter giving the ratios of means, standard deviations, and coefficients of variation. \Yith one exception the ratios of ex-conjugant to non-conjugant mean fission rates in £40« are very close to 1.0. whereas the ratios in /•SI*/ vary from 0.33 to 0.61. The single exception in the case of £40o is in the second experiment. Observations in this experiment were dis- TAHLE XIII Comparison of ex-conjugants and non-conjugants of clones E40a and ESla. centage mortality. Per- Clom- /m i ( Ilone £8 la Exper- iment Period Ex- con j. Non- conj. Ex-conj. minus non-conj. Ex-conj. minus non-conj. Ex- conj. Non- conj. 1 2 |an. 12-25, 1931 Feb. 9-14, 1931 12.5% 14.3% o % 0 % 12.5% 14.3% 43.8% 40.9% 81.3% 4(1.9% 37.: 0 % 3 4 Feb. 15-28, 1931 .... Mar. 18-Apr. 1, 1931 71.9% 47.5% 75.0% 40.0% -3.1% 7.5% 18.2% 22.1% 99.5% 92.9% 81.3% 70.8% continued after three days because of a defect in the culture method which resulted in great reduction in fission rates, and extraordinarily high mortality. For this reason, the data of this experiment are of doubtful value. The ratios in the other experiments were quite con- sistent: 1.14, 1.04, and (J.90 for 7-4()t/ ; 0.54, 0.61, and 0.48 for ES\a. It then- fore appears that, under favorable conditions, conjugation regu- larly makes little or no difference in the mean fission rate of £40a, but ularly reduces to about one-half the mean fission rate of ESla. The- ratios nf standard deviations .show no consistent differences be- tween the two clones. ( )n the other hand, the ratios of coefficients of \ariation are consistently higher in /:X1; ].<>') as compared with 1.00; 10.87 as compared with 5.38; 4.1n with 2.01. It is thus clear that conjugation regularly increases relative variability in the race /'.S1f the decided difference.^ between these two clones, it ap- peared of great interest to attempt cross conjugation between them. Thi> was done and twenty pairs of conjngants were obtained in which cue member «i each pair was from the clone f:4()a. the other member of the pair from the clone /:S1. per cent. Comparing thi^ figure of sterility to that found for the control males we find a differ- ence of 2.3 ± 8.09 per cent. — a difference too small to be significant. Treatment of lar\;e at the ages of 72-84 hours rendered 33.33 per cent of tin- males sterile-, while of males from larva- treated at the ages of 24—36 hours, 50 per cent were found to be incapable of reproduction. The percentage of sterility found in the group of males derived from lar\a- treated at 72-84 hours of age differs from that of the control ••up by 19.05 ± 8.29 per cent. Males obtained from larvae treated at 24 -3d hours of age differ in sterility from controls by 35.72 ± 6.7 per cent. EFFECTS OF X-RAYS ON FERTILITY DROSOPHILA 297 SI § 8 a C*3 ?% a PS H *«>i *T* CM) rCi a Difference in percentage of fertility Average percentage of fertility Mfference percentage : sterility Percentag sterile be ile &£ B't in. - S rs -* oo oc oo »O i/S iri CN ^-1 rO OO •^ CN (M ro C | -^ 10 «0 IO >O CN »-l f*5 CN »— 1 -H -H 41 ^41 -H -H C*3 -* GO rg O ^ ^H vo r^< I/O >O t-~ t^~ OO OO O CO tN T-I t* O <— i 00 "t O -t1 '-' 'O r— re O 10 vO I-— l-» 1C CN O oo O f*3 O <^ Cs O O tN t-~ *-H OO ro OO OO MD ON t~~ c<3 -H-H-H-H-H-H •C O f-; CD '-i CO N OS l/"> ON ^ ^C \O re PC O 1-- PC \O PC O PC ON "-1 •* ^— i PC LO PC "^ *-^ CN LO O O O O O C O O "b "b "b o o " ^i 10 o io i— i i — to ^— i *— I CN ^ — i *— i ^ — i PC 4) -= -3 2'»x W. G. MOORE The sterility found in the various groups of females parallels that observed in males treated at corresponding ages. When females were treated as adults 33.33 per cent of the individuals were sterile. Sub- tracting the percentage of sterile controls from this figure leaves a dif- ference of 19.05 ±9.1 per cent. The group of females from larvae irradiated at the ages of 72-84 hours contained very few sterile flies, — one in eleven, or 9.0 per cent. Here the percentage of sterile individuals was less than in the control group. However, it may be seen that a high degree of .-terility maintained among females from larvae treated at 24-36 hours of age. In this group 41.7 per cent of the individuals were TABLE II A Table of Calculations Comparison of the fertility of treated and control flies Percentage of "ility in controls Percentage of fertility in treated Difference in percentage Difference probable error 86 ± 1.5 Adult d" 44 ±5.0 42 ± 5.2 8.07 86 ± 1.5 72-84 hr. d" 58 ±5.2 28 ± 5.4 5.2 86 ± 1.5 24-36 hr. d" 52 ±2.3 34 ± 2.8 12.1 86 ± 1.5 Adult 9 70 ± 1.0 16 ± 1.8 8.9 86 ± 1.5 72-84 hr. 9 71 ± 3.4 15 ±3.8 3.94 86 ± 1.5 24-36 hr. 9 81.3 ± 2.5 4.7 ± 2.5 1.92 Comparison of fertility in sex of treated flies Percentage of fertility Percentage of fertility Difference in percentage Difference probable error Adult 9 70 ±1.0 72-84 hr. 9 71 ±3.4 24-36 hr. 9 81.3 ± 2.5 Adult cf1 44 ± 5.0 72-84 hr. d" 58 ±5.2 24-36 hr. rf1 52 ± 2.3 26 ± 5.8 13 ± 6.2 29.3 ± 3.4 4.9 2.1 8.6 sterili/.ed. The sterility in this case differs from that found in control females by 26.89 ± 8.3 per cent, the difference being 3.24 times the probable error. /•"<•;•/// //y of Eggs Deposited The table of calculations (Table II) was devised to compare the fertility of treated groups to the fertility displayed by the control series. From the data on control Hies we find the mean fertility of the group to be N'I • 1.5 per cent. When adult males were treated with irradiations, tin- mean fertilitv of eggs subjected to fertilization by them was 44 per cent with a probable error of ± 5 per cent. This group of treated EFFECTS OF X-RAYS ON FERTILITY DROSOPHILA 299 males differs in fertility from the controls by 42 ± 5.2 per cent, the difference being 8.7 times its probable error. Males treated in larval stages exhibit no such marked reduction in ability to fertilize eggs as do males irradiated after metamorphosis. The fertility of eggs col- lected from matings of males X-rayed at 72-84 hours of age amounts to 58 per cent of all eggs secured from the group, the probable error being ±5.2 per cent. The mean fertility of this group is found to differ from that of the controls by 28 ± 5.4 per cent ; a difference 5.2 times the probable error. The fertility of males treated at the larval ages of 24-36 hours is slightly lower than that of males irradiated when 72-84 hours old. Of 923 eggs collected from females crossed to males that were X-rayed at 24-36 hours of age, 52 ± 2.3 per cent produced adults. This figure (52 ± 2.3 per cent) subtracted from the fertility displayed by controls (86 ±1.5 per cent) leaves the significant differ- ence of 34 ± 2.8 per cent. Thus far the fertility of treated males only has been considered. A review of the data obtained from treated females, Table II, shows that eggs of females, irradiated thirty hours after emerging from the pupal stage, exhibit a mean fertility of 70 ± 1.0 per cent. A comparison of the egg fertility of this group to that of the control group reveals that, due to the effects of irradiation, 16 ± 1.8 per cent of the eggs from females treated as adults failed to complete their development. This figure, 8.9 times its probable error, is highly significant. The mean fertility of eggs from females treated as 72-84 hour larvse was 71 ± 3.4 per cent, or 15 ± 3.8 per cent less than that of controls. However, treatment of 24—36 hour larvae reduced the fertility of eggs deposited by females from such larvse by only 4.7 ± 2.5 per cent. Since this figure is only 1.92 times its probable error no significance can be at- tached to it. Permanence of Effects upon Fertility and Sterility The volume of data on the permanence of the effects of irradiation upon fertility is such as to exclude it from this publication and a brief review of the findings is included in the section on discussion. From a study of the data contained in Table III it may be seen that in some cases flies which were fertile at 9 days after treatment became sterile within 25 days after irradiation. Nine such cases appear among 52 treated individuals. Of 14 flies found sterile at 9 days after treat- ment, 3 regained their fertility within 25 days, or 21.4 per cent of the flies were temporarily sterilized by irradiations. On the other hand, 6.66 per cent of the control flies found fertile 9 days after mating were sterile in the 25-day period, while none of the controls that were sterile 300 W. G. MOORE TAP.LE III . 1 Table Comparing the Sterile Flies of Two Periods 9-17 days after treatment ! i.ll Nlllir Eggs Flies Sterile EH - Flies Sterile A cf 6 83 2 100 () 1 A cf 9 7 3 37 o 1 A 9 100 39 56 o 1 A 9 13 0 0 1 0 o 1 A 9 14 0 0 1 7 o 1 A 9 15 0 0 1 21 o 1 C of 10 59 10 () o 1 C cf 11 0 0 1 100 89 C cf 13 45 0 1 6 o 1 C c? 14 0 0 1 100 55 C cf 15 0 0 1 6 0 1 C 9 -' 100 82 24 o 1 Bo"1 4 55 1 9 0 1 Be?1 7 100 88 36 0 1 B c?1 10 100 2 3 0 1 B cf 15 0 0 1 24 0 1 B cf 16 0 0 1 0 0 1 B cf 18 0 0 1 0 0 1 B cf 19 29 0 1 38 0 1 B 9 8 50 42 1 (1 1 B 9 12 6 0 1 27 o 1 B 9 14 36 o 1 o o 1 B 9 17 18 0 1 11 3 Total . ... 788 269 14 606 147 20 25-33 days after treatment Controls I) 24 7') 33 38 0 1 D 31 100 0 1 100 0 1 D 32 100 0 1 100 0 1 1 > 33 100 0 1 100 0 1 Total 379 J3 3 338 0 4 lays after mating produced offspring during the 25-day period of mating. DISCUSSION Little more can in- -aid hy way of discussing the sterility effects of irradiation than to point out the percentage differences between the atcd and control groups. It is unnecessary to apply a mathematical formula to determine that there is no difference in the sterility of males EFFECTS OF X-RAYS ON FERTILITY DROSOPHILA 301 treated as adults and that of the control males. A larger percentage of sterility is found if males derived from larvae irradiated at 72-84 hours of age are tested. In this group the percentage of sterile individuals differs from that of the controls by 19.05 ± 8.29 per cent, a figure not mathematically significant but perhaps suggestive. However, males ob- tained from larvae treated at 24—36 hours of age differ in sterility from controls by 35.72 ± 6.7 per cent. In this case there is no doubt that X-rays produced sterile individuals, since the difference is 5.33 times the probable error. Analysis of the data on females shows the sterility of only one group to differ significantly from the sterility found for the control females. The percentage of sterile flies found among females from larva X-rayed at 24-36 hours of age differs from that of the control females by 26.89 db 8.3 per cent. Therefore, in the case of either males or females taken from larvae treated at 24—36 hours of age, irradiation produces sterile individuals. It is also quite probable that X-rays produced sterility in the other treated groups tested, but the dosage used was too light and the numbers were too small for obtaining definite proof. In normal stocks of Drosophila the proportion of eggs which hatch at room temperature is remarkably constant and the percentage of fertile eggs deposited is extremely high. Packard (1927) reported an average fertility of 97 per cent, basing his report upon the number of larvae appearing in the cultures. The thirty fertile controls used in this ex- periment deposited 2979 eggs, 86 per cent of which completed meta- morphosis. When we consider that Packard used a different strain of flies and based the fertility upon larvae rather than adults that emerged, this figure compares very favorably with the fertility which he observed. Table II compares the fertility of the treated groups to the fertility displayed by the control series. When the mean fertilities of the differ- ent groups of treated males are compared to the fertility of eggs ob- tained from controls, we find the differences ranging from 42 ± 5.2 per cent, if males treated as adults are compared, to 28 ± 5.4 per cent, if male larvae irradiated at 72-84 hours of age are used. It appears at once that the reductions of fertility effected by irradiations are signifi- cant. Clearly, the action of X-rays upon the germ cells of the male rendered a large proportion of those cells incapable of entering into the fertilization process, or produced effects within the sperm cells which inhibited the development of the zygotes of which such cells have become a part. From the experimental data it is obvious that the fertility of males as defined in this paper is reduced by the action of X-rays irrespective of the developmental stage irradiated. The highest degree of fertility 302 \V. G. MOORE displayed by any of the three groups of males tested represents a reduc- tion in fertility of 28 ± 5.4 per cent, while the greatest reduction of 42 ± 5.2 per cent is evidence of the extreme effects of X-rays upon the fertility of males treated during adult life. These statements are, how- ever, to be regarded, not as facts which must recur since the difference between any two groups of treated males (the greatest difference being 14 ± 7.2 per cent, or 1.94 times the probable error) is so small as to be insignificant. In case these suggestions point in the right direction, as the data indicate, it is well to note that only the fertility of eggs is con- siden-d, sterile flies being eliminated. This might be called the "net" fertility. 1 f we consider the sterile flies produced in the various groups plus the egg fertility, opposite results are obtained. That is to say, the "gross" fertility of adult males is affected less than that of males treated in the larval stages. Going briefly into the findings of the fertility of irradiated females, we see that the fertilities of only those females from larvae treated at 7J-84 hours of age and the females irradiated as adults differ signifi- cantly from that of the control group. Treatment of 24 to 36-hour larv;e reduced the fertility of eggs deposited by females from such larvae by only 4.7 ± 2.5 per cent. Since this figure is only 1.92 times its prob- able error, we cannot conclude that irradiation reduces the fertility of eggs if the females are treated at the larval age of 24-36 hours. This result may hi- due to the severe action of X-rays, which produces sterile individuals in a large number of the cases in which the gonads are af- fected, or to an interruption of developmental processes which inhibits the appearance of the individual whose gonads arc affected by the ab- sorption of radiant energy. Be that as it may, the data show clearly that the fertility of females treated in the later stages of development is reduced by the action of irradiations. It is interesting to note that there' is no significant difference in the egg fertility observed if the different groups of treated males are com- pared, or it' the three groups of irradiated females are compared with one another. On the other hand, if we compare the fertility of males to the fertility of female- treated in the same stage of development, we find in each case, with one exception, that the fertility of eggs is more affected by treating males than by treating females. The fertility of treated adult male-, is 26 ± 5.8 per cent lower than that of treated adult females. This figure i- 1.5 times the probable error and without doubt i- significant of an inherent difference1 in the germ cells of the sexes which accounts for the greater susceptibility of the sperm cells to ir- radiations. The males and females irradiated at the larval age of 72-84 EFFECTS OF X-RAYS ON FERTILITY DROSOPHILA 303 hours show a difference in fertility of 13 ± 6.2 per cent, favoring the females, or a figure 2.1 times its probable error, which is slightly in- dicative of a greater reduction in the fertility of males than in the fertility of females. The fertility of the males derived from larvae treated earlier in life (24-36 hours) differs greatly from that of the females. In this case the fertility of the males is 29 ± 3.4 per cent lower than the fertility of females. Summarizing the effects of X-rays upon the fertility of flies treated at different stages of development we may say, if sterile flies are elim- inated, that (1) except by treatment of female larvae 24—36 hours of age, the fertility of either sex may be reduced by X-ray treatment ir- respective of the age at which the flies are irradiated, (2) the effects of irradiation upon the fertility of female Drosophila are at a minimum when 24 to 36-hour larvae are treated, (3) there is no significant differ- ence in the fertility of a sex, regardless of the age at which the flies are treated, but (4) it is evident that if treatment is administered to op- posite sexes of Drosophila in the same stages of development, X-rays of the dosage used affect the fertility of males to a greater degree than that of females, irrespective of the age at which the flies are subjected to treatment. It is curious to note that a dosage of X-rays considerably higher than that here used had been found by Muller to have the opposite effect, — reducing the fertility more in adult females than in adult males. Two questions arise : Are the effects which irradiations produce upon fertility of a permanent nature ? Is sterility, once induced, to remain a characteristic of the individual throughout life, or is such sterility of a temporary nature? To determine the answers to these important and interesting questions two periods of egg collection were observed. These were fully described above as the 9 to 17-day period and the 25 to 33-day period and will be referred to as such in the following discussion. Comparisons of average fertilities found for the groups during the 9 to 17-day period with those found for the 25 to 33-day period appear at first glance to indicate that eggs deposited in the second period ex- hibited a higher degree of fertility than eggs deposited in the first period. Considering the number of flies included in these calculations it is hardly advisable to depend upon a comparison of the fertilities exhibited by the groups during the two periods. The numbers are so small and the variations in the fertility of the individual flies so great that the differ- ence in fertility exhibited by one or two flies may influence the average percentage of the group in such manner as to make it slightly higher for the 25 to 33-day period, whereas had these few flies varied as the 304 \v. G. MOORE remainder of the group the average would have been lower for the 25 to 33-day period than for the 9 to 17-day test. But if the fertilities exhibited by tin- individual flics at nine days after irradiation are com- pared to those exhibited twenty-five days after treatment, it becomes evident that some of the flies were more fertile in the second than in the first period, while others were less fertile in the second period than in the first. In either sex, irrespective of the stage in development at which the flies were treated, the fertility displayed twenty-five days after irradiation was often different from that exhibited nine days after the administration of X-rays. Although it is certain that the fertility of untreated flies decreases with age, we cannot say whether the fertility of irradiated individuals will increase or decrease with the passage of time. \Ye can, however, say with some degree of certainty that the percentage of fertile eggs deposited, or fertilized by treated flies, varies greatly at different dates after irradiation. After X-ray treatment the rate of fertilization is highly variable irrespective of the sex which has been subjected to irradiations. In other words, a large number of eggs deposited by treated flies during one period may be viable, while of those collected at some other date relatively few may be capable of development. From a close study of the data included in Table III we find a few instances in which flies were fertile at nine days after treatment but became sterile within twenty-five days after irradiation. Of the fertile treated flies 17.3 per cent became sterile between the seventeenth day and the twenty- fourth day following X-ray treatment. Of fourteen flies sterile at nine days after treatment, three regained their fertility within twenty-five days, 21.4 per cent of the sterile flies were only tem- porarily so. There is, perhaps, no significant difference in the per- centage of flies regaining fertility and the percentage becoming sterile, but significance may be attached to the fact that some flies sterilized by X-rays were found to become fertile with the passage of time following irradiation while none of the four Hies found sterile among the controls ever became fertile. ( >n the other hand, ().(>(> per cent of the control flies found fertile nine da\s after mating were sterile in the 25-day period, indicating a natural tendency for normal flies to become sterile with age. Krom the second period of egg collection (25-33 days after treat- ment of flies, or after mating controls) we learn that (1) the percentage of fertile eg^s deposited at nine davs and twenty-five days after irradia- tion is not constant, (2) flies may be rendered temporarily sterile by X-ray treatment, but sterile untreated flies fail to become fertile with time and ( .} ) untreated flies decrease in fertility with age. EFFECTS OF X-RAYS ON FERTILITY DROSOPHILA 305 It seems reasonable to conclude that the effects of irradiation upon the fertility of Drosophila melanogaster are not permanent, and that fer- tilization may be effected rather irregularly by treated flies. Certainly, since flies are temporarily sterilized by X-rays, the action of irradiations in producing sterile individuals is not always permanent. The author wishes to express thanks to Professor J. T. Patterson and to Professor H. J. Muller for their many helpful suggestions during the course of the experiments. CONCLUSIONS 1. With the moderate dosage used, the degree of fertility exhibited by Drosophila eggs is more affected by the treatment of males than by the irradiation of females of the same age, regardless of the stage in development at which irradiations are administered. Females from cul- tures of larvae treated at 24-36 hours of age display, when fertile at all, a higher degree of fertility than flies of either sex, irrespective of the age at which they are irradiated. 2. Irradiation of larvae 72-84 hours of age sterilizes a smaller pro- portion of the flies of either sex than treatment of any other larval stage, and less sterility is found among females when larvae are treated at this age than in any other group of flies tested. 3. When larvae only are irradiated more males are sterilized than females. 4. The effects of irradiation upon fertility are not permanent. Ir- radiated flies are variable in the degree of fertility which they exhibit at different dates after treatment. Fertility m,ay increase with the passage of time following treatment, but the fertility of control flies decreases with age. 5. The action of X-rays in producing sterile flies may often have only a temporary effect. LITERATURE CITED HANSON, F. B., 1928. The Effects of X-rays on Productivity and the Sex Ratio in Drosophila melanogaster. Am. Nat., 62: 352. HANSON, F. B., AND FRANCIS R. FERRIS, 1929. A Quantitative Study of Fecundity in Drosophila melanogaster. Jour. E.vper. Zoo!., 54: 485. MAVOR, J. W., 1927. A Comparison of the Susceptibility to X-rays of Drosophila melanogaster at Various Stages of Its Life-Cycle. Jour. E.vper. Zool., 47 : 63. MULLER, H. J., 1928. The Problem of Genie Modification. Zcitschr. f. ind. Abst. u. Vererb. Sup., 1: 234. PACKARD, CHARLES, 1927. The Quantitative Biological Effects of X-rays of Dif- ferent Wave-Lengths. Jour. Cancer Res., 11: 1. PACKARD, CHARLES, 1928. A Comparison of the Quantitative Biological Effects of Gamma and X-rays. Jour. Cancer Res., 12: 60. THE EFFI-VT ' IF TKMI'ERATURE ON THE LEG POSTURE AXD SI'KED OF CREEPING IN THE ANT LASIUS T. CUXLIFFE BARNES AXD HENRY I. KOHX i i BORN ZOOLOGICAL LABORATORY, YALE UNIVERSITY INTRODUCTION The present paper is an attempt to investigate the influence of tem- perature upon the locomotion of the ants, Lasius niger and L. umbratiis. In particular it deals with the correlation between temperature and leg- spread, and temperature and speed. Shapley has determined the correlation between temperature and speed of progression for several species of ants. His first paper (Shapley, 1920) reported that, for 1/wuictopuin apicitlatitni, "... an empirical curve is obtained that for any temperature throughout a range of 30° centigrade gives the speed with an average probable error of 5 per cent for one observation. Conversely, from a single observation of the ant-speed, the temperature can be predicted within 1 degree centi- grade." In his second paper (1924), Shapley applied the Yan't Hoff- . \rrhenius equation to his data. He stated: "A constant value of /* at about 11,000, for temperatures from 20° C. to 35° C., appears to be justified. But further observation is necessary to test the indicated variability of p. for temperatures below 20° C." Assuming a break at 20° C., he approximates // as 20,000 for the lower temperatures. Hoagland (1931) has studied the influence of temperature on learn- ing in ants, but this does not concern us here. \Yheeler's book (1913) may be consulted for an account of the ecological aspect of the influence of temperatures on ants. Fielde (1905), in particular, has observed the effect of temperature on the development of ants. LEG-SPREAD AXD TI.MIM.KATURE The relationship between leg-spread and temperature was deter- mined as follows: Ants were permitted to crawl over smoked kymograph paper, lying in a covered Petri dish. The dish was placed in cither a dark, constant-temperature chamber, or in a glass vessel immersed in tlie wafer of a thermostat. In the latter case, light was eliminated by the addition of India ink to the water. 306 TEMPERATURE AND LOCOMOTION IN ANT 307 The records were then shellacked, after which microphotographs were made, with the helpful advice of Professor A. Petrunkevitch. The trails made by the metathoracic legs are easily seen and studied on the microphotographs (Fig. 1). Specimens of L. umbratus, taken from a genetically homogeneous colony, were utilized. In all cases it 13 3 mm. FIG. 1. Microphotographs of the footprints of Lasius umbratus (No. 124) on smoked kymograph paper. These records were taken in the dark at 21° and at 13° C. after one-half hour for thermal adjustment. The greater leg-spread at 21° is evident but is more striking over a wider temperature difference. The long streaks are made by the hind legs. These photographs are not re-touched. was found that as the temperature rose, the legs were held further apart. Since the increase in leg-spread per degree rise in temperature is very small, the phenomenon is best illustrated over a fairly large temperature difference. 308 T. C. BARNES AND H. I. KOHX A typical instance of this phenomenon is demonstrated by Specimen 124 (Figs. 1 and 2). At 12 C. the leg-spread was 2.36 ± .31 mm.; at 21° C, 2.81 ± .20 nun. ; at 25° C, 2.88 ± .23 mm. Sri.KD AND TKMPERATURE Methods Tin- thermostat consisted of a large glass jar rilled with water. The temperature \vas regulated by a mercury toluene contact and by the addition of ice or hot water. Hubbies of compressed air thoroughly stirred the system. The ants crawled in a glass tube 15 cm. long, of 5 mm. internal diameter, on which lines were etched at 5 cm. intervals. The ends of 25° 13° FIG. 2. Trails made by Lashts umbratits crawling over smoked kymograph paper; the actual tracings are covered with white paint to demonstrate the width of the paths. Upper tracing ant No. 124, 25° C. Lower tracing, same individual at 13°. At the higher temperature the hind legs are further apart and with many specimens make an unbroken tracing : see text. this tube were lightly plugged with moist cotton and were attached by rubber tubing to spirals of glass tubing, the free ends of which extended out of the thermostat. Thus air adjusted to a given temperature could circulate through the system. This we think necessary since ants died when left overnight if the 15 cm. glass tube was corked at both ends, but survived if it was lightly plugged with cotton. A thermometer placed in the 15 cm. glass tube indicated that about ten minutes was necessary for the air within the tube to reach the same temperature as the surrounding water. Actively crawling ants were taken from the two genetically homo- geneous formicaries of /,. unibnilus and L. nigcr, placed in the 15 cm. glass tube. and. after an interval of at least ten minutes, were clocked with a stup watch as they crawled between the 5 cm. lines. \Yhen necessary the tube was gently tapped to induce crawling. Actively crawling specimens were used because observations over a long period indicate a rhythm of activity and rest in the colony as a whole. TEMPERATURE AND LOCOMOTION IN ANT 309 A diffuse light of approximately 25-foot candles flooded the entire thermostat. \Ye do not regard the light conditions as a disturbing factor. Data The data accumulated for L. itinbratus are graphed in Fig. 3. The ordinates represent the log of the rate of progression: X 100. time in seconds for 5 cm. The abscissae are the reciprocals of the absolute temperatures. These units were used so that the data could be applied to the Van't Hoff- Arrhenius equation. 1.60 1.40 1.20 1.00 .80 I .0033 .0034 .0035 FIG. 3. Rate of creeping as controlled by temperature in Lasiits itinbratus. The log of the progression rate, — - X 100, is plotted against time in sees, for 5 cm. the reciprocal of the absolute temperature (abscissae). Each point is the average of 8-20 stop watch readings. Because of the scattering of points, no certain line could be drawn. Each point on the graph represents from 8 to 20 readings. The temperature range is limited by the facts that below 10° C. it is very difficult to induce crawling, while above 30° C. the specimens are injured, some dying at the latter temperature. The graph clearly shows that the speed of progression is influenced by temperature. On the basis of centimeters per second, the increase is from .3 cm. per second (average) at 10° C., to 1.6 cm. per second at 24° C. Sample time records for 5 cm. runs are shown in Table I. The value of p., the energy of activation of the catalyst, in the Van't Hoff-Arrhenius equation, was called the temperature characteristic by 310 T. C. EARXKS AND H. I. KOHN TABLE I Rate of locomotion. Mean time, based on 8 trials, ant crawling 5 centimeters in each. Temperature Individual Time 'C. p 120 (umbratus) seconds 11.6 18 114 (umbratus) 4.3 16 97 (niger) 7.1 20 97 (niger) 5.5 25 99 (niger) 2.8 Crozier (1(>24), when he applied the equation to hiological phenomena. The form of the equation we used was: Rl 2\Tl Rl is the rate at T1, and R2 is the rate at T~. T is the absolute tempera- ture. In using the equation a line is drawn through the graphed points, and the rates are then calculated from it. In Fig. 3 it will be noticed that there is a greater scatter of points in the higher range. Furthermore, since the lower range also exhibits scattering, we do not put too much faith in the rectilinear relation. However, we have calculated a value of //. for it. which is 22,500 calories. The data accumulated for L. ii'ujcr are graphed in Fig. 4. The linear relationship between temperature and speed is evident, and a sharp break at 20° C. is easily seen. From 15° to 20° C. //, has a value of 10,700 1.51 1.31 1.11 .0033 .0034 FK, 4. K.iir nf crn-ping as controlled by temperature in f.iishts niger. Ab- <1 ordinuti -s mi s.-nnc 1>;i>^ as in Fig. 3. Note the clear-cut break at 20° C, the critical temperature. TEMPERATURE AND LOCOMOTION IN ANT 311 TABLE II Values of n (Temperature Characteristic Determined) Species Activity Temperatures °C. M L. umbratus Speed of progression 10-19 22 500 L. umbratus Frequency of leg movements 12-25 1? 700 L. niger Speed of progression 16-20 10 700 L. niger , Speed of progression 20-25 2? 900 calories; from 20° to 25° C., a value of 22,900 calories. Sample time records are shown in Table I. The foregoing values of /t are based on speed, which itself must be dependent in part on the frequency of leg movement — a difficult quantity to determine. Several attempts to do this, however, were made. The method was simply to observe the ant in the thermostat and clock the time for ten steps of the metathoracic leg. At low temperatures this is not difficult, but in the higher range the frequency determined can- not be very accurate. L. umbratus, Specimen No. 128, showed the fol- lowing mean frequencies per second, based on 7 readings for each temperature: 12° C, 2.13; 18° C., 3.45; 25° C., 5.55. This gives //, the value of 12,700 calories. DISCUSSION In the case of L. umbratus we have experienced difficulty in fitting a curve to the experimental data for speed of progression, so that we might utilize the Van't Hoff-Arrhenius equation. Our data plainly show that speed of progression increases as the temperature rises, as do leg-spread and frequency of leg movement. Furthermore, leg-spread and frequency of leg movement no doubt largely determine speed of progression. But each of these three has its own rate of increase. We desire to point out that the locomotion of ants involves a number of complex factors. Crozier and Stier (1925) have shown that in the tent caterpillar speed of progression is not a simple exponential function of l/Tabs, but that frequency of abdominal peristaltic locomotor waves, a factor of the former, is, giving the value of 12,200 calories. Crozier (1925) pointed out that the value for ^ of 12,200 calories in arthropods has been found for various rhythmic types of neuro-muscular activity (respiratory excepted). These include speed of progression of a diplopod, the calls of the katydid and cricket, and the flashing of the firefly. Also, our colleague Professor G. Evelyn Hutchinson (1928) 21 312 T. C. I!. \KXKS AND H. I. KOHX has found a n of 11.5UO for the >peed of progression of an interesting South African i>opnd. J'n-ntoiciis cupcnsis. CrozR-r's fis for Shaple\ -'s 1920 Liometopum data were 25,900 for below W C, and 12,200 for fn.m 16 to 38° C. Our findings roughly approximate these, except in the case of L. umbratus, as can be seen in Table 11. It should be remembered that the data for speed of progres- sion in L. ninhiMtiis are none ton good a basis for the calculation of //. SUMMARY 1. In the progression of L. nuibratus the leg-spread and the fre- i|Uency of leg movements increase as the temperature rises. _'. '11 1 of progression as inlluenced by temperature- in L. niger shows a critical temperature at 20° C., and yields two values for p.: 10,700 between 16-20° C., and 22,900 between 20-25° C. LITERATURE CITED MII;, \V. J., 1924. Critical Thermal Increment for the Locomotion of a Diplo- pod. Jmir. (icn. Phyxiol., 7: 123. CKOXIF.R, \Y. J., AND T. B. STIER, 1925. Temperature C'haracteri.stic for Locomotor Activity in Tent Caterpillars. Jour. (icu. /'hysi,>l., 9: 49. FIKI.IIK. A. M., 1905. Trmperaiure as a Factor in the Dr\ rl. .pment of Ants. Bio!. Bull., 9: 361. II lAia.AMi, II., 1931. A Study of the Physiology of Learning in Ants. Jour. Gen. Psycho!.. 5: 21. Ili-'K iiiv-ox, G. E., 1928. On the Temperature Characteristics of Two Biological Processes. South .-Ifrlcan Jour. Sci., 25: 338. SHAPLEY, H., 1920. Thermokinetics of Liometopum apiculatum Mayr. Proc. Nat. Acad. Sci., 6: 204. SiiAi'i.i.v. II., ]<;24. Xote on the Thermokinetics of Dolichoderine Ants. Proc. Nat. Acad. Sci., 10: 436. WIIKI.I.I-.K, \\'. M., 1913. Ant>. Xe\v ^"ork : Columbia University Pr< OSMOTIC PROPERTIES OF THE ERYTHROCYT IV. Is THE PERMEABILITY OF THE ERYTHROCYTE TO DECREASED BY NARCOTICS? M. H. JACOBS AND ARTHUR K. PARPART (From the Department of Physiology, University of Pennsylvania and the Marine Biological Laboratory, Woods Hole, Massachusetts) I In the course of studies connected with his well-known " Haftdruck " theory, Traube (1908) made the observation that certain lipoid-soluble substances, such as amyl alcohol, which are also known to possess nar- cotic properties, may, under appropriate conditions, exert an anti- hemolytic effect. This effect he interpreted as being due to " eine Ver- dickung der Lipoidschicht welche die Stabilitat der Blutkorperchen gegen andere Hamolytika andern muss." A few years later Arrhenius and Bubanovic (1913) obtained similar results with chloroform, ethyl ether, ethyl and amyl alcohols, and benzene. As an example of these results one typical experiment with a hypotonic solution containing chloroform may be cited. In this experiment it was found after one hour at 37° C., followed by several more in the ice-box, that the per- centage of hemolysis decreased from 60 per cent in the absence of chloroform to a minimum of 37 per cent in the presence of 0.2 per cent, rising again at higher concentrations. Arrhenius and Bubanovic did not hesitate to conclude that " diese Wirkung beruht vermutlich auf einer Verlangsamung des Eindringens von Wasser in die Zellen." Among the more recent workers in this field, Yoshitomi (1920) found after exposures of 2 to 18 hours to hypotonic solutions, a lesser degree of hemolysis in the presence of certain concentrations of ether, chloro- form, chloretone and amylene hydrate than in their absence. He sug- gested that this effect might be due either to a less ready entrance of water into, or a more ready escape of salts from, the cells. Jarisch (1921) also obtained an inhibition of osmotic hemolysis with alcohol, ether, amylene hydrate and urethane. While he did not state in very precise terms his conception of the antihemolytic action of these sub- stances, his use in this connection of the term " wasserhemmend " would seem to indicate that his views were not unlike those of Arrhenius and Bubanovic. 313 314 M. H. JACOBS AXD A. K. PARPART Since these and similar results obtained with the erythrocyte have frequently been cited in -upport of the view that narcosis is associated with a decreased cell permeability, it seems necessary for us to point out that such evidence is entirely inconclusive. In the first place, osmotic hemolysis is a complicated process involving not only (a) the entrance of water into the crythrocyte, but (b) the escape of hemoglobin, (c} the possible los> from the cell of salts and other osmotically active substances (see in this connection Ponder and Saslow, 1931) as well as (rf) a vanity of possible chani;e> of diverse nature, not directly associated with permeability, in what is commonly and rather loosely called the "osmotic resistance" of the cell. While it is undoubtedly true that anything that decreases the permeability of the cell to water (factor o) will, in general, tend to delay osmotic hemolysis, the reverse statement is by no means true. A delay might equally well be produced by the operation of any or all of factors b, c, and d. \ more fundamental objection, however, to evidence of the type mentioned above, is that it is very unlikely that the investigators in question obtained from their experiments any real information as to the rate of hemolysis, which is the thing of greatest importance in connection with questions of permeability to water. It is, of course, conceivable that if at the end of some single arbitrarily selected time a lesser degree of hemolysis is obtained in the presence than in the absence of a nar- cotic, the difference might be due to a slower rate of progress in the former case towards the same final end state of the system. On the other hand, the possibility must lie considered that the narcotic exerts its effect primarilv on the degree of hemolysis ultimately attained; in that case, with no knowledge of the position of equilibrium towards which the system is proceeding — or which it may indeed have reached at the time the observation is made — no valid conclusions whatever can be drawn as to the fundamental rate of the hemolytic process. Though this principle would seem to be a self-evident one, it has • •n very frequently disregarded in the past, not merely in studies on hemolysis, but in other fields of physiological work as well. It should In- most strongly emphasized, therefore, that while it is possible to study a position ,,f equilibrium with no exact knowledge of the rate at \vhieli it is attained, the reverse procedure of attempting to draw con- clusions coneerning a rate, with no information whatever as to the end state which the given system is approaching, is entirely unwarranted and n lead only to confusion (see in this connection Jacobs, 1928). In the pre.M-nt paper the distinction between the effects of certain urethanes on the "equilibrium" and the "rate" factors concerned in osmotic will be illustrated; and it will further be pointed out that since NARCOTICS AND PERMEABILITY 315 as far as we are aware the work of previous investigators in this field makes no such distinction, it can be expected to throw no real light on the question of the permeability of the erythrocyte to water. II As a first step in the separation of the two types of factors, it seemed important to obtain curves representing the entire course of hemolysis from its beginning until further change had ceased. This necessary type of information, which, as has been mentioned, has apparently not been supplied by previous workers, is particularly easy to obtain by the method of one of the authors (Jacobs, 1930). A typical experiment is represented in Fig. 1 in which the erythrocytes were those of the ox and the narcotic was ethyl urethane. The blood in this, as in all the 10 20 30 40 50 60 70 80 90 TIME-MINUTES FIG. 1. Course of hemolysis of ox blood in buffered 0.090 M NaCl in the presence (lower curve) and absence (upper curve) of 0.3 M ethyl urethane: pH, 7.4 ; temperature, 20° C. other experiments here described, was thoroughly and almost instan- taneously mixed with the solutions used in the proportion of approxi- mately 1 to 500. Kymograph records were made of the ensuing hemolysis and from these the curves in the figure were reconstructed. Because of the enormous importance of pH and temperature changes in experiments involving equilibria, the precautions described by Jacobs and Parpart (1931) were employed in all such experiments, the pH being kept almost constant in this case at approximately 7.4, and the temperature at 20° C. An inspection of Fig. 1 shows very clearly that under the conditions of this experiment the effect of ethyl urethane upon the final end-point reached by the system is far more striking than any possible effect it might have upon the rate of hemolysis as such. Thus, in the absence of the urethane, a degree of hemolysis of approximately 61 per cent was attained in between 20 and 30 minutes, and after that time no further 316 M. H. JACOBS AND A. K. PARPART change occurred; in its ] HTM nee a final degree of hemolysis of 48 per cent was reached in about the >ame time, and this likewise underwent no further change. Yen >imilar results were obtained in a number of other experiments. It i- evident, therefore, that no information about the rate of henioK-is ran be secured in such cases by observing at the end of some arbitrarily selected time the mere degree of hemolysis that then happens to exist. It is to be noted that the antihemolytic effect of ethyl urethane is not a simple osmotic one due to the greater total concentration of the solution containing it. Such an effect could be obtained only with a non-pene- trating MI! stance, since otherwise the solute molecules would distribute themselves inside and outside the cell in such a way that their osmotic ects would everywhere balance. In the case of the urethanes. how- r. not only is the penetrating power for cells in general known to be extremely hi^h, but a number of hematokrit measurements made in the urse of these experiments showed the absence of any measurable osmotic effects on cell volume. As a matter of fact, though the con- centration of the ethyl urethane in this particular experiment was 0.3 M. the antihemolytic effect was only that which would have been produced osmotically by an increase in concentration of possibly 0.006 M in a solution of a non-penetrating non-electrolyte. As to the effect of ethyl urethane on the fundamental rate of hemolysis, apart from that on the final degree attained, this cannot be determined by a mere inspection of such curves as those in Fig. 1. Even though a measurably longer time were required to attain a given degree of hemolysis in tin- presence than in the absence of the narcotic, it would be impossible to be certain, without a fairly complicated mathe- matical analysis of the results, how much of the observed effect was due merely to the shift in the final equilibrium. Since, therefore, even under the relativelv favorable conditions provided by the possession of two complete hemolysis curves it is very difficult to draw conclusions about the effect of a narcotje on the rate of the process what information of vahn- could conceivably be obtained from the knowledge of only a single point on each curve? A.S a matter of fact, most of the observations mentioned in the introductory paragraph were- taken at times long after those at which the final equilibrium mu-t have been attained and could, therefore, by no possibility throw any liidit upon the rate of hemolysis, and by implication, upon the possible rate of entrance of water into the ('11-. Though, as will be shown later, the conclusion of previous in- ve-ti.^ators that the rate of entrance of water into the erythrocyte is .-lowed by the presence of a narcotic may in itself be entirely correct, it may not validly be drawn from the data they have presented. NARCOTICS AND PERMEABILITY 317 Having found in experiments of the type of that represented in Fig. 1 that the final equilibrium condition is usually reached within 20 or 30 minutes, and in any case in less than an hour, it seemed desirable to study by more quantitative methods than those previously used the effects of varying the concentration of the narcotic. In Table I are given the results of one such experiment with ethyl urethane in which the percentages of hemolysis reached in one hour in the presence of dif- ferent concentrations of this substance were determined. The concen- tration of NaCl used was selected as a favorable one for the particular sample of ox blood employed. The pH in this case was approximately 7.35 and the temperature, as before, 20° C. It will be noted that as the concentration of urethane increased, the degree of hemolysis decreased until a maximum effect was reached probably somewhere between TABLE I Degree of hemolysis attained by ox crylhrocytes in one hour in a buffered 0.085 M Nad solution containing different concentrations of ethyl urethane. pH, 7.35; temperature, 20° C. Concentration Percentage of of urethane hemolysis 75 0.0078 75 0.0156 74 0.0313 74 0.0625 70 0.125 70 0.25 68 0.5 73 1.0 100 0.25 M and 0.5 M. Beyond this point a hemolytic effect of the narcotic became evident. The general results of this entirely typical experiment with ethyl urethane do not differ in principle from those obtained by Arrhenius and Bubanovic with chloroform, if it be assumed, as was almost certainly the case, that what these investigators measured was the final position of equilibrium of the system. In another experiment, whose results are given in Table II, the gen- eral procedure was reversed by keeping the concentration of urethane constant at 0.1 M and varying that of the hypotonic solution. In this way the antihemolytic effect of the narcotic on a considerable propor- tion of all the erythrocytes in the blood could be observed. It will be noted that although the results show certain minor irregularities, there is in no case any departure from the previously observed antihemolytic effect. Similar, though less complete, results were also obtained with several other urethanes, as well as with ethyl alcohol. Taking to- gether these results and those reported by previous workers, the evidence 318 M. H. JACOBS AND A. K. PARPART seems to be entirely consistent that the degree of hemolysis ultimately attained in a hypotonic N'lution may be reduced by a variety of narcotic substances in proper concentrations. This fact, however, though of interest in other ways, throws little or no light upon the question of the effect of naroitics upon the permeability of the erythrocyte to water. \Yhat is needed is information not about the final equilibrium but about the fundamental rate at which hemolysis occurs. Ill It has been pointed out in another place by one of the authors (Jacob-. 1('JS) that in cases where the position of equilibrium of a TABLE II Degree of hemolysis attained by ox erythrocytes in one hour in buffered NaCl solutions of different concentrations in the absence and presence of 0.1 M ethyl urethane. pH, approximately 7.35; temperature, 20° C. Percentage c f Hemolysis Urethane absent Urethane present 0.098 82 78 0.099 78 74 0.100 73 69 0.101 71 65 0.102 69 59 0.103 67 55 0.104 63 53 0.105 57 48 0.106 53 44 0.107 50 37 0.108 45 38 0.109 41 33 0.110 34 22 hcinolytic system is influenced by the >ame factor whose effect on the rate of hemnlysi.s it i> desired to study, the general rule should be fol- lowed of keeping the s\>irm at all times as far away from equilibrium conditions as possible. Thus, in investigating the effect of temperature upon the rate of osmotic hemolysis, it was found that where only water m- very strongly hypotonic .solutions were employed consistent and plausible results eonld lie obtained, while with less strongly hypotonic the results were erratic and at first sight inexplicable. Simi- larly, in the case of narcotics, which have been shown in the preceding lion to affect in a striking manner the degree of hemolysis finally at- tained, the only satisfactory method of studying their effect on the rate NARCOTICS AND PERMEABILITY 319 of hcmolysis as such would be to work with very strongly hypotonic solutions, or preferably, with distilled water. Most methods for study- ing hemolysis are much too slow for use in experiments of this type— which doubtless accounts for the fact that they have apparently not here- tofore been made. The method of one of the authors (Jacobs, 1930) is entirely suitable for this purpose, however, and some results obtained with it may now be described. In Table III are indicated the effects on hemolysis by water of the addition of different amounts of ethyl, n-butyl, i-amyl and phenyl urethanes. Though for such short times the experimental errors are TABLE III Effect of Various Urethanes on the Time Required for 75 Per Cent Hemolysis of Ox Blood by Distilled Water Ethyl Urethane n-Butyl Urethane i-Amyl Urethane Phenyl Urethane * Concen- Time Concen- Time Concen- Time Concen- Time tration tration tration tration seconds seconds seconds seconds — 1.5 — 1.3 — 1.30 — 1.25 0.0313 1.5 0.0015 1.4 0.0003 1.30 0.00025 1.25 0.0625 1.5 0.0031 1.4 0.0013 1.28 0.0005 1.30 0.125 1.7 0.0062 1.5 0.0025 1.32 0.001 1.30 0.25 2.0 0.0125 1.6 0.005 1.42 0.002 1.35 0.5 2.2 0.025 1.5 0.01 1.50 0.004 1.32 1.0 2.6 0.05 2.1 0.008 1.52 * See also Table V. relatively large, it will be noted that the results are on the whole entirely consistent and that for each substance there is a slight but unmistakable retardation of hemolysis as the concentration increases. Furthermore, the effectiveness of the different urethanes, as might be expected, proves to be very different. As judged by the dilutions at which a retardation of hemolysis first becomes apparent, the order of effectiveness is: ethyl < n-butyl < i-amyl < phenyl. This is not only the order in which narcotic effects are usually mani- fested by these substances, but there is a rough quantitative agreement between the concentrations at which hemolysis begins to be retarded and those found by Dr. E. B. Harvey to be effective in reversibly suppressing the cleavage of the Arbacia egg. The latter concentrations, as cited by Lucke (1931), are as follows : ethyl, 0.05 M to 0.2 M ; n-butyl, 0.0125 M to 0.05 M; i-amyl. 0.005 M to' 0.01 M; and phenyl, 0.00125 M to 0.005 M. 320 M. H. JACOBS AND A. K. PARPART In the case of ethyl nrethane, where the effective concentration is fairly high, it is conceivable ihat at least a part of the delay in hemolysis may be due to osmotic factors. F.ven a substance that penetrates a cell as rapidly as a urethane might, if sufficiently concentrated, slow to a measurable extent tin.- rate of attainment of the final osmotic equilibrium between the cell and its surroundings, while having no direct effect upon the position of the equilibrium. Whatever may be the validity of this objection in the case of ethyl urethane. however, it is certain that it can- not hold in the case of the other three substances, where the effective concentrations are of the order of 0.01 M to 0.001 M. It has been -hown elsewhere by one of the authors (Jacobs, 1932) that even in the : a completely non-penetrating non-electrolyte such as saccharose, the time required for hemolysis in a 0.01 M solution differs from that in TABLE IV Times required for 75 per cent hemolysis of ox blood in NaCl solutions containing different amounts of ethyl urethane. Temperature, 20° C. • nt ration of I Concentration of XaCl Water 0.02 M 0.04 M 0.06 M 0.07 M 0.073 M 0.076 M 0.08 M seconds 1.5 1.5 1.5 1.7 2.0 2.2 2.6 seconds 4.0 4.2 4.0 4.0 4.7 4.8 5.1 seconds 5.7 5.6 5.8 6.2 6.5 7.(l 7.1 seconds 7.0 7.5 7J 8.3 9.1 9.7 10.3 seconds 10.3 11.7 12.5 16.0 20.3 39.0 32.6? seconds 18.0 19.8 18.9 27,s 54 11(1 274 seconds 22.5 26.8 34.4 56.5 103.5 280 317 seconds 70.2 73.4 96.5 156 332 832 420 0.0313 M 0.0025 M . 0.125 M . 0.25 M. . 0.5 M . . 1.0 M water only to a barely measurable extent, the difference for ox erythro- cytes being perhaps 0.1 second. With lower concentrations than 0.01 M of substances of extremely high penetrating power such as the urethanes, direct osmotic effect-- could certainly not be measured; and any effects that could be measured would therefore necessarily be of a more specific nature. Though the retardation of hemolysis in water by urethanes is apparently always small in amount and the errors of the determinations arc relatively large, the results obtained in these and in other experiments have been sufficiently consistent to leave little doubt that the substances in question are able to affect the rate of hemolysis under conditions where any shift in the final theoretical equilibrium is of negligible im- portance. Such an effect, though erroneously inferred by other workers from their experiment-, has not, we believe, previously been demon- strated. NARCOTICS AND PERMEABILITY 321 In concluding the presentation of experimental data there may be added in Tables IV and V figures showing the gradual transition from conditions where the fundamental rate factor is primarily concerned in determining the time of hemolysis to those where the equilibrium factor tends to dominate the situation. In the light of the facts presented in the preceding section, it is evident that the striking effects produced by all the substances in the most concentrated salt solutions are due chiefly to a change of the hemolytic end-point in the direction of a reduced final degree of hemolysis. The rate factor as such cannot be studied in such solutions. In the case of ethyl urethane, one additional factor appears in the last figure of the last column, namely, a direct hemolytic effect of TABLE V Times required for 75 per cent hemolysis of ox blood in NaCl solutions containing different amounts of pJienyl urethane. Temperature, 20° C. Each figure for water and for the lower concentrations of NaCl is the average of four determinations. Concentration of Urethane Concentration of NaCl Water 0.02 M 0.04 M 0.06 M 0.075 M 0.08 M 0.085 M seconds 1.50 1.50 1.55 1.52 1.70 1.95 seconds 3.87 3.88 3.85 3.90 3.93 seconds 5.17 5.17 5.23 5.20 5.30 5.38 seconds 6.92 6.70 6.72 7.05 7.58 7.95 seconds 11.1 11.8 12.8 14.5 48.5 54.2 seconds 19.8 28.5 30.0 76.5 86.0 121.0 seconds 41 46 55 115 870 0.0005 M 0 001 M 0.002 M 0.004 M 0.008 M the urethane in very high concentrations ; but this factor is of compara- tively small importance in connection with the present experiments. One other point suggested by the results obtained with i-amyl and phenyl urethanes may be mentioned briefly, namely, that whereas the retarding effect of urethanes on hemolysis by water never failed to appear in our experiments, it was sometimes more doubtful in the case of the most dilute salt solutions, e.g., 0.02 M. The experiment chosen for representation in Table V was selected to show this condition, which was, however, not invariably observed. We are unable to suggest a reason why the retarding effect of certain urethanes may at times be more pronounced in the absence than in the presence of salts, but may point out that a somewhat similar but much more striking effect has been observed by Lucke (1931) in the case of the Arbacia egg. His ex- planation is that if the entrance of water into the cell has already been strongly retarded by the presence of salts (especially those of calcium), narcotics are unable to produce any further effect. In the case of the 322 M. H. JACOBS AND A. K. PARPART erythrocyte it will be shown in a later paper that very low concentra- tions of salts, including XaC'l. are effective in some non-osmotic man- ner in retarding osmotic IK-UK >1\ sis. It is conceivable, therefore, though by no means certain, that a similar principle is involved in the two cases. IV It has been shown in the preceding sections that various urethanes in the proper concentrations are able to reduce the degree of hemolysis finally attained in certain hypotonic salt solutions. This increase in the osmotic resistance of the cells, which has been noted by a number of previous observers, is an equilibrium rather than a rate effect and is best seen in Dilutions whose concentrations are such as to cause the disap- pearance of some but not all of the erythrocytes in the given sample of blood. It has also been shown that under conditions where complete hemolysis is very rapidly produced and where possible changes in the theoretical position of equilibrium of the system are of negligible im- portance, narcotics are able to bring about a slight but consistent slowing of the rate of the process. This effect, which is best seen in distilled water and which has frequently been confused with the one first men- tioned, has not as far as we are aware, previously been demonstrated, though from the standpoint of cell permeability it is the more important of the two. Its possible nature will be considered after attention has first been given to the more striking and better known change in the osmotic resistance of the cells. As has been mentioned above, there are at least four different ways in which the osmotic resistance of the erythrocyte might be affected (factors a. /'. r and d on page 314). Of these factors, the first, namely, a change in permeability to water, may almost certainly be ruled out as a possible cause of any change in the position of final equilibrium of the In molytic system. ( >nlv if the cells at some point became completely impermeable to water could this factor alone do more than change the rate at which the equilibrium is reached; the equilibrium itself would remain unaltered. As a matter of fact there is an abundance of evi- dence that the erythrocytes do not at any time become completely im- permeable to water. On the other hand, a changed | enneability to salts and other osmoti- cally active substances contained within the cell (factor r) might con- ceivably alter its osmotic resistance. Since hemolysis by hypotonic solu- tions i- due to an excess of osmotic pressure within the cell, any escape of materials that reduced this excess would not only slow the rate of hemolysis in all cases, but in certain critical cases would prevent its oc- currence altogether. The possibility of a leakage of salts from the NARCOTICS AND PERMEABILITY 323 erythrocyte in ordinary osmotic experiments has been emphasized by Ponder and Saslow (1931), and it is by no means inconceivable, or in- deed unlikely, on a priori grounds that such a leakage might be favored by narcotics. In sufficiently high concentrations these substances tend by destroying the erythrocyte to permit a very ready escape of materials from its interior ; and it is entirely possible that in lower concentrations they might injure its surface sufficiently to increase any loss of electro- lytes already in progress. Though this theory has a certain degree of plausibility, it neverthe- less seems necessary to discard it in view of the direct evidence obtained by Siebeck (1922) that narcotics actually reduce to an easily measurable extent the rate of exchange of ions between the cell and its surroundings, and that furnished by Joel (1915) that the gradual increase in the elec- trical conductance of a suspension of erythrocytes is slowed rather than accelerated in the presence of such substances. As far as the available evidence goes, the effect of ordinary concentrations of narcotics would seem, if this factor were of importance, to be in the direction of reducing rather than of increasing the osmotic resistance of the erythrocyte. Turning next to factor d, which involves some change or changes in the osmotic resistance of the cell not associated with permeability factors, the possibility suggested by Traube that the narcotic may produce in some way a thickening or a strengthening of the cell membrane and so oppose hemolysis may first be considered. Such an explanation ap- pears to be an unlikely one in view of the fact that the surface of the erythrocyte seems normally to offer little resistance to osmotic volume changes. (See in this connection Jacobs 1931, 1932.) As a matter of fact, the increased resistance in the presence of, for example, 0.3 M ethyl urethane, which is by no means the greatest effect we have ob- served, may correspond to a change in the critical concentration of NaCl by 0.003 M, amounting in terms of osmotic pressure to perhaps one- eighth of an atmosphere. That the delicate cell membrane could be strengthened to support this excess of pressure does not seem very likely. A much more plausible possibility is that the narcotic may in some way have a tendency to cause a diminution in the volume of the cell and so to oppose its swelling in hypotonic solutions. Effects of this sort are already known in the case of other agents. For example, the increased osmotic resistance of erythrocytes in alkaline media and at high tem- peratures (Jacobs and Parpart, 1931) and in solutions of non-electro- lytes (Jacobs, 1932) is probably to be accounted for in this way. As a matter of fact, v. Knaffl-Lenz (1918) has reported a decrease in the volumes of erythrocytes, as measured by the hematokrit, on the addition 324 M. II. JACOl'.S AXU A. K. PARPART of certain narcotics, though the times required to produce this effect in his experiments were much longer than those involved in the present series; and. furthermore, his results were indecisive in the case of the only urethanc IK- used. \\ c have been unable to detect with certainty by the hemat< ikrit method any such differences in volume in the case of ethyl urethanc solutions, though in view of the rather large errors of the hematokrit method and the very slight volume changes required to pro- duce a considerable difference in the observed percentage of hemolysis (Jacobs and I'arpart, 1^31) we do not feel that this possibility has been entirely ruled out. The la>t factor that will be discussed is the second of those men- tinned above, namely, the escape of hemoglobin from the cell. This lor has frequently been neglected in studies on osmotic hemolysis in the past owing, no doubt, to the old belief that hemolysis is produced by an actual bursting of the cell when the internal pressure has reached a sufficiently high point. If this were the mechanism of hemolysis, then the escape of hemoglobin would, in fact, be an unimportant part of the process. It seems certain, however, from the phenomenon of " re- \ersible hemolysis." so-called, that the cell is not ordinarily ruptured by mild hemolytic agents, but that at a certain time, as a result of stretching or some other change in its surrounding membrane, the latter becomes permeable to the hemoglobin contained within the cell. This permeabil- ity to hemoglobin is reached in such a sudden and definite manner that osmotic hemolysis is apparently an " all-or-none " phenomenon, i.e., up to a certain point no hemoglobin escapes from the cell ; beyond that point an almost infinitesimal increase in the volume of the cell results in the lice outward diffusion of all of its hemoglobin (Saslow, 1929; Parpart, 1931). I n fortunately, we know ton little at present about the physical state of the hemoglobin within the cell and the possible effects of changes in this state on its diffusibility. It is usually assumed, however, in the absence of evidence to the contrary, that under all usual conditions w<- have to do with a Dimple aqueous solution of hemoglobin and that the possibility of its escape from the cell depends merely on the character of the cell membrane. The assumption that the escape of hemoglobin from the erythrocyte depends primarily on the cell membrane may or may not be true. It is of interest, however, to see whether it can be made the basis of a plaus- ible explanation of the effect of narcotics on osmotic resistance. There is considerable evidence that at the- surface of the erythrocyte in addition to lipc.id substances which gi\e to the cell certain of its physical prop- erties i Mudd, S.. and K. Ii. II. Mudd. T»Jf>) and which perhaps deter- mine its free p.-rmeability to all lipoid-soluble substances, there are NARCOTICS AND PERMEABILITY regions through which water, ions, and non-lipoid-soluble organic sub- stances of low molecular weight can pass. Though the exact structural nature of these regions is not known, they may, at least in a semi- figurative sense, be called " pores." A further discussion of this theory as applied to the crythrocyte is given by Mond and Hoffman (1928) and by Jacobs (1931). Whatever may be our ideas of the exact nature of the hypothetical '' pores " in the cell membrane, it must not be forgotten that certain purely objective facts are well known; namely, that non-lipoid-soluble molecules of sufficiently low molecular weight pass through the wall of the erythrocyte readily, those of higher molecular weight more slowly, and those whose molecular weight (or molecular volume) exceeds a certain size fail to do so at all. The hemoglobin molecule, of course, enormously exceeds the critical size for penetration. Nevertheless, in osmotic hemolysis a point is somewhere reached where rather suddenly the cell becomes permeable to hemoglobin. Without attaching too literal a meaning to the statement, we may say that at this point the " pores " have been enlarged sufficiently to permit the escape of this molecule. Now we have a certain amount of experimental evidence that nar- cotics are able — presumably by adsorption — to diminish the size of the pores in artificial membranes, or, at any rate, to render more difficult the passage of certain substances through these membranes (Anselmino, 1928 a, b}. Suppose that the same were true of the erythrocyte at the point where it undergoes hemolysis. In this case, the presence of a sufficient concentration of a narcotic substance might be expected to con- vert a " pore " that would otherwise just permit the passage of hemo- globin into one that would just fail to permit it. Further swelling would be necessary to cause hemolysis. The osmotic resistance of the cell would thereby be raised, just as it is known to be in fact. Furthermore, the effectiveness of weakly adsorbed narcotics would be less than that of strongly adsorbed ones and, again, there is a parallel between the adsorbability of different urethanes and their ability to prevent hemolysis. Accepting in a purely tentative manner this explanation of the effect of narcotics upon the final equilibrium of a hemolytic system, how would such an explanation fit the known facts concerning the rate at which hemolysis occurs in very strongly hypotonic solutions? It is entirely conceivable that in such solutions the rate of osmotic hemolysis might be affected either by a slowing of the rate of entrance of water or by a slowing, — though not a prevention — of the escape of hemoglobin by a delay in the attainment of the proper condition of the pores, or by a combination of both factors. Since it is unlikely, with a rate of in- crease of cell volume as rapid as that in distilled water, that the delay 326 M. H. JACOBS AXD A. K. PARPART in the escape of hemoglobin would be very great, it seems entirely pos- sible that at least a part of the observed effect of narcotics on the rate of hemolysis by water may be clue to an actually decreased rate of pene- tration of this substance. It is to be noted, however, that the possible effect must in any case be rather slight. In this connection it is of interest to consider the work of Siebeck (1922) on the effect of narcotics on the rate of passage of ions between the erythroe\tc and its surroundings and also that of Anselmino and Hoenig (1930) on the entrance of the non-electrolytes erythritol, ara- binose. xylose, etc. In the former case, actual chemical analyses were made at several intervals and it is therefore virtually certain that the permeability of the cell to the substances in question was dealt with directly. In the work of Anselmino and Hoenig, though the methods were not quite so direct, it is also very likely that their interpretation of their results as indicating a production by narcotics of a decreased permeability to various slowly penetrating non-electrolytes is correct. It is perhaps significant that the decrease in permeability to ions and to rather slowly penetrating non-electrolytes is much greater than any de- crease for water that could be inferred from the present experiments. If the " pore " theory were correct, it would be expected that the hypo- thetical diminution of the pore diameter produced by narcotics would exert an effect upon permeability which would become proportionately greater as the size of the molecule increased. The water molecule, being the smallest of those commonly supposed to enter the erythrocyte in this manner, would be affected least of all. It shonld be emphasized that this explanation of the manner in which narcotics may conceivably affect osmotic hemolysis is suggested merely as a convenient working hypothesis. Its chief advantages are that it explains in essentially the same manner both " rate " and " equilibrium " effects and that, as far as we are aware, it is not incompatible with any known facts. It is by no means necessary, however, that the rate and equilibrium effects should be explained in the same way; in the case of temperature, for example (Jacobs, 1928), they seem almost certainly to be of a different nature. It is entirely possible that at any time facts may come to light with which the present theory is inconsistent ; in that case it may readily be abandoned without greatly changing the signifi- cance of the experimental data here presented. \Ye are glad to acknowledge our indebtedness to Dr. Balduin Lucke tor supplying most of the urethanes used in this work, and to Ethel R. I'arpart and ( ;. E. Shattuck for assistance in connection with several of the 'its. NARCOTICS AND PERMEABILITY 327 SUMMARY 1. The observation of previous investigators that narcotic substances in proper concentrations tend to oppose osmotic hemolysis is confirmed in the case of several urethanes. 2. It is shown that the conclusion frequently drawn from such ob- servations, that the antihemolytic effect of narcotics is due to a decreased permeability of the erythrocyte to water, is unwarranted by the existing experimental evidence. The necessity for a separation of '" rate " and " equilibrium " factors in studies on osmotic hemolysis is emphasized. 3. It is shown by experiments in which these factors are properly separated that a slight but measurable retardation of osmotic hemolysis may be produced by low concentrations of urethanes. The possible nature of the mechanism of this retardation, which may perhaps in part involve a decreased permeability of the cell to water, is discussed. BIBLIOGRAPHY AXSELMIXO, K. J., 1928a. P finger's Arch., 220: 524. AXSELMIXO, K. J., 1928ft. Biochcm. Zcitschr., 192: 390. AXSELMINO, K. J., AXD E. HoENiG, 1930. Pfliigcr's Arch., 225: 56. ARRHENIUS, S., AXD F. BUBAXOVIC, 1913. Mcddcl. fran K. Vct-Akad. Nobel- institut, 2: No. 32, p. 1. JACOBS, M. H., 1928. Am. Nat., 62: 289. - 1930. Biol. Bull., 58: 104. - 1931. Ergcbn. d. Biol.. 7: 1. - 1932. Biol. Bull.. 62: 178. JACOBS, M. H., AXD A. K. PARPART, 1931. Biol. Bull.. 60: 95. JARISCH, A., 1921. Pfliigcr's Arch., 186: 299. JOEL, A., 1915. Pfliigcr's Arch., 161: 5. VON KXAFFL-LEXZ, E., 1918. Pfliiger's Arch., 171: 51. LUCKK, B., 1931. Biol. Bull., 60: 72. MOXD, R., AXD F. HOFFMAXN, 1928. Pfliigcr's Arch., 219: 467. MUDD, S., AND B. H. MUDD, 1926. Jour. Ex per. Mcd., 43: 127. PARPART, A. K., 1931. Biol. Bull., 61: 500. POXDER, E., AND G. SASLow, 1931. Jour. PhysioL, 73: 267. SASLOW, G., 1929. Quart. Jour. E.vpcr. PhysioL, 19: 329. SIEBECK, R., 1922. Arch. f. e.rpcr. Patlwl.' 95: 93. TRAUBE, J., 1908. Biochcm. Zcitschr., 10: 371. YOSHITOMI, T., 1920. Ada. Schol. Mcd. Univ. Imp. Kioto, 3: 338. 22 THE LIFE I1IST< >KY < >F PARORCHIS AVITUS (LIXTON) A TREMAT( IDE FR( >M THE CLOACA OF THE GULL HORACE \V. STl.'XKARD AND RAYMOND M. CABLE (Prom .Y.Ti1 Yi>rk University and the Marine Biological Laboratory, Woods Hole, Mass. ) The study of parasitic flat worms in the United States began with Joseph Leidy. who described a number of larval and adult trematodes. For many years progress consisted in the description of new species of adult flukes. More recently, attention has been directed to a study of the larval stages and at present one hundred and sixty-six ccrcariae have been described in the ( "nited States. < hily twenty of these larva?, how- ever, have been definitely correlated with their adult forms. It is ob- vious therefore that further advance is dependent upon knowledge of the life history of these species. Lack of such information keeps the litera- ture burdened with almost twice as many specific names as are necessary, since each larval trematode whose adult stage1 is not recognized, is de- scribed as a distinct species. Furthermore, and of more significance, knowledge of the successive stages in development is indispensable for correct identification, taxonomy, and any fundamental work on the physiology and control of these parasites. The fact that fifteen of the twenty life histories solved in this country have been described in the last four years indicates the intensitv of present interest in life history studies. This paper adds another to the list of known life histories and describe- more completely the results reported by the writers in a recent preliminary note (Stunkard and Cable, 1931). Experiments n inducted at the Marine Biological Laboratory during the summer of 1(>31 have shown that Ccrcaria scnsifcra (Stunkard and Shaw, 1931) is the larva of Purorcliis ai'itns (Linton, 1914). Parorchis azntus was originally described from the cloaca of the herring gull, Larns argcntatns. It is very similar to Parorchis acanthus, a species described by Xicoll fl'iori) from the lmr-a and cloaca of Larns argcntatns and first given the name Zcnyorchis acanthus. Subsequently Nicoll (1907a) erei t'-d tin- genus Par orchis to supplant Zcnyorchis (preoccupied) and named P. acanthus as type species. 1'itrorcliis acanthus has been the subject of important studies by l-'.n-lish investigators. Nicoll (1907/>) gave a more complete description of the species and reported its occurrence in the common gull, Larns canux. In the same work he transferred Dlstomnm plttaciuni Braun, 328 LIFE HISTORY PARORCHIS AVITUS 329 1902 to the genus Parorchis. Nicoll, in a letter to Linton, expressed the belief that Parorchis avilus is probably identical with P. acanthus. Lin- ton (1928) however, showed that P. avitus differs from P. acanthus in the same respects that P. acantJuis differs from Parorchis (Distomuui) pittaciuin, namely, the ratio of sucker diameters and the extent of uterine convolution. Lebour (1914) compared young stages of P. acanthus with Ccrcaria pur pure? Lebour, 1907 from Purpura lapillus, and con- cluded that this cercaria is the larval stage of Parorchis acanthus. She also emphasized the close affinity of the genus ParorcJiis to the genus Echinostomum and expressed the belief that in the life history of P. acanthus there must be a secondary intermediate host, probably a bivalve mollusk, as is the case with Echinostomum sccunduui (Lebour, 1908). In a later paper, Lebour and Elmhirst (1922) reported the encystment of cercariee of Parorchis acanthus, together with the cercarise of Echino- stoinuin sccunduni in the foot and mantle of Cardium cdulc and Mytilus cdulis. No feeding experiments were made in determining the life his- tory of P. acanthus, and conclusions were based entirely on morphologi- cal comparisons. METHODS AND OBSERVATIONS The present study consisted of controlled feeding experiments, since this is the only conclusive method of determining the life history. The oyster drill, Urosalpinx cincrcus, was used exclusively as a source of cercarise since the study was practically completed before it was dis- covered that Thais (Purpura) lapillus also served as host for this species. The infected snails were isolated in finger bowls. Cercarise escaped from them in large numbers and encysted on the bottom of the bowl. These cysts were scraped off with a scalpel and concentrated by decanta- tion. Attempts were made to infest terns, guinea pigs, white rats, and mice. Although the natural host of P. avitus is the herring gull, Lams arycntatus, for the experiments, nestlings of two species of terns, the common tern. Sterna hirundo, and the roseate tern, Sterna dougalli, were employed since they were more easily obtained. The newly hatched nestlings were taken from a tern rookery on Big Weepecket Island, near Woods Hole, and were kept alive in the laboratory as long as possible. These birds, which had received no food while in the nest, were fed the deep flesh of mackerel and flounders, since trematode cysts are not as abundant in these tissues as in the skin and flesh near the surface of the body. Fishes showing any signs of being parasitized were not used for food. The nestlings were kept in the laboratory for two days, or until they began to take food readily, before infestation was attempted. They were then fed large numbers of cysts by means of a pipette. The H. W. STTNKAKI) AXD R. M. CABLE first lot of 6 birds was fed cysts to determine whether infestation could be established. Five of these .lied at night and in all cases were so badly disintegrated by tin following morning that the alimentary tract had almost completely disappeared. The sixth bird refused to eat on the tenth day and was killed. Twelve immature worms in various stages of development were recovered from the cloaca. The second lot con- sisted of eight birds. Six were fed cysts on the same day and two were kept a- controls. The nestlings lived for varying lengths of time, two of the rose-ate species surviving as long as 15 days. Table I gives the results obtained with this series. TABLE I No. Species Fed Cysts Autopsied No. Worms 1 S. hirundo July 11 July 16 11 _' II 11 1 1 " 17 13 3 < ( it control " 17 none 4 tt tl 1 1 " 18 n 5 II II July 11 " 20 3 *6 11 II " " 21 none 7 11 dougalli t < " 26 13 8 ii it 1 1 " 26 28 * This bird died during the night and was badly disintegrated when autopsied, which may explain the absence of worms. It is seen that every bird which was fed cysts, and on which a favor- able autopsy could be made, became infected, while the two controls were negative. The oldest worms were recovered from the two roseate terns which lived 15 days after infestation was established. Although not sexually mature, these specimens were sufficiently developed to be identified positively as I'urorcliis trritns; this identification has been con- firmed by Professor Kdwin Linton. I'urthcr attempts at experimental infection were made, using mam- mals instead of birds. Four guinea pigs, nine rats and six mice were f'-d large- numbers of enc\sted larva- and were autopsied after varying lengths ot time. Xo worms were recovered from any of these animals. \ltliough the larv;c may possibly excyst in these mammals, it is evident that conditions are not suitable for establishing infection. STAGES IN THE LIKE CYCLE The -exually mature worms are viviparous and miracidia occur free in the terminal part of the uterus. Fach miracidium contains a single, fully forn.ed n-dia and there is no ^porocyst generation in the life his- y. These stages were described by Linton (1914). The redise LIFE HISTORY PARORCHIS AVITUS 331 (Figs. 1 and 2) and cercariae (Fig. 3) were described by Stunkard and Shaw (1931). It should be recorded that in addition to Urosalpinx cincrcns, Thais (Pnrpnra} lapillns also has been found to serve as the intermediate host of ParorcJiis avitus. Of 108 specimens of Thais lapil- lits collected August 1st at Weepecket Island, 8 were found to be in- fected. The appearance of the parasitized gonad and digestive gland of this snail was similar to that reported for Urosalpin.v cinercns (Stunkard and Shaw, 1931). Of 610 Urosalpinx examined during the summer of 1931, 25 were found to be infested. This determination is based on the emergence of cercariae from the isolated snails. The in- festation of Thais, however, was found by crushing and examining them. Most of the cercariae encyst within 48 hours after leaving the snail. Encystment may be accelerated by mechanical stimulation such as stir- ring or shaking. Concentrated solutions of vital dyes also induce cyst formation. Before encysting, the cercarias attach themselves by means of the suckers to the bottom of the dish or to the slide as the case may be, and become comparatively quiet for a few moments. The cystogenous material is then exuded, enclosing the body in a viscous mass. This material adheres to the tail, pushing it off from the body as additional cystogenous material is extruded. The tail may remain attached to the cyst for some time or may break off as a result of its vigorous movement. Decaudation before encystment has been observed a few times but this" is not the rule. In either case, the tail soon disintegrates. The cystog- enous material is poured out very quickly and remains in a semi-liquid state for about a minute, as indicated by its yielding to the movement of the enclosed worm. In a few minutes, however, the cyst becomes hardened. The cyst wall (Figs. 4 and 5), which is flattened on the side of attachment, consists of two layers, an outer thick and brittle one, and an inner membrane which is thin and tough. The outer wall is often broken when the cysts are scraped from the bottom of the dish with a scalpel. The larva becomes shrunken immediately if the inner cyst membrane is torn and the worm comes in contact with sea water. This interesting phenomenon seems to be due to the hypertonicity of the sea water, although it may be the result of mechanical injury. If the former is the case, encystment must involve an important change in the body of the worm which before encystment can live perfectly well in sea water. The cyst is fairly transparent and the larva can be clearly observed within it. The position of the larva in the cyst is shown in Figs. 4 and 5. The acetabulum is often forced to the side of the median plane because of its large size. A fluid fills the space between the larva and the cyst wall. It is clear and in some cases contains small granules which have 332 H. \V. STUXKARD AND R. M. CABLE probably escaped from the excretory system of the worm. The main excretory tubules with their concretions are easily observed through the cyst wall. The larva moves for several hours after encystment and then becomes quiescent. Movement may be induced by warming very slightly; it ceases upon cooling. The encysted worms may remain alive for an extended period if kept in water, but they do not withstand any considerable' desiccation. They have been made to move two weeks after encystment ; it is not known how long they can survive. Specimens of M\'tilns cdiilis were placed in finger bowls containing hundreds of cercari;e in order to determine whether the larv?e would encyst within any part of the mollusk. These experiments were all negative. In one case, three cysts were found on the surface of the foot and mantle, and a few others were seen in masses of mucus in the mantle cavity. In the same specimen, large numbers of cercarise en- -;ed on tlu- outside of the shell, most of them near the incurrent siphon. Many of them struck the shell around the incurrent opening, adhered, and immediately encysted. The significance of these results will be brought out in the discussion. Since birds could be infected by feeding the encysted worms, and since the larvae encyst on any available object, it is apparent that a second intermediate or transfer host is not an essential stage in the life cycle of the parasite. It is clear that the final hosts are infected by accidentally ingesting the larvae with food on which they have encysted. The worms excyst in the intestine of the bird and develop to maturity in the cloacal region. In autopsies of experimentally infested birds, a few worms were found just behind the intestinal ceca. but none in the ceca them- selves. Since sexually mature worms were not obtained, the following description is based on specimens recovered 15 days after infection. These worms possess all of the characteristic adult structures except the distended goimducts. They are white in color and are seen only with difficulty "ii the \\-all of the cloaca. It is very difficult to remove the worms from the surface of the cloacal epithelium on account of the po\verful suction of the acctalmlnm. After removal to Ringer-Locke's solution, specimens have been kept alive for twelve hours before fixation. It is not known how much longer they would have survived. If not removed from the duaca soon after the death of the host, however, they die and disintegrate rapidly along with the tissue to which they adhere. The intestine of a bird is frequently disintegrated 6 to 8 hours after its death, which explains the negative results in the first series of experi- mei When the worms are placed in Ringer-Locke's solution, they move actively for about an hour. Tin's movement is very similar to that of the LIFE HISTORY PARORCHIS AVITUS 333 cercariae. The worms, although flat and leaf-like, tend to lie on one side, i.e., with the ventral surface in a vertical rather than a horizontal position, movement being effected by flexing and extending of the pre- acetabular region of the body. After this period of active movement during which the suckers are not used, the worms settle to the bottom of the dish, become attached by means of the suckers, and move in meas- uring worm fashion. This means of locomotion is more effective than the erratic dorso-ventral lashing of the forebody. Finally, however, the worms become quiescent on the bottom of the dish, remaining in some cases sufficiently still to be drawn with the camera lucida ; when stimu- lated by moving the dish or touching them with a needle, they again swim actively for a time. The morphology of Parorchis ai'itus was described by Linton (1914). Study of the present specimens confirms the observations of Linton and permits certain additions to the description of the species. The average measurements of thirteen 15-day worms stained and mounted are: length, 1.8mm.; width of oral sucker, 0.236mm.; width of pharynx, 0.1 mm.; and width of acetabulum 0.433 mm. The details of spination are shown in Fig. 6. There are about 68 spines in the collar. The papillae, noted in the larva?, still persist on the anterior end of the 15-day worm and are most clearly seen on a well extended but not flattened living specimen. The antero-ventral surface is the most heavily spined region of the body. The alternate rows of closely set spines extend backward in regular order to the genital ridge. Behind this region, there is a gradual thinning out on either side of and posterior to the acetabu- lum, with a few scattered spines extending to the extreme posterior end. The body is flattened dorsoventrally, especially in the postacetabular region, although with sexual maturity this portion will undoubtedly enlarge. The excretory system, which is one of the most striking features of these specimens, differs markedly from the condition in the cercarial stage described by Stunkard and Shaw. In the cercaria there is an excretory vesicle at the caudal end of the body and collecting ducts pass forward on either side to the region between the pharynx and oral sucker where they turn posteriad. The recurrent ducts divide at the level of the intestinal bifurcation into anterior and posterior branches. Each of the anterior and posterior branches bears three groups of flame cells, with three cells in each group. In specimens removed from the cloaca of the tern the ascending collecting ducts of the cercaria have developed complicated series of evaginations which tend to form a reticulum and which ramify through the body of the worm. This ramification of the ascending ducts has not been observed in encysted larvae and probably 334 H. \V. STUXKARD AND R. M. CABLE does not appear before excystmciit. A number of very small worms were recovered from the tern that was fed cysts intermittently. The youngest of these-, which had probably been excysted not more than two or three days, exhibited to a considerable degree the complex ramifica- tion of the main excretory trunks. In general, the excretory system agrees with that of /'. ticaiitltits as described by Xicoll (1907 b). The excretory vesicle i> a conspicuous, irregularly shaped, transparent sac at the posterior end of the- body, and empties through a dorsal excretory pore ( Fig. <>). The main lateral excretory trunks extend from the antero-lateral regions of the vesicle and branch in a very complicated manner. The rami extend in three general directions, laterad, mediad, and anteriad. The median branches either end blindly or anastomose with corresponding branches of the opposite trunk. The anterior branches are. for the most part, continuations of the main trunk which extend around either side of the acetabulum. anastomosing in some cases, and sending a number of sinuses into the region just above the acetabulum. The lateral branches extend laterally and somewhat ven- trally. connecting with a more or less continuous row of large sinuses which extends along either side of the body. Tiny branches extend from the sinuses to the very edges of the body. The two rows of lateral sinuses extend posteriorly and in some cases fuse, making a con- nection behind and below the excretory vesicle. Anteriorly, the anterior sinu.scs decrease in si/.e and send branches into the tissue just dorsal to the oral sucker. On each side, at the level between the oral sucker and EXPLANATION OF PLATE IMC. 1. A young rcdia \vith an immature ccrcaria and several germ balls (42 X). FlG. -'. An older redia containing only mature cercariae, one of which is seen ipiim through the birth pmv I -4J • ). FIG. .v A mature cercaria (155 X)- •4. An encysted cercaria. tup view (132X). IML. 5. An cnr\ >trd ccrcaria, side view (lo2X). IMI.. o. Ventral view of a 15-day worm ( o() X). LETTERING A Ac.'tahulum / Intestine C Cen ii ia O Ovary • .liar OS Oral Sucker rus Sac 1\11T Recurrent Excretory Tubule / 1 opha SG Shell Gland /-./' Excretorj I' SV Seminal Vesicle / / ' i . lc T Testis / I U Uterus m Hall V X'itellaria • I 'me LIFE HISTORY PARORCIIIS AVITUS 335 336 H. \V. STUXKAK1) AND R. M. CABLE pharynx, the recurrent excretory tubules extend inward and backward from the lateral sinuses, dividing at the level of the intestinal bifurcation into anterior and posterior branches, as described in the cercaria. It is seen, then, in the adult that the complex sinus system as a whole cor- responds to the main lateral trunks of the cercaria and, consequently, serves to connect the recurrent excretory tubules with the vesicle. The main tubules are ciliated and the entire system, with the exception of the recurrent tubules and their branches, contains numerous concretions. The movement of these excretory bodies in the sinuses is of assistance in tracing tin- connections of the system. The excretory fluid is forced through the spaces by movements of the worm, the direction of the flow depending on the movement. After living worms have been kept in Kinder- Locke's solution for some time, the sinuses and bladder collapse, expelling clouds of granules and considerable amounts of fluid through the excretory pore. This makes it necessary to study the excretory sys- tem in freshly removed material, it" the precise relations of the vesicle and sinuses are to be observed. The worm may live, however, for many hours after these structures have become indistinct. The preacetabular region of the body also becomes opaque long before the worm dies, mak- ing observations on the finer details of the excretory system in this region more difficult. The reproductive system (Fig. 6) of the 15-day worm is not fully developed, although in a few cases motile spermatozoa were observed in the seminal vesicle, and one case of copulation was noted. The uterus lie-ins as a simple duct which extends forward from the ootype. As development proceeds it forms lateral convolutions, a process which continues as the worm becomes more mature and large numbers of eggs are produced. DISCUSSION Since the close affinity of I'ar orchis and the echinostomes is apparent t'roni morphological comparisons, and since the cercarire of Echinosto- iiiitin sccinnliiin were found by Lebour ( 1(K)8) to encyst in bivalve mol- lusk-. she expected the same behavior on the part of cercarise of P. Several fads, however, seem to indicate that Lebour and irsfs f 1'L'Jj (-(inclusions in regard to the encystment of cercariae of /'. (iciiiitlnis (C. f>urpnrt enter the tail of Cercaria purpurcc; (6) the arrangement of the gland cells in the cercarite is different; and (7) the larvae of P. avitus do not encyst in the foot or mantle of Mytilus cdulis as has been reported for /'. acanthus. SUMMARY The life history of Parorcliis avitus has been experimentally traced. The cercari;e occur in the marine snails, Urosalpin.v cinercus and Thais (Pur pur a ) la pill us. Adults have been obtained from the cloaca of the common tern. Sterna hirundo, and the roseate tern, Sterna dougalli, ,-ifter feeding the young birds with encysted larva?. It has been shown that a specific secondary intermediate host is not '•ntial for the completion of the life history; only a means of trans- ference is necessary. Additional morphological differences between Parorchis avitus and Parorchis acanthus are described. BIBLIOGRAPHY Ion \sox. J. C.. 1920. The Life History of Echinostomum revolutum (Froelich). Univ. (•«////. /'»/>/. Zool, 19: 335. LEBOIK, M. \'.. 1908. A Contribution to thr Life History of Echinostomum sivundum (Xicoll). I\irasil.. 1: 352. l.ip.oi u. M. \'., 1<>14. Some Larval Tivmatodcs from Millport. Parosit.. 7: 1. 1.11:01 R. M. \'.. AND R. KI.MIIIKST, 1922. A Contribution towards the Life His- tory of Parorchis acanthus ( Xicoll), a Trematode in the Herring Gull. Jour. Mar. Hwl. Ass'n., Plymouth, N.S., 12: K29. LINTON, I-'., 1914. Xotes on a Viviparous. Di^tome. Proc. U. S. Nat. Mas.. 46: 551. LlNTON, I-"... 1'L'S. Xotes on Trematode Parasites of Birds. Proc. U. S. Nat. Mtts., 73: 1. NICOLL, \\'.. I'^io. Some Xe\v and Little-known Trematodes. Ann. Mag. Nat. Hist., (7) 17: 513. Xiioi.i.. \\'.. 107/'. I'.iionhis acanthus, the Type of a New Genus of Trematodes. "/. Jour. Micrns. Si-i,. 51: 345. STUM i I \V.. AND !\. M. CABLE, I'^l. A Trematode from the Cloaca of the Gull. Science, X.S.. 74: 43S. i, 11. \V.. AND < !' SB \u. 1'Ml. The Effect of Dilution of Sea Water on ill'- Vmity and Longevity of Certain Marine Cercaria?, with De- scriptions (,f Two Xew Species. Biol. Bull.. 61: 242. INDEX ^DOLPH, EDWARD F. The vapor tension relations of frogs, 112. Adrenalin, effect on blood pressure, Squalus acanthias, 17. Amoeba proteus, chemical needs, 205. Anaerobiosis, effect on eggs and sperm of sea urchin, starfish, and Nereis, 46. Ant, temperature effect on leg posture and speed of creeping, 306. Arbacia punctulata, development of half and quarter eggs, and strongly centrifuged whole eggs, 155. — , physical and chemical con- stants of egg, 141. AREY, LESLIE B. The formation and structure of the glochidial cyst, 212. Ascidians of the Bermudas, 77. "DARD, PHILIP. See Lundstrom and Bard, 1. BARNES, T. CUNLIFFE, and H. I. KOHN. The effect of temperature on the leg posture and speed of creeping in the ant, Lasius, 306. BARRON, E. S. G. Studies on cell metabolism. I. The oxygen con- sumption of Nereis eggs before and after fertilization, 42. — , - — . The effect of anaerobiosis on the eggs and sperm of sea urchin, starfish and Nereis and fertilization under anaerobic conditions, 46. BENNITT, RUDOLF, and AMANDA DICKSON MERRICK. Migration of the proxi- mal retinal pigment in the crayfish in relation to oxygen deficiency, 168. BERRILL, N. J. Ascidians of the Ber- mudas, 77. Blood, of skate, properties of, 23. - pressure, effect of adrenalin on, in Squalus acanthias, 17. — , reducing substances in, of Limulus polyphemus, 37. /^"ABLE, RAYMOND M. See Stunkard ' and Cable, 328. Centrifugation, effect on development of whole eggs of Arbacia punctulata, 155. Chemical needs of Amoeba proteus, 205. Chick embryo, fat metabolism under artificial incubation, 54. COLMAN, JOHN. A statistical test of the species concept in Littorina, 223. Conjugation, physiological effects, racial differences in, Paramecium aurelia, 258. Crayfish, oxygen deficiency and mi- gration of proximal retinal pigment, 168. Creeping, temperature effect on speed, in ant, 306. Culture and chemical needs of Amoeba proteus, 205. J)AILEY, M. E. See Fremont-Smith and Dailey, 37. DILL, D. B., H. T. EDWARDS, and M. FLORKIN. Properties of the blood of the skate (Raia ocellata), 23. Drosophila melanogaster, fertility of, effect of X-rays, 294. gCHINARACHNIUS parma, loco- motor organs, 195. EDWARDS, H. T. See Dill, Edwards, and Florkin, 23. Epibdella melleni Maccallum, life- history, 89. Erythrocyte, permeability to water, effect of narcotics, 313. — , question of permeability to hydro- gen ions, 63. — , rate of hemolysis in hypotonic solutions of non-electrolytes, 178. "pAT metabolism, chick embryo, under artificial incubation, 54. FAURE-FREMIET, E. Strombidium Cal- kinsi, a new thigmotactic species, 201. Fertility, effects of X-rays, in Drosophila melanogaster, 294. Fertilization, under anaerobic conditions, in sea urchin, starfish and Nereis, 46. FLORKIN, M. See Dill, Edwards, and Florkin, 23. 339 340 INDEX FREMONT-SMITH, F., and M. I. !>MI EY. The nature of the reducing sub- stances in the blood serum of Limulus polyphemiis and in the serum, cerebrospinal iluid and aque- ous humor <>f certain elasmobranchs, 37. Frogs, vapor tension relations, 112. pI.OCHiniAL cyst, formation and structure. J12. Golgi and plastid zones compared, 126. T_T. \II\I.KT, WILLIAM I". Studies on the chemical needs of Amoeba pniteus: a culture method, 205. HAKVKY, Hthel Browne. The develop- ment of half and quarter eggs of . \rbacia punctulata and of strongly centrifuged whole eggs, 155. HARVEY, I-;. NKUTOX. Physical and chemical constants of the egg of the sea urchin, Arbacia punctulata, 141. llemolysis, rate of, by erythrocyte, in hypotonic solutions of non-elec- trolytes, 178. Hydrogen ions, question of permeability of erythrocyte to, 63. llypophysial control of cutaneous pig- mentation in elasmobranch fish, 1. JACOBS, MERKEL H. Osmotic prop- J erties of the erythrocyte. III. The applicability of osmotic laws to the rate of hemolysis in hypotonic solutions of non-electrolytes, 178. — , - — , and A. K. PARPART. Os- motic properties of the erythrocyte. IV. Is the permeability of the erythrocyte to water decreased by narcotics? 313. — , - — , and A. K. I'AKI-ART. Is the er\ tln-o.Ate permeable to hydrogen ions? 63. I \n\. TIII.O. I.., and I.. K. Ki UN. The I it\- -hi story of Kpibdella melleni M.irr;illum l'>27, a monogenetic i rrmatodr parasitic on marine fishes, 89. I ' ( >HN, II. I. See Barnes and Kohn, Ki UN, I.. R. See Jalin and Knhn. 89. T A^ll S, tempera! nre effect on leg po»l lire Mild speed of creeping, 306. I.imiilus |)olyphcmus, blood of, reducing sul in, 37. Littorina, species concept, statistical test, 223. Locomotion, temperature effect on, in ant, 306. — , organs of, in Echinarachnius par ma, 195. LUNDSTROM, H. M., and P. BARD. Hy- pophysial control of cutaneous pigmentation in an elasmobranch fish, 1. LUTZ, B. R. See Wyman and Lutz, 17. — , - — , and L. C. WYMAN. Reflex cardiac inhibition of branchio- vascular origin in the elasmobranch Squalus acanthias, 10. LYNCH, C. J. See Sonneborn and Lynch, 258. ly/rERRK'K, AMANDA DICKSON. Bennitt and Merrick, 168. Metabolism, cell, of Nereis eggs before and after fertilization, 42. MOOKK. \V. < .. The effects of X-rays on fertility in Drosophila melanogaster treated at different stages in de- velopment, 294. Mussels, fresh-water, glochidial cysts, formation and structure, 212. •MTARCOTICS, and permeability of erythrocyte to water, 313. Nereis eggs, oxygen consumption before and after fertilization, 42. — , fertilization under anaerobic con- ditions, -I'-. QSMOSIS, effect of narcotics on permeability of erythrocyte to water, 313. — , permeability of erythrocyte to hydrogen ions, 63. — , rate of hemolysis of erythrocyte in solutions of non-electrolytes, 178. Oxygen consumption of Nereis eggs be- fore and after fertilization, 42. - deficiency, and migration of proxi- mal retinal pigment in crayfish, 168. pARAMECIl'M aurelia, inherited va- riation during vegetative reproduc- tion, 244. — , racial differences in early physiological effects of conjugation in, 258. Parasites, formation and structure of glochidial cyst, 212. INDEX 341 — , Epibdella melleni Maccallum 1927, life-history, 89. PARKER, GEORGE H., and MARGARET VAN ALSTYNE. Locomotor organs of Echinarachnius parma, 195. Parorchis avitus (Linton), life history, 328. PARPART, A. K. See Jacobs and Par- part, 63. Permeability, erythrocyte to hydrogen ions, 63. — , erythrocyte to water, effect of narcotics on, 313. Physical and chemical constants, egg of Arbacia punctulata, 141. Pigmentation, cutaneous, in elasmo- branch fish, hypophysial control of, 1. Plastid and Golgi zones compared, 126. T5ACIAL differences in early physi- ological effects of conjugation in Paramecium aurelia, 258. RAFFEL, DANIEL. Inherited variation arising during vegetative repro- duction in Paramecium aurelia, 244. Raia ocellata, blood, properties of, 23. Reducing substances, of blood, of Limulus polyphemus, 37. Reflex cardiac inhibition of branchio- vascular origin in Squalus acanthias, 10. Retina, oxygen deficiency and migration of proximal pigment, in crayfish, 168. ROMANOFF, ALEXIS L. Fat metabolism of the chick embryo under standard conditions of artificial incubation, 54. OEA urchin, effect of centrifuging on development of whole eggs, develop- ment of half and quarter eggs, 155. — , egg of, physical and chemical constants, 141. — , fertilization under anaerobic conditions, 46. Skate, blood of, properties of, 23. SONNEBORN, T. M., and C. J. LYNCH. Racial differences in the early physiological effects of conjugation in Paramecium aurelia, 258. Species concept in Littorina, statistical test, 223. Squalus acanthias, blood pressure, effect of adrenalin on, 17. — , reflex cardiac inhibition of branchio-vascular origin, 10. Starfish, fertilization under anaerobic conditions, 46. Strombidium Calkinsi, 201. STUNKARD, H. W., and R. M. CABLE. The life history of Parorchis avitus (Linton), a trematode from the cloaca of the gull, 328. 'pEMPERATURE, effect on leg pos- ture and speed of creeping in ants, 306. Trematode, Epibdella melleni Maccallum 1927, life-history, 89. ALSTYNE, M. See Parker and Van Alstyne, 195. Variation, inherited, rise of, during vege- tative reproduction in Paramecium aurelia, 244. , T. ELLIOT. A comparison of the plastid with the Golgi zone, 126. WYMAN, L. C. See Lutz and YVyman, 10. — , - — , and B. R. LUTZ. The effect of adrenalin on the blood pressure of the elasmobranch, Squalus acan- thias, 17. , effect on fertility of Dro- sophila melanogaster, 294. Volume LXII Number 1 THE BIOLOGICAL BULLETIN PUBLISHED BY THE MARINE BIOLOGICAL LABORATORY Editorial Board GARY N. CALKINS, E. G. CONKLIN, Princeton University E. N. HARVEY, Princeton University SELIG HECHT, Columbia University LEIGH HOADLEY, Harvard University M. H. JACOBS, University of Pennsylvania H. S. JENNINGS, Johns Hopkins University E. E. JUST, Howard University Columbia University FRANK R. LlLLEE, University of Chicago CARL R. MOORE, University of Chicago GEORGE T. MOORE, Missouri Botanical Garden T. H. MORGAN, California Institute of Technology G. H. PARKER, Harvard University W. M. WHEELER, Harvard University EDMUND B. WILSON, Columbia University ALFRED C. REDFIELD, Harvard University Managing Editor <(U~. FEBRUARY, 1932 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, Marine Biological Laboratory, Woods Hole, Mass., between May 1 and November 1 and to the Institute of Biology, Divinity Avenue, Cambridge, Mass., during the remainder of the year. I .nterectober 10, 1902, at Lancaster, Pa., as second-class matter under Act of Congress of July 16, 1894. CONTENTS Page LUNDSTROM, H. M., AND P. BARD Hypophysial Control of Cutaneous Pigmentation in an Elas- mobranch Fish 1 LUTZ, B. R., AND L. C. WYMAN Reflex Cardiac Inhibition of Branchio-vascular Origin in the Elasmobranch Squalus acanthias 10 WYMAN, L. C., AND B. R. LUTZ The Effect of Adrenalin on the Blood Pressure of the Elasmo- branch, Squalus acanthias 17 DHL, D. B., H. T. EDWARDS, AND M. FLORKIN Properties of the Blood of the Skate (Raia oscillata) 23 FREMONT-SMITH, F., AND M. E. DAILEY The Nature of the Reducing Substances in the Blood Serum of Limulus polyphemus and in the Serum, Cerebrospinal Fluid and Aqueous Humor of Certain Elasmobranchs 37 BARRON, E. S. G. Studies on Cell Metabolism. I. The oxygen consumption of Nereis eggs before and after fertilization 42 BARRON, E. S. G. The Effect of Anaerobiosis on the Eggs and Sperm of Sea Urchin, Starfish and Nereis and Fertilization under Anaerobic Conditions 46 ROMANOFF, ALEXIS L. Fat Metabolism of the Chick Embryo under Standard Con- ditions of Artificial Incubation 54 JACOBS, M. H., AND A. K. PARPART Is the Erythrocyte Permeable to Hydrogen Ions? 63 BERRILL, N. J. Ascidians of the Bermudas 77 JAHN, THEO. L., AND L. R. KUHN The Life-history of Epibdella melleni Maccallum 1927, a Monogenetic Trematode Parasitic on Marine Fishes 89 ADOLPH, EDWARD F. The Vapor Tension Relations of Frogs 112 WEIER, T. ELLIOT A Comparison of the Plastid with the Golgi Zone 126 Volume LXII Number 2 THE BIOLOGICAL BULLETIN PUBLISHED BY THE MARINE BIOLOGICAL LABORATORY Editorial Board GARY N. CALKINS, E. G. CONKLIN, Princeton University E. N. HARVEY, Princeton University SELIG HECHT, Columbia University LEIGH HOADLEY, Harvard University M. H. JACOBS, University of Pennsylvania H. S. JENNINGS, Johns Hopkins University E. E. JUST, Howard University Columbia University FRANK R. LILLIE, University of Chicago CARL R. MOORE, University of Chicago GEORGE T. MOORE, Missouri Botanical Garden T. H. MORGAN, California Institute of Technology G. H. PARKER, Harvard University W. M. WHEELER, Harvard University EDMUND B. WILSON, Columbia University ALFRED C. REDFIELD, Harvard University Managing Editor APRIL, 1932 Printed and Issued by LANCASTER PRESS, Inc. PRINCE 8C LEMON STS. LANCASTER, PA. THE BIOLOGICAL BULLETIN is issued six times a year. Single numbers, SI. 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: \Vheldon & 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, Marine Biological Laboratory, Woods Hole, Mass., between May 1 and November 1 and to the Institute of Biology, Divinity Avenue, Cambridge, Mass., during the remainder of the year. Entered October 10, 1902, at Lancaster, Pa., as second-class matter under Act of Congress of July 16, 1894. LANCASTER PRESS, INC. LANCASTER, PA. CONTENTS Page HARVEY, E. NEWTON Physical and Chemical Constants of the Egg of the Sea Urchin, Arbacia punctulata 141 HARVEY, ETHEL BROWNE The Development of Half and Quarter Eggs of Arbacia punctulata and of Strongly Centrifuged Whole Eggs 155 BENNITT, RUDOLF, AND AMANDA DICKSON MERRICK Migration of the Proximal Retinal Pigment in the Crayfish in Relation to Oxygen Deficiency 168 JACOBS, MERKEL H. Osmotic Properties of the Erythrocyte. III. The applicability of osmotic laws to the rate of hemolysis in hypotonic solutions of non-electrolytes 178 PARKER, GEORGE H., AND MARGARET VAN ALSTYNE Locomotor Organs of Echinarachnius parma 195 FAURE-FREMIET, E. Strombidium Calkinsi, a New Thigmotactic Species 201 HAHNERT, WILLIAM F. Studies on the Chemical Needs of Amoeba proteus: a Culture Method 205 AREY, LESLIE B. The Formation and Structure of the Glochidial Cyst 212 Volume LXII Number 3 THE BIOLOGICAL BULLETIN PUBLISHED BY THE MARINE BIOLOGICAL LABORATORY Editorial Board GARY N. CALKINS, Columbia University E. G. CONKLIN, Princeton University FRANK R. LlLLIE, University of Chicago E. N. HARVEY, Princeton University CARL R. MOORE, University of Chicago SELIG HECHT, Columbia University GEORGE T. MOORE, Missouri Botanical Garden LEIGH HOADLEY, Harvard 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, 1932 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, \V.C. 2. Communications relative to manuscripts should be sent to the Managing Editor, Marine Biological Laboratory, Woods Hole, Mass., between May 1 and November 1 and to the Institute of Biology, Divinity Avenue, Cambridge, Mass., during the remainder of the year. Knterecl October 10, 1902, at Lancaster, Pa., as second-class matter under Act of Congress of July 16, 1894. • CONTENTS Page COLMAN, JOHN A Statistical Test of the Species Concept in Littorina 223 RAFFEL, DANIEL Inherited Variation Arising during Vegetative Reproduction in Paramecium aurelia 244 SONNEBORN, T. M., AND C. J. LYNCH Racial Differences in the Early Physiological Effects of Con- jugation in Paramecium aurelia 258 MOORE, W. G. The Effects of X-rays on Fertility in Drosophila melanogaster Treated at Different Stages in Development 294 BARNES, T. CUNLIFFE, AND HENRY I. KOHN The Effect of Temperature on the Leg Posture and Speed of Creeping in the Ant, Lasius 306 JACOBS, M. H., AND ARTHUR K. PARPART Osmotic Properties of the Erythrocyte. IV. Is the per- meability of the erythrocyte to water decreased by nar- cotics? 313 STUNKARD, HORACE W., AND RAYMOND M. CABLE The Life History of Parorchis avitus (Linton), a Trematode from the Cloaca of the Gull . . 328 INDEX. 339 M.Bi; *"<>! LIBRARY 17IE 3