THE BIOLOGICAL BULLETIN PUBLISHED BY THE MARINE BIOLOGICAL LABORATORY Editorial Board E. G. CONKLIN, Princeton University E. N. HARVEY, Princeton University SELIG HECHT, Columbia University LEIGH HOADLEY, Harvard University L. IRVING, Swarthmore College M. H. JACOBS, University of Pennsylvania H. S. JENNINGS, Johns Hopkins University FRANK R. LILLIE, University of Chicago CARL R. MOORE, University of Chicago GEORGE T. MOORE, Missouri Botanical Garden G. H. PARKER, Harvard University A. C. REDFIELD, Harvard University F. SCHRADER, Columbia University DOUGLAS WHITAKER, Stanford University H. B. STEINBACH, Washington University Managing Editor VOLUME 90 FEBRUARY TO JUNE, 1946 Printed and Issued by LANCASTER PRESS, Inc. PRINCE 8C LEMON STS. LANCASTER, PA. 11 THE BIOLOGICAL BULLETIN is issued six times a year at the Lancaster Press, Inc., Prince and Lemon Streets, Lancaster, Penn- sylvania. Subscriptions and similar matter should be addressed to The Biological Bulletin, Marine Biological Laboratory, Woods Hole, Massachusetts. Agent for Great Britain: Wheldon and Wesley, Limited, 2, 3 and 4 Arthur Street, New Oxford Street, London, W. C. 2. Single numbers, $1.75. Subscription per volume (three issues), $4.50. Communications relative to manuscripts should be sent to the Managing Editor, Marine Biological Laboratory, Woods Hole, Massachusetts, between July 1 and September 1, and to the De- partment of Zoology, Washington University, St. Louis, Missouri, during the remainder of the year. Entered as second-class matter May 17, 1930, at the post office at Lancaster, Pa., under the Act of August 24, 1912. LANCASTER PRESS, INC., LANCASTER, PA. CONTENTS No. 1. FEBRUARY, 1946 PAGE KOZLOFF, EUGENE N. Studies on ciliates of the family Ancistrocomidae Chatton and Lwoff (order Holotricha, suborder Thigmotricha). I. Hypocomina tegularum sp. nov. and Crebricoma gen. nov 1 SHORT, ROBERT B. Observations on the giant amoeba, Amoeba carolinensis (Wilson, 1900) . . 8 SCHRADER, FRANZ The elimination of chromosomes in the meiotic divisions of Brachyste- thus rubromaculatus Dallas 19 HALL, C. E., M. A. JAKUS, AND F. O. SCHMITT An investigation of cross striations and myosin filaments in muscle 32 SMITH, F. G. WALTON Effect of water currents upon the attachment and growth of barnacles ... 51 D'ANGELO, ETHEL CLANCY Micrurgical studies on Chironomus salivary gland chromosomes 71 SERIAL PUBLICATIONS ADDED TO THE MARINE BIOLOGICAL LABORATORY AND THE WOODS HOLE OCEANOGRAPHIC INSTITUTE LIBRARY SINCE FEBRUARY, 1943. 88 No. 2. APRIL, 1946 COSTELLO, DONALD PAUL The swimming of Leptosynapta 93 RICHARDS, A. GLENN AND LAURENCE K. CUTKOMP Correlation between the possession of a chitinous cuticle and sensitivity to DDT 97 CARLSON, J. GORDON Protoplasmic viscosity changes in different regions of the grasshopper neuroblast during mitosis 109 MILLER, MILTON A. Toxic effects of copper on attachment and growth of Bugula neritina .... 122 WALFORD, LIONEL A. New graphic method of describing the growth of animals 141 BODENSTEIN, DIETRICH Investigation on the locus of action of DDT in flics (Drosophila) 148 HOPKINS, DWIGHT L. The contractile vacuole and the adjustment to changing concentration in fresh water amoebae. 158 in 60067 iv CONTENTS \o. 3. JUNE, 1946 HAYASHI, TERU Dilution medium and survival of the spermatozoa of Arbacia punctu- lata. II. Effect of the medium on respiration 177 1 1 KXLEY, CATHERIN E The effects of lithium chloride on the fertilized eggs of Nereis limbata .... 188 KOZLOFF, EUGENE N. Studies on ciliates of the family Ancistrocomidae Chatton and Lwoff (order Holotricha, suborder Thigmotricha). II. Hypocomides mytili Chatton and Lwoff, Hypocomides botulac sp. nov., Hypocomides parva sp. nov., Hypocomides kelliae sp. nov., and Insignicoma vcnusta gen. nov., sp. nov 200 TYLER, ALBERT Natural heteroagglutinins in the body-fluids and seminal fluids of various invertebrates 213 KRUGELIS, EDITH JUDITH Distribution and properties of intracellular alkaline phosphatases 220 WILLIAMS, CARROLL M. Physiology of insect diapause: The role of the brain in the production and termination of pupal dormancy in the giant silkworm, Platysamia cecropia 234 LOOSANOFF, VICTOR L. AND CHARLES A. NOMEJKO Feeding of oysters in relation to tidal stages and to periods of light and darkness 244 SCHRADER, FRANZ Autosomal elimination and preferential segregation in the harlequin lobe of certain Discocephalini (Hemiptera) 265 RUGH, ROBERTS The effect of the adult anterior pituitary hormone on the tadpoles and the immature male frogs of the bullfrog, Rana catesbiana 291 THOMAS HUNT MORGAN SEPTEMBER 25, 1 866— DECEMBER 4, 1945 Member of the Editorial Board 1902-1945 Vol. 90, No. 1 February, 1946 THE BIOLOGICAL BULLETIN PUBLISHED BY THE MARINE BIOLOGICAL LABORATORY STUDIES ON CILIATES OF THE FAMILY ANCISTROCOMIDAE CHATTON AND LWOFF (ORDER HOLOTRICHA, SUBORDER THIGMOTRICHA) I. HYPOCOMINA TEGULARUM SP. NOV. AND CREBRICOMA GEN. NOV. EUGENE N. KOZLOFF Lewis and Clnrk Collct/c, Portland, Oreuon INTRODUCTION The genus Hypocoiuina was proposed by Chatton and Lwoff ( 1924) to include a single species, Hvpoconut putcUaniin Lichtenstein (1921 ), parasitic on the pallial branchiae of Patella cacntlca L. Subsequently, Raabe (1934) described as Hypo- coiuina cariuata a ciliate from Mytilits cdulis L. which Jarocki (1935) pointed out obviously does not belong to the genus Hypocomina. Raabe used the name H\p<>- couiiiui carinata again in 1938, however, and no literature has come to my attention in which the form in question is assigned to another genus. In the present paper I will describe as Hypocoinina tcgitlaruin sp. nov. an ancistrocomid ciliate from the ctenidium of Tcgnla bntnnca ( Philippi ) ; and on the basis of my studies on the morphology of a ciliate which I presume to be identical with "Hypocomina" cannata, I will establish the position of Raabe's species in a new genus, Crcbriconut. HYPOCOMINA TEGULARUM SP. NOV. (Plate I, Figs. 4, 5) The body is elongated and compressed dorso-ventrally. Fifteen living individ- uals taken at random ranged in length from 26 p. to 36 p., in width from 12 p. to 17 p., and in thickness from 9 p to 12 p, averaging about 31 p. by 15 p by 1 1 p. As seen in dorsal view the curvature of the left margin of the ciliate is a little more pronounced than that of the right margin, although this asymmetry is not conspicuous and in most fixed specimens is barely apparent. The body is widest near the middle and is rounded posteriorly. The anterior end is attenuated, bent ventrally, and truncate at the tip. The reduced ciliary system, to be described presently, is disposed in a shallow, relatively flat depression occupying the anterior half of the ventral surface ; the dorsal surface and that part of the ventral surface posterior to the thigmotactic field are convex. EUGENE X. KOZLOFF A contractile suctorial tentacle enables the ciliate to attach itself to the epithelial cells o! tlie ctenidiuin of the host and to suck out their contents. I have not yet suc- ceeded in determining the exact nature of the tentacle in the ancistroconiid ciliates which 1 have examined and the accounts of other authors regarding its structure are likewise unsatisfactory. In Hypocomina tct / / 'ypacoinclla nuicoiiuic and "Hypocomina" cannata to hyper- distension of the canal due to unequal infiltration by the fixing reagent or as a result of its action upon moribund, partly plasmoly/.ed individuals. It is interesting to note in this connection, however, that some ancistroconiid ciliates rarely show this phenomenon upon fixation and the canal can be traced to within a short distance of the posterior end of the body. The delicate thigmotactic cilia of Hypocomina tc(/itliirnin are about 6-7 // in length and are disposed in nine longitudinal rows. All the rows originate about 3 p. or 4/i posterior to the base of the suctorial tentacle and are about one-half the length of the body ( Plate I. fig. 4). The first five rows from the right side of the ciliate are appreciably longer and sometimes appear to lie closer together than the other lour rows. Chatton and Lwoff (1924) reported that in Hypocomina patellarum a ridge-like eminence ( carcne ) divides the ciliarv area into two unequal fields. These authors did not, however, state how many of the ten ciliary rows described by Liechtenstein (1921 ) for //. patellarum are in each complex and did not give any facts concerning the relative lengths of the rows comprising the two fields. In Hypocomina tegularum I have not discerned an unmistakable eminence separating the five rows on the right !rom the tour rows on the lelt. although in >ome living and fixed individuals the distance between the distal portions of the fifth and sixth rows is somewhat greater than the distance between the other rows. Ex I'l.. \.V.\TIO.\ (W I'l.ATK 1 FIGURE 1. ( rebricoma curiiuitu (Raabe) ronili. n<>v. Ventral a^u-et. Schaudiun's fixative- iron heniatoxyliu. I )ra\vn with aid of camera lucida. 1130. l'!i,rim ri»lit side, troin life. FIGURE -t. Hypocomina tegularum sp. nov. \entral aspect. Schaudinn's fixative-iron hematoxylin. Drawn \\ith aid of camera lucida. < 18/0. l-K,ri. nov. Diagnosis: Length 26^-36^, average about 31/t; width 12^-17/x, average about 15 p.; thickness 9//-12//, average about 11 p.. The anterior end is attenu- ated, bent ventrally, and provided with a contractile suctorial tentacle continuous with an internal tubular canal. The ciliary rows are nine in number and are dis- posed in a shallow depression occupying the anterior half of the ventral surface. The first five rows from the right are slightly longer than the other four rows. The macronucleus is spherical and is situated in the posterior half of the body. The micronucleus lies anterior to the middle of the body. Ectoparasitic on the ctenidium of Tcyula bntnnca (Philippi) (Carmel Bay, California). Syntypes are in the col- lection of the author. CREBRKOMA <;KN. NOV. CREBRICOMA CARINATA (RAABE) COMB. NOV. ( Figure 1. Plate 1, Figs. 1-3) I have distinguished two species of ancistrocomid ciliates parasitic on the epi- thelium of the palps and ctenidial filaments of Myiilns cdulis L. from various locali- ties in San Francisco Hay. One of these is Hypocomides inytili Chatton and Lwol'f ; the other I presume to be identical with "Hypocomina" carinata Kaabe, lor which I propose a new genus, Crcbricoiiia. M v observations on the morphology JI am indebted to I >r. I). T. Mad )ousjal for hi* kindness in arran.uinsi my trip to (..'armcl which made possible tlie collection of the original material of the ciliate described herein. CILIATES OF THE FAMILY ANCISTROCOMIDAE. I 3 of this species permit me to augment the original description given by Raabe. It seems advisable, therefore, to present herein a revised description of Crebricoma carinata. The body is elongated and somewhat compressed dorso-ventrally. Fifteen liv- ing individuals taken at random ranged in length from 58 p. to 71 p., in width from 27 p. to 39 /A, and in thickness from 22 p. to 31 p., averaging about 64 p. by 31 p. by 25 p. The anterior end is narrowed, bent ventrally, and in dorsal view is seen to end in an oblique truncation, the right side of which projects a little more than the left side (Plate I, figs. 1, 2). The body is widest near the middle and is rounded posteriorly. The ciliary system, to be described presently, is disposed for the most part on the shallow concavity occupying the anterior two-thirds of the ventral sur- face ; the dorsal surface and that part of the ventral surface posterior to the thigmo- tactic area are convex. c FIGURE 1. Crebricoma carinata (Raabe) comb. nov. Distribution of ciliary rows. Semi- diagrammatic representations based on camera lucida drawings of specimens fixed in Schaudinn's fluid and impregnated with activated silver albumose (protargol). A. Ventral aspect, B. Dorsal aspect, C. View of anterior end. The suctorial tentacle is situated on the right side of the anterior truncation. I am at present unable to make any conclusive statement with regard to the structure and contractile properties of the tentacle in this species. The internal tubular canal continuous with the tentacle cannot be satisfactorily distinguished in living material, but can be demonstrated in some fixed specimens which have been stained with iron hematoxylin. It appears to pass at first dorsally and then ventrally and obliquely to the right. I have not succeeded in tracing it posteriorly for more than a short dis- tance. A broad lighter area behind the anterior part of the canal is frequently ob- served in fixed individuals. As described by Raabe, the ciliary system of Crebricoma carinata consists of about twenty closely-set rows bordered on the right by two longer and more widely- spaced ro\vs and on the left by three such rows. According to my observations, however, the more widely-spaced rows on the left side do not represent a complex separate from the closely-set rows. The number of rows in the thigmotactic field 6 EUGENE N. KOZLOFF of C. carinata is very difficult to determine, since some of the rows on the left are lateral in position and several of them originate on the dorsal surface. In one in- dividual impregnated with activated silver albumose I counted thirty-lour rows and in another individual, thirty-six rows. In all specimens which I have examined carefully the number of ciliary rows exceeded thirty-two. The two widely-spaced rows on the right side of the thigmotactic field originate on the ventral surface near the base of the suctorial tentacle. The outer row is the longer and is about two-thirds the length of the body (Fig. 1A). All the rows of the second complex, with the exception of three or four rows on the extreme left which are about as widely-spaced as the two long rows on the right side, are very closely-set and are one-half to two-thirds the length of the body, becoming progres- sively longer toward the left. Several of the rows on the left side of the thigmo- tactic field originate on the left lateral margin or on the dorsal surface (Fig. IB; Plate I, fig. 2). The arrangement of the ciliary rows at the anterior end thus forms an incomplete suture (Fig. 1C) reminiscent of the anterior field of other holotrich- ous ciliates. The cilia of Crcbriconia carinata are about 10-11 /JL in length. When the organ- ism is attached to the epithelium of the palps or ctenidial filaments of the host the cilia usually exhibit only a sluggish movement. Occasionally one or two small groups of cilia near the anterior end beat energetically. When dissociated from the host C. carinata swims actively and its cilia beat metachronously. Raabe proposed the specific name carinata in allusion to a keel-like prominence on the dorsal surface. I have seen such prominences on seriously shrunken fixed specimens of Crcbriconia carinata and on some living individuals which were obvi- ously undergoing plasmolysis. I have not, however, been able to distinguish them on ciliates freshly removed from the host. Raabe also described a sunken space between the keel-like prominence and the outermost ciliary row on the right side, which he said "vermutlich dem von Chatton und Lwoff bei anderen Hypocomiden beschriebenen 'vestige de frange adorale' entspricht." I do not believe such a de- pression exists in normal individuals, although I have observed groove-like and slit-like depressions form on the lateral and dorsal surfaces of moribund specimens of Crcbricoma carinata, as well as Hypocomides mytili. The cytoplasm is colorless and contains a few small refractile granules, which presumably are lipoid droplets, and food inclusions. I have not discerned any large food vacuoles in this species. The contractile vacuole (Plate I, fig. 2, cv) opens to the exterior on the ventral surface near the middle of the body. There appears to be no permanent opening in the pellicle. The sausage-shaped or ovoid macronucleus (Plate I, fig. 1, ma) is situated in the posterior half of the body. In preparations stained with iron hematoxylin or the Feulgen nuclear reaction the chromatin is aggregated into a dense reticuium en- closing vacuole-like clear spaces of varying size. Viewed dorsally, the longitudinal axis of the macronucleus is ordinarily placed obliquely to the longitudinal axis of the body. In most specimens the anterior end of the macronucleus is directed dorsally, while the posterior end is directed ventrally. In ten individuals fixed in Schaudinn's fluid and stained by the Feulgen nuclear reaction the macronucleus ranged in length from 13 /A to 24.3 /j. and in width from 5.6 /j. to 11.7 /J-. The spherical micronucleus (Plate I, iig. 1, mi) is readily detected in the living ciliate. It is commonly situated near the dorsal surface close to the anterior end of CILIATES OF THE FAMILY ANCISTROCOMIDAE. I / the macronucleus. Following fixation the micronucleus stains lightly with iron hematoxylin and the Feulgen reaction. The chromatin appears to be homogeneous. In ten individuals fixed in Schaudinn's fluid and stained by the Feulgen reaction the diameter of the micronucleus ranged from 2.7 p. to 3.6 /A. Crebricoma gen. nov. Diagnosis : The body is elongated, somewhat flattened dorso-ventrally, and nar- rowed anteriorly. The anterior end is provided with a contractile suctorial tentacle continuous with an internal tubular canal. The ciliary rows are numerous and ex- cept for some which originate on the dorsal surface or left lateral margin are dis- posed on the ventral surface. Two long, widely-spaced rows on the right side of the ciliary system appear to form a complex separate from the remaining rows, which are for the most part closely-set. The arrangement of the ciliary rows at the anterior end of the body forms an incomplete suture. Genotype : Crebricoma carinata (Raabe) comb. nov. (= Hypocomina carinata Raabe). Crebricoma carinata (Raabe) comb. nov. Diagnosis: Length 5S//.-71/A, average about 64 /A; width 27 p-39 p, average about 31 jtt; thickness 22/^-31 p., average about 25 /A. Two widely-spaced rows of cilia on the right side, the outer of which is about two-thirds the length of the body, and a series of more than thirty rows, which with the exception of three or four on the left are closely-set, comprise the ciliary system. The closely-set rows are one- half to two-thirds the length of the body, becoming progressively longer toward the left. The macronucleus is sausage-shaped or ovoid. The micronucleus is spherical. Ectoparasitic on the palps and branchial filaments of Mytilus edulis L. LITERATURE CITED CHATTON, E., AND A. LWOFF, 1924. Sur 1'evolution des infusoires des lamellibranches ; mor- phologic comparee des hypocomides. Les nouveaux genres Hypocomina et Hypo- comella. C. R. Acad. Sci. Paris, 178 : 1928. CHATTON, E., AND A. LWOFF, 1939. Sur la systematique de la tribu des thigmotriches rhyn- cho'ides. Les deux families des Hypocomidae Biitschli et des Ancistrocomidae n. fam. Les deux genres nouveaux, Heterocoma et Parhypocoma. C. R. Acad. Sci. Paris, 209 : 429. JAROCKI, J., 1935. Studies on ciliates from fresh-water molluscs. I. General remarks on pro- tozoan parasites of Pulmonata. Transfer experiments with species of Heterocineta and Chaetogaster limnaei, their additional host. Some new hypocomid ciliates. Bull. int. Acad. Sci. Cracovie, Cl Sci. math, not., £(//), 1935: 201. LICHTENSTEIN, J., 1921. Hypocoma patellarum n. sp., acinetien parasite de Patella caerulea L. C. R. Soc. Biol, 85 : 796. RAABE, Z., 1934. Uber einige an den Kiemen von Mytilus edulis L. und Macoma balthica (L.) parasitierende Ciliaten-Arten. Ann. Mus. zool. polon., 10: 289. RAABE, Z., 1938. Weitere Untersuchungen an parasitischen Ciliaten aus dem polnischen Teil der Ostsee. II. Ciliata Thigmotricha aus den Familien : Hypocomidae Biitschli und Sphaenophryidae Ch. & L\v. Ann. Mus. zool. polon., 13 : 41. OBSERVATIONS ON THE GIANT AMOEBA, AMOEBA CAROLINENSIS (WILSON, 1900) x ROBERT B. SHORT Miller School of Biology, University of Virginia INTRODUCTION During the past few years there has been considerable controversy over the name of the giant fresh-water amoeba which Wilson (1900) described and named Pelomyxa carolinensis. He placed this amoeba in the genus Pelomyxa evidently because it has many nuclei. Mast (1938) and Rice (1945) agree with Wilson in regard to the name of this form. The latter author says that the principal characteristic of the genus Pclomyxa is its many nuclei. Greeff (1874), who established the genus, states that besides vacuoles there are in Pelomyxa palustris three kinds of characteristic structures : (1) Nuclei, (2) Hyaline and homogeneous bodies of spherical, ellipsoidal, or ir- regular shape, and (3) Stabchen, little rods (later considered to be symbiotic bac- teria). He does say that the large number of nuclei forms a principal characteris- tic of the genus Pelomy.va. However, since GreefFs description of P. palustris, pelomyxae have been described with one or few nuclei (P. Belevskii Penard, 1888; P. binucleata (Gruber, 1885) Penard, 1902; P. paradoxa Penard, 1902; P. lentis- shna Schaeffer, 1918; P. scJiicdti Schaeffer, 1918), and amoeboid forms other than pelomyxae have been described with more than one nucleus. Amoeba proteus has been observed in this laboratory with as many as eight nuclei.2 Also the hyaline spherical bodies are not found exclusively in Pelomyxa. In view of these facts, Penard (1902) states that the chief criterion for the genus Pelomyxa is the pres- ence of symbiotic bacteria in the cytoplasm. He says, "S'il me fallait done carac- teriser le genre Pelomyxa, je le ferais a peu pres en ces terms: 'Amibes a mouve- ments lents, toujours pourvues de bacteries symbiotiques.' ' He then points out that Wilson's rhizopod has in common with the Pclomyxa only the Glanskorpcr and that this form has nothing to distinguish it from the genus Amoeba. Schaeffer (1918) in describing Pclomyxa lentissima states, "Other inclusions in the ectoplasm are the bacterial rods, distinctive of the genus." Later in his ac- count of P. schiedti he says, "The bacterial rods, the presence of which characterizes the genus Pelomyxa, are found in considerable numbers in schiedti." Bourne (1891) in speaking of P. viridis describes the organism as, "densely packed with bacteria." Thus it is seen that the chief characteristic of the genus Pclomyxa is the pres- ence of symbiotic bacteria in the cytoplasm. Since the giant amoeba does not possess these bacteria, it is not a Pelomyxa. 1 The author wishes to express his grateful appreciation to Dr. B. D. Reynolds for his sug- gestions and helpful advice in regard to this work and to thank Dr. C. M. Gilbert for help in taking the photomicrographs. 2 Personal communication from Dr. B. D. Reynolds. 8 OBSERVATIONS ON THE GIANT AMOEBA Schaeffer (1926, 1938a) maintains that Wilson's amoeba represents Roesel's (1755) "der kleine Proteus," and should therefore be called Chaos chaos Linnaeus. But Roesel's description is entirely inadequate to establish a species or even a genus. He undoubtedly had an amoeboid form, but it is impossible to find out the exact structure of his "Proteus." He mentions granules (Korncni) but says nothing of vacuoles, which in the giant amoeba are much larger than the crystals (granules). Schaeffer seems to base his conclusion mainly on the size and shape of the "Proteus." He contends that Roesel was a reliable investigator and that his fig- ures may therefore be credited with general accuracy. As Schaeffer (1926) points out, "Roesel states the natural size of his amoeba in the rounded (spherical?) form to have been the same as figure A, which measures about 1660^, in diameter." Yet Schaeffer (1938a) says, "Chaos chaos has a distinctive size range which fluctuates around 500 /*, diameter." Thus, Schaeffer himself gives the diameter of the giant amoeba as about V. that of Roesel's "Proteus." There are also striking differences between Roesel's figures and description of a binary fission in his "Proteus" and Schaeffer's (1938b) photographs and account of the "3-daughter division of the giant amoeba." Mast and Johnson (1931) review the data presented by Roesel and express the opinion that the latter possibly was dealing with a myxomycete. Rice (1945) states, "It is impossible to ascertain the exact structure of Roesel's 'der kleine Proteus.' ' Thus since both the generic and specific names. Chaos chaos, are based on Roesel's inadequate description, they are not valid. Hegner and Taliaferro (1924) in referring to Wilson's large amoeba use the name Amoeba carolinensis for it, though they give no reason for doing so. The present investigation adds evidence to support the use of this name. Therefore, Amoeba carolincnsis (Wilson. 1900) Hegner and Taliaferro, 1924 is considered to be the correct name of the organism used in these experiments. CYTOLOGY AND PHYSIOLOGY MATERIALS AND METHODS The original stock culture of Amoeba carolincnsis was procured from the Gen- eral Biological Supply House, Chicago, Illinois, in November 1944. About every two weeks subcultures were made in the following manner: Spring water containing five wheat grains per 100 cc. of water was boiled for a few minutes and left un- covered for several days. One or two of the grains of wheat were put into a butter dish and about 90 cc. of the boiled water was poured in, giving a depth of % inches. The amoebae along with food organisms, which consisted chiefly of rotifers, Chiloinonas f>aniiucciuin, Colpidiiim colpoda, and Paraincciitin caiidatiim. were then added. In the best cultures a water mold grew on the wheat grains and many amoebae often were found among the hyphae of the mold. These stock cul- tures had been maintained for a little more than four months when the experiments began. In addition to the wheat culture medium ( W. C. M.) a hay medium ( H. C. M. ) was made as follows: 10 grams of chopped timothy hay were added to 1000 cc. of spring water and boiled for about 15 minutes. The solution was filtered to remove the solid particles and then put into test tubes with cotton stoppers and antoclaved on two successive days at 10 pounds pressure. From these test tubes the medium 10 ROMKRT B. SHORT was taken to start new cultures during the experiments. After trying various con- centrations nf this hay medium it was found that the amoehae and food organisms remained healthier and increased more rapidly if the stock solution was diluted with t'our parts of hoiled and aerated spring water. Cultures of Chilomonas pnrauiccinin and Paramecium caudatuin were established in both the wheat and the hay culture media to serve as food organisms fur the amoebae during the experiments. Subcultures were made from these about every two weeks. Forty specimens of Amoeba curolincnsis were washed six times in boiled spring water. Ten were transferred to two stender dishes (five to each dish) containing paramecia in YY. C. M. In a similar manner ten were transferred to two stender dishes containing chilomonads in \V. C. M.; ten. to two stender dishes containing paramecia in H. C. M. ; and ten, to two stender dishes containing chilomonads in H. C. M. Each day the cultures were inspected and the supply of food organisms wa> replenished if it was low. The amoebae were counted daily for seven-day periods and then on the seventh day (in some cases a few days later) new cul- tures were made from the ones of the preceding week, so that by the end of the experiment most of the amoebae had been growing in one kind of medium and feeding on one kind of food organism for at least 46 days. With the exception of two cultures, the pH was taken on each culture on or after the seventh day. Dur- ing the experiment, observations as to the relative size and form of the amoebae were made; some of the amoebae in each culture were examined in hanging drops under the compound microscope; and at the end of the experiment photomicrographs were taken of a few amoebae in the wheat-culture medium. RESULTS After the first two weeks there were marked differences in the rates of repro- duction, especially between those amoebae grown in \Y. C. M. and fed on paramecia and the other cultures, as can be seen in Table I. During the first two weeks there seems to be a period of adjustment when the reproduction rates ot the different cul- tures do not vary significantly. However, during the third to sixth weeks the amoebae grown in \Y. C. M. and fed on paramecia increased rapidly — from 10 to 45 during the sixth week. The other cultures showed little or no growth. The hydrogen-ion concentrations remained rather constant. The pH's varied from 7.0 to (S.I, with that of the H. C. M. containing paramecia usually a little higher than the others. The room temperature varied from about 20° to 26° C. There was no significant difference in size of the amoebae in the various cult- tures, but there was a difference in form. Those growing in \Y. C. M. and fed on paramecia were often somewhat disc-shaped with short, blunt pseudopods radiating from all sides as shown in Figure la. Others were usually monopodal or bipodal with blunt pseudopods as shown in Figure lb. Clumping of the disc-shaped amoeba was ol ^served five or six times. The amoebae in W. C. M. fed on chilomonads were more flattened, with thinner pM-ndopoiK which were often at right angles to the main part of the body and tre- quently branched (Fig. Ic). The pseudopods often had little knob-like swellings at their distal ends (Fig. Id). The amoebae in these' cultures occasionally were monopodal in torm. OBSERVATIONS ON THE GIANT AMOEBA 11 TAHLK I Effect of culture media and food organisms on the reproduction rates of Amoeba carolinensis. At the beginning of each week amoebae were put into new cultures corresponding to the con- ditions under which they previously had been growing, so that the end results represent cumu- lative effects. Since there always were two cultures of each type, the pH's of both cultures are given. I Wheat medium \\ heat medium I lav medium Hav medium with paramecia with chilomonads with paramecia with cliilnnionads Number Number Number Number of of of of amoebae % amoebae % amoebae % amoebae •; in- pH in- pll in- pll in- pll crease crease crease crease First day Sev- enth day First day Sev- enth day First day Sixth day First day Sev- enth day First 10 17 709? 7.5 10 22 120% 7.5 10 18 80% 7.7 10 1 1 10% 7.5 week 7.6 7.5 7.8 7.5 Sev- enth day .Second 10 15 50% 7.6 9 14 55 \% 7.5 10 11 10% 8.0 10 10 0% 7.8 week 7.6 7.6 7.9 7.7 Third 10 30 200', 10 10 o%, 7.8 10 16 60% 8.0 10 10 0% 8.0 week 7.9 8.1 8.1 Fourth 10 48 380' , 7.0 10 10 0% 7.6 10 10 0% 7.6 10 11 10% 7.9 week 7.1 7.6 7.4 8.0 Fifth 10 57 470', 7.5 10 13 30%, 7.6 10 5 -50', 7.6 10 10 0', 7.6 week 7.4 7.6 7.4 7.6 Sixth 10 45 ,.50' , 7.3 10 12 20' , 7.4 5 4 -20', 7.8 10 10 0', 7.3 week 7.4 7.4 7.6 7.2 In H. C. M. the animals differed very little from each other as far as external appearance is concerned. They were, for the most part, monopodal or bipodal. Sometimes, however, those with the chilomonad diet exhibited the branched form characteristic of the amoebae in \Y. C. M. with chilomonads. Those amoebae writh the paramecium diet were usually slightly larger than the ones feeding on chilo- monads, and during the last two weeks one specimen grew to be the largest amoeba in any of the cultures. Very large vacuoles were apparent in the amoebae in H. C. M. with the paramecium diet, and some individuals were spherical. These last two characteristics are probably associated with degeneration. In Wilson's description (1900) of this giant amoeba, he mentions minute, elongate, and fusiform crystals in the endoplasm. \Yilber (1942) describes the crystals in more detail and says that there are two types: (1) plate-like crystals, and (2) bipyramidal crystals, which are the more numerous. He states that these crystals (presumably both kinds) are formed from food in the food vacuoles. Wil- son also mentions spherical bodies 8 p. in diameter and smaller, which resemble oil drops. Wilber gives the size range of these bodies as 2.5-8 /j. in diameter, and I-1 KOl'.KKT I',. SHORT .*>< ft - 1. Photomicrographs of amoebae thrown in wheat culture medium, a and I), those with parameeia as food organisms, e and d, those with chilomonads as food organisms. conclude^ ( 1('45) that they are funned fnun the crystals and the vacuole refractive bodies. He then says that the refractive spherical bodies function as reserve l<>od in the amoeba. \Yhile examining the stock and experimental amoebae under the compound microscope, differences in regard to the bi])yramidal crystals and spherical bodies \verc noteil. The plate-like crystals were so few in number that it was difficult to make a valid comparison in regard to their relative numbers and sixes. OBSERVATIONS ON THK (ilANT AMOKI5A 13 In the stock cultures the amoebae had few spherical bodies, the largest <>f which were about 5 /< in diameter. The largest bipyramidal crystals were about 2.4 p. long. The amoebae grown in W. C. M. with a diet of paramecia showed some larger bipyramidal crystals 2.8/^ long and a great number of smaller ones. There were few spherical bodies and the largest were only 3.1 ju. in diameter. Those specimens grown in W. C. M. and fed on chilomonads showed practically the same condition in regard to the bipyramidal crystals. However, there were very many large spherical bodies measuring 8.5 ^ in diameter. The spherical bodies of the amoebae grown in H. C. M. and fed paramecia were slightly more numerous and a little larger than the ones in the amoebae grown in W. C. M. with a paramecium diet. There were a few larger bipyramidal crystals 4.2 ju. long, more 2.8 p. long and some smaller ones. The amoebae grown in H. C. M. with chilomonads showed many large bi- pyramidal crystals 4.2 p, long, and the spherical bodies were very large and numer- ous. Many of the latter were 9.8/A in diameter, and some were slightly flattened on one side so that they looked like a three-quarter-full moon. DISCUSSION With the exception of size and number of nuclei. Amoeba prof ens and A. caro- linensis are very much alike. Schaeffer (1926) states, "My study of this -matter (a comparison of A. protcns and A. carolinensis} leads me to include difflnens (proteus) in the same genus with Chaos (the giant amoeba). These two species, as a matter of fact, resemble each other more closely than most other species within one genus." From Wilber's (1942) description of the cytology of A. carolinensis it is seen that the giant amoeba resembles A. protcns very closely in regard to the shape and structure of the nuclei and the cytoplasmic inclusions. Wilber (1945) has also shown that these two forms are very much alike in respect to the function of the nuclei, the formation of the contractile vacuoles, and the formation and function of the spherical bodies. When the data of these experiments are compared with similar work on A. protcns, it is shown that these two amoebae resemble each other in nutritional re- quirements. The numbers and relative sizes of the crystals and spherical bodies in the amoebae in my experiments agree in general with the results of similar studies by Mast (1939) and Mast and Hahnert (1935) on A. protcns. Concerning the numerous spherical bodies in the chilomonad-fed amoebae, Mast and Hahnert say, "This is of considerable importance for, as stated above, it indicates that the neutral red staining droplets found in abundance in Chilonionas but not in Colpidium function in the formation of the spherical bodies in Ann>cl>a." My experiments in- dicate that this is also the case in A. carolinensis. In examining the data of these nutrition studies it is seen that in general the amoebae with the most numerous and largest crystals and spherical bodies had the lowest reproduction rates. In view of the fact that the spherical bodies serve as reserve food (Wilber, 1945), these experiments indicate that even though chilo- monads may be ingested and digested by the amoebae, this diet is not adequate for normal reproduction. The reproduction rates in the different cultures agree with the results of Reynolds (1938) and Williamson (1944) for A. carolinensis, but do not correspond with Williamson's data on A. proteus. 14 knl'.KKT II. SHOk'T Tlu- relative sixes ol the amoebae in the various cultures do not agree very closch with the observations during similar experiments by Mast (1939) on A. protcus and \\'illiamson (1944) on ./. pro/cits and A. ctirolincnsis. In my experiments there was no increase in the si/e of the amoebae which had eaten ciliates, as found by both these authors, nor did the chilomonad-fed amoebae decrease in size, as found bv Williamson. Tn regard to the lorrn ol . /. carolinensis while utilizing paramecia as food organ- isms, the results are in accord with those of Mast and Root (1916) for A. proteus and \\~illiamsou ( 1°44) for ./. protcns and A. cai'ulinciisis. NUCLEAR DIVISION MATERIALS AND METHODS Specimens were taken from the stock cultures and fixed with Carnoy's. Helling's, and Bouin's fixatives. Those fixed with Carnoy's and Belling's fluids were stained with Delafield's and Heidenhain's haematoxylin. Those fixed in Bouin's fixative were stained with Heidenhain's haematoxylin. Some specimens were sectioned at 10 /j.. Since those fixed with Carnoy's fixative and stained with Heidenhain's haematoxylin showed the nuclear structure best and stained the cytoplasmic inclu- sions very little, all the remaining work <>n nuclear divisions was done with these reagents. In order to secure nuclear divisions, slides were made at various hours of the day. All the specimens showing nuclear division stages were fixed between 5:45 p.m. and 11 :00 p.m. and were more or less spherical in form. The experiments are not extensive enough to warrant anv statement as to a periodicity of mitosis. This seeming periodicity was probably a result of better selection toward the end of the experiment. OBSERVATIONS AND DESCRIPTIONS ticstiiif/ nucleus The resting nucleus of A. carolinensis ( Fig. 2a) is disc-shaped, and measures about 24 X 10 p.. Immediately beneath a distinct nuclear membrane and adhering closely to it are darkly staining granules (peripheral granules), the largest of which are about 2 /t in diameter. The interior of the nucleus has a finely granular ap- pearance and stains more lightlv than the peripheral granules. Schaeffer (1938) states that all the nuclei of an individual divide at the same time. Yet, nuclei at slightly different stages of division can be found in the same animal. Prophase In earlv prophase the nucleus enlarges apparently bv a pulling awav ot the mem- brane from the central granules leaving a large endosome about l('/i in greatest diameter. The nucleus now measures 27 : 15. 4 /z. Most of the peripheral granules are still rather small and stain more darklv than the endosome. Some of them however are slightly larger and less deeply stained than those in the resting nucleus. Kaint strands running from the endosome to the periphery are now apparent. In a little later stage ( Fig. 21> and c) the nucleus has become more spherical measuring about 27 ( 22 //. Some of the peripheral granules have become loosened OBSERVATIONS ON THE GIANT AMOEBA 15 » — • FIGURE 2. Camera lucida drawings of the stages of mitosis in .huocha carolincnsis. a. resting nucleus ; b. propliase, face view ; c. prophase, edge view ; d. very late prophase showing chromosomes and spindle forming ; e. metaphase, edge view ; f. metaphase, polar view ; g. late anaphase; h. early telophase, showing one plate with granules at the pole: i. a little later stage showing one pole with a delicate membrane and granules; j. late telophase with membrane fully formed. All drawings X 1660. 16 R( (BERT I',. SHORT from tin- membrane, are more spherical, and show a lighter area in the center. The endosome has become smaller and more compact, and now stains about as deeply as the peripheral granules. At this statue the endosome is a thin disc measuring 9.3 ju in diameter and 1 .5 /*, in thickness. The reticulum connecting the endosome with the periphery is more evident now. Very late prophase (Fig. 2d) shows the chromosomes becoming arranged on the plate which is 14.2 /JL in diameter. Spindle fibers are distinct and the peripheral granules have practically disappeared. A lighter area in the cytoplasm immediately surrounding the nucleus is visible. The nucleus shown in Figure 2d was the only nucleus in the animal in this condition. The remaining nuclei were in metaphase or early anaphase. Metaphase At metaphase (Fig. 2e and f) the nuclear membrane has disappeared. There are. however, around the periphery of the plate delicate, blister-like structures which may be remnants of the nuclear membrane. The chromosomes, which are spherical or ellipsoidal, and perhaps about 300 in number , are arranged on a discoidal plate 13. 2 fa in diameter. The split halves of the chromosomes in some cases can be seen. The spindle fibers are at right angles to the chromosome plate and are about 4 p, long. No centrioles or granules are apparent at the ends of the spindle fibers. There is still a lighter area in the cytoplasm around the figure. Anaphase During anaphase (Fig. 2g) the chromosome plate splits into two daughter plates which diminish in diameter with the chromosomes becoming so closely aggregated that they no longer can be distinguished individually. In late anaphase the plates measure 8.4/x in diameter and are, for the most part, flat. A few are slightly arched or saucer-shaped. The spindle between the chromosome groups has in most cases become twisted as if both plates had rotated in opposite directions. The polar libers seem to be liner and more numerous than the interzonal ones, and the polar areas appear slightly darker than the surrounding cytoplasm. The polar fibers have not shortened during anaphase but are in most cases 5 \\. long, and the outside ones are inclined at an angle of about 60° to the plates. In an amoeba containing thirty-three nuclei, twenty-nine are in late anaphase and four are in early telophase. The nuclei in anaphase have their polar fibers in- clined at an angle of 60° to the chromosome plates. The four in early telophase have granules arranged on a more or less hemispherical surface or membrane (Fig. 2h and i). Twenty-nine of the spindles lie so that the distance between daughter chromosome groups can readily be measured. The average distance between groups is 34 //. The shortest distance is 10//, and the longest distance is 62//. Some of the chromosome plates are tilted at angles to each other so that one of a pair is seen in edge view, while the other is seen in polar view. Some pairs of plates, both showing in edge view, are twisted at angles of 30 to 150 to each other. In the twcntv-one anaphase spindles which lend themselves to analysis, twenty have the spindle between the chromosome groups twisted clockwise, or to the apparent right (Fig. 2g). ( )ne spindle shows no twisting. This constancy in direction indicates that the twisting is caused not by external forces in the cytoplasm, but by forces inherent in the spindle apparatus. OBSERVATIONS ON THE GIANT AMOEBA 17 | Telophase In early telophase (Fig. 2h) granules appear at the distal ends of the polar fibers. A delicate, more or less hemispherical, membrane forms at the poles and the ends of the fibers nearest the poles disappear (Fig. 2i). Later telophase (Fig. 2j) shows the plates larger in diameter, less dense and more finely granular, with the membrane more flattened. The granules are more densely packed on the median sides of the nuclei where the membrane is scarcely visible except near the edge of the disc. The spindle fibers have completely disap- peared by this time. DISCUSSION Mitosis in A, carolinensis as herein described is very similar to that of A. prote-us as described by Chalkley and Daniel (1933), Chalkley (1936), and Liesche (1938). The figures and descriptions of the nuclear division stages of A. proteus agree with those of A. carolinensis with the following exceptions: 1. The nuclei and mitotic figures of A. carolinensis are approximately half the size of those of A. proteus. Also the chromosome number of A. carolinensis (prob- ably near 300) seems to be about half that given by Liesche (500-600) for A. proteus. 2. No "spireme" (Liesche) stage was observed in A. carolinensis. 3. No granules (Chalkley and Daniel) were visible at the distal extremities of the spindle fibers at metaphase. 4. During metaphase and anaphase grouping of the distal ends of the polar spindle fibers is not so pronounced in A. carolinensis, and during anaphase the chromosome plates are only slightly arched. SUMMARY 1. It is concluded that the giant amoeba described by Wilson (1900) is not a Pelovnyxa but belongs to the genus Amoeba and should be designated A. carolinen- sis. Chaos chaos is considered to be invalid as a name, owing to the fact that it is based on Roesel's inadequate description. 2. Nutritional studies indicate that A. carolinensis has food requirements similar to those of A. proteus and that these two species react in a similar manner to the same type of food. 3. Nuclear division is described for A. carolinensis. The stages are similar to those described for A. proteus, and except for smaller size, correlated with a smaller number of chromosomes, the differences are insignificant. LITERATURE CITED BOURNE, A. G., 1891. On Pelomyxa viridis, sp. n., and on the vesicular nature of the protoplasm. Quart. Jour. Mic. Sci., 32 : 357-374. CHALKLEY, H. W., 1936. The behavior of the karyosome and the "peripheral chromatin" dur- ing mitosis and interkinesis in Amoeba proteus with particular reference to the morpho- logic distribution of nucleic acid as indicated by the Feulgen reaction. Jour. Morph., 60: 13-29. CHALKLEY, H. \V., AND G. E. DANIEL, 1933. The relation between the form of the living cell and the nuclear phases of division in Amoeba proteus (Leidy). Physiol. Zool., 4: 592- 619. 18 ROBERT B. SHORT GREEFF, R., 1874. Pelomyxa palustris (Pelobius), ein amobenartiger Organismus dcs siissen Wassers. Arch. f. Mikr. Anat., 10: 51-73. HEGNER, R. W., AND W. H. TALIAFERRO, 1924. Human protozoology. LIESCHE, W., 1938. Die Kern- und Fortpflanzungsverhaltnisse von Amoeba proteus (Pall.). Arch, f. Protistcnk., 91 : 135-186. MAST, S. O., 1938. Amoeba and Pelomyxa vs. Chaos. Turto.v News, 16: 56-57. MAST, S. O., 1939. The relation between kind of food, growth, and structure in Amoeba. Biol. Bull, 77 : 391-398. MAST, S. O., AND W. F. HAHNERT, 1935. Feeding, digestion, and starvation in Amoeba proteus (Leidy). PhysioL Zool., 8: 255-272. MAST, S. O., AND P. L. JOHNSON, 1931. Concerning the scientific name of the common large amoeba, usually designated Amoeba proteus (Leidy). Arch. f. Protistcnk., 75: 14-30. MAST, S. O., AND F. M. ROOT, 1916. Observations on Ameba feeding on Infusoria, and their bearing on the surface tension theory. Proc. Natl. Acad. Sci, 2: 188-189. PENARD, E., 1902. Faitnc Rhisopodique du Bassin dn Lcnian. Geneva. REYNOLDS, B. D., 1938. The effect of different food organisms on the growth rate of Chaos carolinensis and the length of time such organisms will live in food vacuoles. Va. Acad. Sci, Proc. 1937-1938: 42. RICE, N. E., 1945. Pelomyxa carolinensis (Wilson) or Chaos chaos (Linnaeus)? Biol. Bull.. 88: 139-143. ROESEL VON ROSENHOF, A. J., 1755. Der kleine Proteus. Dcr Insecten-Beliistigung., 3: 622-624. SCHAEFFER, A. A., 1918. Three new species of amebas : Amoeba bigemma nov. spec., Pelomyxa lentissima nov. spec., and P. schiedti nov. spec. Trans. Aincr. Mic. Soc., 37: 79-96. SCHAEFFER, A. A., 1926. Taxonomy of the Amebas. Cam. Inst. U'ash. DC ft. Afar. Biol., 24: 1-116. SCHAEFFER, A. A., 1938a. Further data on the name Chaos chaos Linnaeus as referring to the giant amoeba of Roesel. Tiirto.v Nczvs, 16 : 96-97. SCHAEFFER, A. A., 1938b. Significance of 3-daughter division in the giant amoeba. Tiirto.v News, 16 : 157-160. WILBER, C. G., 1942. The cytology of Pelomyxa carolinensis. Traus. Anicr. Mic. Soc., 61: 227-235. WII.BER, C. G., 1945. Origin and function of the protoplasmic constituents in Pelomyxa caro- linensis. Biol. Bull., 88 : 207-219. WILLIAMSON, J., 1944. Nutrition and growth studies of amoeba. Physiol. Zool., 17: 209-228. WILSON, H. V., 1900. Notes on a species of Pelomyxa. Amcr. Nat., 34: 535-550. THE ELIMINATION OF CHROMOSOMES IN THE MEIOTIC DIVISIONS OF BRACHYSTETHUS RUBRO- MACULATUS DALLAS FRANZ SCHRADER Department of Zoology, Columbia University, Xcw York The occurrence of meiotic abnormalities is anything but rare. However, in almost every instance it is the result of mechanical or physiological accidents and therefore sporadic and irregular. Hence, if in a certain species a departure from the normal process always takes a very particular form and is restricted to a certain region of the gonad, it is likely that it is not entirely accidental. If, furthermore, every male or female in the species is affected, it is safe to assume that we are dealing with a basic and well regulated condition, unorthodox though it may be. The case in question is that of Brachystethus rubrowiaculatus, one of the penta- tomid Hemiptera. It holds for all the males of the species. The testis here is composed of four lobes and in the fourth of these (counting from the side where the sperm duct makes its exit) the meiosis follows an abnormal out very definite course. In a way, this is analogous to the case of Loxa (Schrader, 1945a and b), but that is only in the sense that in both the exceptional development takes place in a cer- tain lobe of every testis. The nature of the abnormality is quite different in the two species, as will appear in the account below. My chief interest lies in the possibility that such irregularities may be of use in the analysis of the ordinary mechanism of mitosis and in this as well as several succeeding studies I hope to show that they may throw some light on certain puzzling aspects of the division cycle. MATERIAL AND METHODS The gonads of four males and one female were used in the investigation. The insects were collected near Turrialba, Costa Rica, during April and May, 1944. The material was fixed in Sanfelice and sectioned at 5 to 10^. Gentian violet, haematoxylin and the Feulgen reaction were used in staining, the haematoxylin being especially useful in the analysis of spindle conditions whereas the Feulgen reaction is indispensable in following the maneuvers of the chromosomes. The main manifestations of the mitosis are so clear that they are well shown by pen and ink drawings — a method which in most other cases docs not give a just presentation of spindle conditions. I take pleasure in expressing my gratitude to Dr. E. N. Bressman and Mr. R. A. Nichols of the Inter-American Institute of Agricultural Sciences for facilitating the work. My special thanks are due to Dr. T. J. Grant of the U. S. Department of Agriculture, whose ready helpfulness did so much to further my researches in Costa Rica. NORMAL SPERMATOGENESIS The testis has only four lobes of which the first three show an orthodox sperma- togenesis differing in no essential from that described in so many other pentatomids. 19 20 FRANZ SCHRADER The fourth or "harlequin" lobe has about the same proportions as the rest and does not differ in any discernible way in its general organization. The size differences among the chromosomes as seen in the spermatogonia are not very great. But it is possible to recognize one large, four medium and one small pair of autosomes. The X is a trifle larger than one of the small autosomes while the Y is the smallest chromosome of the complement (Fig. 1). Identifica- tion of the sex chromosomes was checked by examination of oogonial metaphases in the female. The normal meiosis shows the usual prophase conditions, with the leptotene, synaptotene, pachytene, and diplotene stages followed by the confused period in which chromosomal behavior is difficult to analyze. This is succeeded by a diaki- nesis marked by the appearance of beautiful tetrads which, as they condense, are distributed around the nuclear periphery. Just before the breakdown of the nuclear membrane this rather even distribution begins to disappear and the chromosomes may even come in contact with each other in small groups of two or three each (Fig. 2). This collocation reaches its height when the membrane disintegrates. All of the chromosomes then huddle together in the middle of the nuclear space only to separate again almost at once to form the equatorial plate. In the latter they follow the characteristic pentatomid arrangement of the first metaphase, with one or both sex chromosomes taking a central position within a ring of autosomal tetrads (Fig. 3). The sex chromosomes are marked by their heteropycnosis from the leptotene period on. Until the synaptic period they are well separated from each other, but then tend to come together. During the confused period they form a single, rounded chromatin nucleolus and it is not until after the early diakinesis that they j •/ once more become independent of each other. They divide equationally in the first, and segregate in the second division, so that the spermatids carry 6 + X or 6 + Y chromosomes (Figs. 3-5). There is no interkinesis and the two centrioles at each pole have already sepa- rated to assume their position for the second division before the chromosomes have completed the anaphase movement of the first. This precocious behavior of the centrioles is however also encountered in several other pentatomids and does not interfere with the normal -meiotic distribution of the chromosomes except in such extreme cases as Peromatus (Schrader, 1941b). ABNORMAL SPERM ATOGENESIS First division Neither in the spermatogonial nor the meiotic behavior up to the late diakinesis does the harlequin lobe differ from the other three. Six autosomal tetrads are formed which resemble in every way those seen in the normal lobes (Fig. 6). Again, as these tetrads condense into the characteristic dumbbell shaped bodies, they and the sex chromosomes are well distributed around the nuclear periphery. But then arises a difference in behavior from the normal. Here, too, the last phase prior to membrane disintegration witnesses a coming together of the chromo- somes. In this fourth lobe, however, the collocation is both more regular and more pronounced. In nearly all cells the six tetrads form a single chain whose con- stituents maintain contact with the nuclear periphery. This chain is always more BRACHYSTETHUS RUBROMACULATUS DALLAS 21 or less equatorially placed (Fig. 7 and 8). The configuration is evidently the re- sultant of several different forces. Just as in normal cells, the tendency to collocate begins to manifest itself in late diakinesis. But in this fourth lobe the autosomal tetrads are also repelled by the two opposite poles at this early stage. Since the chromosomes are still confined within the nuclear membrane they consequently move into the middle region. The combination of polar repulsion, the tendency to collocate, and adhesion to the still intact nuclear membrane must perforce result in the formation of an equatorial chain. Drawings were made from haematoxylin preparations, except for Figures 12 and 24. All figures magnified approximately 1400 X. Autosomes drawn in outline, and sex chromosomes in solid black throughout. FIGURE A. Normal spermatogenesis. 1. Spermatogonial metaphase; 12 autosomes and X (large) and Y. 2. Late diakinesis; beginning of autosomal clumping. 3. Metaphase of Divi- sion I; X and Y in middle. 4. Metaphase of Division II; X superimposed on Y. 5. Telophases of Division II showing two types of spermatids : 6 autosomes + Y, 6 autosomes + X. As one might expect, there is no co.nstant seriation in such a chain. When the nuclear membrane disintegrates, the chain is converted into an irregular clump just like the clump formed by the tetrads of normal cells that have undergone no such maneuvers (Fig. 9 and 10). Much less frequently is there a formation of two smaller aggregates instead of the single large one. The two sex chromosomes are included in such groupings only by accident and sooner or later they always separate from the autosomes and assume an independent position. This is not necessarily an equatorial one at first (Fig. 8-10). It is, however, in the establishment of the first metaphase that the most striking departures from an orthodox behavior are evinced. As already stated, the break- down of the nuclear membrane is followed immediately by the clumping of the autosomal tetrads in the midregion. In many cells this may, however, be halted temporarily if a central core of spindle fibers is formed quickly between the poles. The clumping autosomes may then be applied to these central fibers in a half moon 22 FRANZ SCHRADER configuration for a moment (Fig. 11). But in any case, before the metaphase spindle has assumed final shape, the autosomal aggregate is shifted out of this gen- eral middle region. Almost always this movement seems to be a sudden one and frequently the aggregate comes to rest rather close to the lateral wall of the cell, the direction of the shift being toward the side and never toward the poles. The cell frequently bulges out on the side on which the aggregate is located (Fig. 14). If instead of one aggregate, two smaller ones have been formed, they undergo similar reactions and often become displaced toward opposite sides of the cell (Fig. 13). The beginning of this autosomal shift sees the X and Y in no definite position although usually in the vicinity of the polar axis. Frequently they are situated at some distance from the equatorial plane and bear neither a constant positional rela- tion to each other nor a mitotic orientation toward the centrioles (Fig. 9 and 10). KK;URE B. Abnormal spermatogenesis. Division 1 from diakinesis to early anaphase. 6. Mid diakinesis showing tetrads of normal appearance. 7. Late diakiiu-sis in polar view ; auto- somal tetrads forming equatorial chain. 8. Late diakinesis in side view. 9. Early stage in clumping of autosomal tetrads. 10. Beginning of movement of autosomal aggregate away from polar axis. 11. Displaced autosomal aggregate applied to central portion of spindle in half moon form; X and Y assuming equatorial position. 12. Feulgen preparation corresponding to Figures 9 or 10 showing that tetrads retain their individuality in the aggregate. 13. Autosomes in two aggregates, hoth shunted out of polar axis. 14. Early anaphase, with X clearly showing "tertiary" split ; characteristic displacement of autosomal aggregate. BRACHYSTETHUS RUBROMACULATUS DALLAS However this situation is quickly altered and as the autosomal aggregate moves toward the side of the cell, the X and Y approach very close to the polar axis and assume a position in the equator with a definite orientation toward the poles (Fig. 11 and 13). In all of the hundred or more cells observed at this stage, such a configuration of autosomes and sex chromosomes is always maintained. Since in the prometaphase the two types of chromosomes form one general group albeit not always in contact with each other, it is clear that the later separation is not an accidental one. The unusual shift of the autosomes is equivalent to an actual extrusion from the mid- region of the cell. Despite their distance from the polar axis, the autosomes continue to be con- nected with both poles by chromosomal spindle fibers. This is at first glance rather surprising since with ordinary stains like gentian violet and haematoxylin the auto- somal aggregate appears as a solid, structureless mass which plainly suggests de- generation. In good Feulgen preparations, however, it becomes clear that the auto- somal tetrads have by no means lost their individuality. All six of them can be easily distinguished, lying in a substance which does not stain with Feulgen and thus reveals the individual chromosomes (Fig. 12). This substance, which with other stains becomes just as dark as the chromosomes themselves, resembles the material surrounding the sex chromosomes of certain reduviid Hemiptera (Troeds- son, 1944) and may also be akin to the "flocculent, whey-like coagulum" which envelops some regions of the chromosomes in Olfersia (Cooper, 1944). In Brachystethus it is evidently formed when the chromosomes clump at prometaphase and it persists through both meiotic divisions. In the middle of the cell, between the two poles, there is a wTell-formed spindle of normal length, which however is smaller in diameter than the spindle of normal cells. This accommodates the two sex chromosomes, which are located in the middle of the spindle substance just as they would be if the autosomes were free to form a ring around them. Despite the presence of the chromosomal fibers which connect the autosomes with the poles, the autosomal aggregate behaves more or less like an inert mass in the mitosis that follows. While the sex chromosomes undergo an equational divi- sion and approach the poles, the autosomes near the periphery of the cell undergo no movement (Fig. 14-17). It is only when the dividing cell elongates and be- comes narrow that the autosomal clump is moved toward the midline (Fig. 18). If it there comes in contact with the expanding interzonal region or "Stemmkorper" it may be swept along into one of the two daughter cells, but very often it is not until the cleavage constriction is being completed that the aggregate is definitely included in either of the resulting cells (Fig. 19 and 20). Apparently the chromo- somal fibers exert little or no traction at this time. So far as one can tell, the in- clusion of the autosomes in either daughter cell occurs entirely at random. It is of interest to note that although in its essentials the equational division of the sex chromosomes is always accomplished successfully, there are certain features that distinguish it from the corresponding process in normal cells. In the first place, the sex chromosomes and especially the X clearly show a tertiary split already at this first metaphase — a split which is not utilized until the first somatic division in the egg (Fig. 13-15). This is sometimes indicated in normal cells also, but never so strikingly as here where it appears after all three of the stains used. The second 24 FRANZ SCHRADEK feature lies in the anaphase behavior of the sex chromosomes. Although both start out from an equatorial position, the early anaphase often shows the chromatids of the Y much closer to the poles than those of the X (Fig. 15). In other words, the Y precedes the X in the poleward movement. When there is such a disparity of movement, there is frequently a shift during mid anaphase through which these four chromatids are all brought into the exact polar axis in a single line. Since as just stated, the chromatids of the Y move more quickly, they constitute the extremities in this tandem arrangement, whereas the two X chromatids are in between (Fig. 16). Such a sedation strongly resembles that which characterizes the second di- 16 18 19 20 FIGURE C. Abnormal spermatogenesis. Division I, anaphase and telophase. 15. Early anaphase with Y preceding to the poles. 16. Early anaphase; X chromatids have moved into polar axis between separating Y chromatids. 17. Mid anaphase. 18. Late anaphase; autosomal aggregate returning to axial region. 19. Early telophase; an X and a Y at each pole with the autosomal aggregate included in upper cell. 20. Late telophase with two autosomal aggregates both included in upper cell. vision of the coccid Protortonia (Schrader, 1931), and there is little doubt that similar forces are involved. This anaphasic shift does not affect the result of the division ; one X and one Y chromatid go into each of the resulting second sperma- tocytes. Thus in the majority of cases, two types of second spennatocytes are produced : one carries an X and a Y, as well as the clumped autosomes ; the other contains only the two sex chromosomes (Fig. 19). If the autosomes have been aggregated in two masses before metaphase, these may both go to the same pole or to opposite poles, apparently at random (Fig. 20). In no case is there a division of the in- dividual tetrads. BRACHYSTETHUS RUBROMACULATUS DALLAS 25 Second division As in normal cells, the centriole at each pole of the first spindle is divided and the two daughter centrioles separate some time before the division has been finished (Fig. 21). Each moves through an arc of 90° and the new polar axis for the second division is therefore at right angles to the first. Even while still in telo- phase, the two sex chromosomes often respond to the new poles and move toward them (Fig. 21). However, this precocious movement is soon reversed and the X and Y then come together in the middle of the new spindle in their orthodox "touch and go" pairing (Fig. 22 and 28) and it is only following this that they separate in the regular segregation toward opposite poles. There is no indication of any interkinesis. The autosomal aggregate may at first remain in the general region where it has entered the cell upon completion of the first division, but in other cases it approaches closely to the two sex chromosomes that are near the new polar axis (Fig. 21). Such an approach is only temporary, however, and before the second division is initiated the aggregate is always extruded from this middle region just as it was in the first division (Fig. 23). The configuration of the second division thus almost duplicates that of the first. FIGURE D. Abnormal spermatogenesis. Division II of cell containing autosomal aggregate (all drawings except Fig. 21 show X below Y). 21. Beginning of Division II ; X and Y reacting to new position of poles. 22. Touch and go pairing of X and Y ; autosomal aggregate shunted away from polar axis. 23. Early anaphase ; X and Y going to opposite poles ; autosomal aggre- gate in characteristic displacement. 24. Feulgen preparation showing that tetrads still appear unaltered in early phase of Division II. 25. Anaphase; autosomal aggregate returning toward polar axis. 26. Telophase ; autosomal aggregate included in cell with Y. 27. Spermatids ; upper cell carries only a micronucleus with the Y half chromatids; lower cell with micronucleus con- taining X half chromatids and autosomal nucleus. 26 FRANZ SCHRADER The autosomal tetrads are still intact and are still imbedded in the substance that stains intensely with gentian violet and haematoxylin. However, the aggregate is usually smaller than it was just prior to the first division, a fact that results from a greater crowding of the tetrads, as can be seen in Feulgen preparations (compare Fig. 24 with Fig. 12). Again, chromosomal fibers are formed (Fig. 22), although the aggregate seems to be moved about as a more or less passive body (Fig. 25 and 26). In short, just as in the first division it is only the sex chromosomes that are in- volved in the regular mitotic mechanism. But now, as in normal second divisions, the process is reductional and the resulting cells receive only an X or a Y (Fig. 26 and 30). The autosomes if present are once more included in either cell as a group, and so far as one can tell, on the basis of chance (Fig. 23-26). The spermatids that result from these two peculiar divisions thus either carry only one sex chromosome (the X or the Y) or else they contain the autosomal aggregate in addition. In other words, there are four main types of spermatids : X ; Y ; X -f- autosomes ; Y-{- autosomes (Fig. 26 and 30) . FIGURE E. Abnormal spermatogenesis. Division II of cell lacking autosomes (all draw- ings show X below Y). 28. Touch and go pairing of X and Y. 29. Early anaphase. 30. Late telophase. -31. Spermatids; upper cell with micronucleus containing the two Y half chromatids, lower with two X half chromatids. If the autosomes are gathered in two instead of the usual single group prior to the first division, the resulting spermatids may of course carry intermediate num- bers of chromosomes. But for the great majority of cells they are transmitted in a single aggregate. Throughout there is no attempt at a division of the individual autosomal tetrad, and it is as an aggregate of tetrads that the autosomes enter the spermatid. Obviously there is some factor which interferes with their mitotic mechanism — a factor which interferes in no way with that of the sex chromosomes. The latter behave just as they do in the normal cells of neighboring lobes of the testis. Spermateleosis The first steps after the telophase of the second division parallel the normal course of events. The transformation of the autosomal mass into the spherical nuclear structure shown in Figure 27 is peculiar, though in later stages it approaches BRACHYSTETHUS RUBROMACULATUS DALLAS 27 closely the condition seen in the normal spermatic!. In these abnormal cells how- ever, the sex chromosome at first forms a micronucleus which is distinct from the larger autosomal nucleus (Fig. 27). But sooner or later, the micronucleus be- comes applied to the large nucleus and gradually merges with it. When, as in most cases, a single sex chromosome constitutes the only chromo- somal constituent of the spermatid, its behavior differs in no way from that in cells containing the autosomes as well. In either case a tiny nucleus is formed by each sex chromosome and in this the constituent half chromatids separate to form two distinct chromosomal bodies. (Fig. 27 and 31). This precocious separation is not surprising when it is remembered that already in the first metaphase these half chromatids can be clearly identified as such. 33 FIGURE F. Spermatids before elongation. 32. Normal spermatids shown to indicate size. 33. Abnormal spermatids at stage corresponding to Figure 32. Largest spermatids contain all autosomes as well as either X or Y. Smaller spermatids carry either X or Y ; latter is the smallest. The process of transformation into the sperm seems at first alike in all sperma- tids and homologous to that observed in the normal lobes. This is true whether they carry only a single chromosome or the autosomes as well. As might be ex- pected, the former are distinctly smaller than the corresponding normal cells and nuclei (Fig. 32 and 33) ; the latter on the other hand are markedly larger. It is also interesting to note that it is often possible to distinguish between the small sperms carrying either an X or a Y. The latter are the smaller of the two (Fig. 33). However, when the rounded sperm heads begin to elongate, those which carry only a sex chromosome begin to stain less intensely than either normal or "auto- somal" abnormal sperms, and as the elongation into the typical tenuous sperm head continues it becomes more and more difficult to trace their further fate even in Feulgen preparations. It is likely that the "sex chromosome sperms" never attain 28 FRANZ SCHKADER the final stages of sperm formation. However, this cannot be made certain without smear preparations for even in very thick sections the long sperm heads are prac- tically always cut and it is impossible to decide whether one is dealing with an ab- normally small sperm or only a portion of a larger sperm. The fate of the giant sperms, however, is subject to no such doubt. They continue their development to the fully formed stage, enter the sperm duct, and are there mingled with the normal sperms from the normal lobes. GENERAL CONSIDERATIONS The cytological features that characterize this abnormal development are thus of the most striking sort. The most obvious one lies in the clumping of the auto- somes. A temporary collocation of chromosomes is of course encountered as a normal occurrence in many species, but is there, confined — as it is in the normal lobes of Brachystethus — to a very brief, almost momentary period immediately after the breakdown of the nuclear membrane in prometaphase. In Brachystethus there is some tendency toward a collocation of autosomes even before this event. But in the abnormal lobe, the clump that is finally formed is not resolved again, and this condition is maintained through both meiotic divisions and into sperm forma- tion. Moreover this collocation of chromosomes is so intimate that only a Feulgen preparation reveals that there has not been an actual fusion. Although the further behavior of this autosomal aggregate suggests strongly that it is shunted about more or less like a passive body during cell division, it nevertheless evinces certain reactions which indicate that it is not completely inert. In diakinesis, these autosomal tetrads duplicate the behavior of normal tetrads in establishing contact with the nuclear periphery; toward the end of the diakinetic period they evince a reaction to the two poles in taking up an equatorial position in contact with the nuclear membrane ; after the disappearance of the latter they form chromosomal fibers to the poles ; and almost simultaneously they move out of the midregion of the cell with a suddenness that bespeaks a forcible displacement. It is possible and even likely that this movement toward the side of the cell is due to the same centriolar repulsion which forces the. chromosomes into the equa- torial plane while they are still held within the confines of the nuclear membrane. An influence of centrioles on the chromosomes prior to the dissolution of the nuclear membrane is observed in certain other forms also ( for instance in Aniso- labis, Schrader, 1941a). But not often is it exerted so as to bring about an equa- torial placement at so early a stage, even though a role in the later final formation of the equatorial plate is frequently assigned to it. It will be realized, however, that even in normal cases, additional factors must function to restrict the chromosomes to the middle of the equatorial plane. Re- pulsion from the two poles alone cannot do that. It is possible that the chromo- somes are thus confined to the midregion simply because of surface tension condi- tions that prevail in the spindle body at the time of metaphase. The escape of the autosomes in the present case might then be attributed to abnormalities in the spindle, say to untoward alterations in the timing of the normal viscosity changes. But that is clearly not the entire explanation since the autosomes are not accom- panied in their displacement by the two sex chromosomes. The latter remain in the spindle in a quite orthodox position. Since the cytoplasm, the centriolar forces, BRACHYSTETHUS RUBROMACULATUS DALLAS 29 and the general spindle conditions are identical for both autosomes and sex chromo- somes, it is therefore in the chromosomes themselves that the explanation must be sought. More specifically, the question to be solved is why the autosomes do not respond to the forces which at metaphase counteract the influence of the centers and confine the chromosomes to the middle of the equatorial plane. The abnormal condition of the autosomes is not indicated by any striking be- havior during the prophase. There is only a more pronounced tendency to assume an equatorial position at prometaphase and the formation of chains that are absent in normal cells. Also, after the aggregate has been formed, Feulgen preparations show the component tetrads to be somewhat swollen and less intensely stained than are the prometaphase tetrads. This probably indicates the first step in a return to a diffuse state. Such a regressive condition may in some way be responsible for the special maneuvers of the autosomes, for the X and Y chromosomes which during this time remain fully condensed behave just like the sex chromosomes of normal cells. But it is not certain that this would suffice as an explanation since in addi- tion to the clumping of the autosomes one must also account for their elimination from the midregion and their failure to take the first steps in division. For it must be remembered that the initial separation of daughter chromosomes is an autonom- ous process which is independent of spindle action. Nevertheless, although the Brachystethus tetrads appear ready for such a step and the line of demarcation be- tween the paired chromosomes is clearly marked and complete, no separation ever occurs. Either there is no mutual repulsion or else the pellicle and the achromatic constituents of the tetrads prevent the normal division. In any case it is clear that some of the basic reactions of the chromosomes have been altered. It is natural to seek a parallel in other cases already on record. The investiga- tion of certain echinoderm hybrids by Baltzer (1910) is especially pertinent. Baltzer found that in certain of these crosses, the paternal chromosomes are elimi- nated during, cleavage. There, too, is a tendency for such chromosomes to clump, and there also a formation of chromosomal spindle fibers occurs nevertheless. Likewise there is sometimes a lateral elimination of these chromosomes although they usually simply lag in the middle of the spindle. However, all this does not occur until the anaphase is well advanced and the various configurations are by no means as constant as they are in Brachystethus. After a careful analysis, Baltzer concluded that it is the chromosomes themselves rather than the plasma that is re- sponsible for the elimination. That such a conclusion is warranted for the case of Brachystethus, as well, has already been pointed out. The difference in behavior between the autosomes and the sex chromosomes in the same plasma makes it un- avoidable. The elimination of chromosomes consequent on a loss of their kinetochores is observed regularly in certain molluscs (Pollister, and Pollister, 1943). Such an explanation is worthy of serious consideration in the case of Brachystethus. However, there can be only a partial loss of the kinetochore activity because some chromosomal fibers are evidently still being formed. Since we are dealing here with a diffuse kinetochore, such a partial loss is easily conceivable, but it must be confessed that the akinetic chromosomes of molluscs present very different elimina- tion pictures than those seen in Brachystethus. The conditions in such coccids as Phenacoccus (Hughes-Schrader, 1935) may also be germane to the present case. There, too, certain chromosomes betray a 30 FRANZ SCHRADER tendency to clump, and it is these chromosomes that become inert and degenerate within one or two succeeding divisions. Nevertheless, these chromosomes form normal chromosomal fibers. But all these other instances are themselves in need of further explanation. The only conclusion that protrudes itself in such an analysis is that the abnormality is to be sought in the basic organization of these forms and is not a superficial and accidental one. The alterations that are involved primarily affect the chromosomes and influence their reactions to each other and to the mitotic mechanism. EVOLUTIONARY ASPECTS It is very questionable whether spermatids carrying only an X or a Y chromo- some ever develop into mature sperms. That, however, is not true of the giant sperms which in addition to a sex chromosome carry the full set of autosomes un- affected by any meiotic mitosis (i.e., four times the number of autosomes contained in a normal sperm). But though these large sperms enter the sperm duct and mingle with the normal sperms, it is doubtful whether they ever become functional in the hereditary sense. Of the seven specimens of Brachystethus that I have ex- amined, none was marked by unusual morphological features such as confidently might be expected if a sperm nucleus with a quadruple set of autosomes joins the haploid nucleus of a normal egg. But if these numerous giant sperms play no direct role in the hereditary mecha- nism of the species, it becomes a matter of wonder that such an extensive develop- ment of abnormal gametes could have withstood the effects of natural selection. For it must be remembered that the testis of Brachystethus has only four lobes and if the sperms of one of these do not function in the genetic determination of the embryo, we are confronted with a prodigious waste which is added to that which occurs normally in the reproduction of most male animals. Such a waste is paralleled in several species of Loxa (Schrader, 1945a and b) where there is likewise an abnormal lobe in every testis. Although the nature of the aberrancy is different in the two cases, they are similar in that both encompass the production of a huge number of sperms which carry many times the normal number of chromosomes. It seems strange that two processes so diverse in their abnormalities should culminate in gametes wrhose general characters are so similar. The explanation may lie in the fact that though these large sperms take no part in the direct control of the heredity of the species, they may nevertheless be important in its welfare. Since they are normal in every respect except size, it is more than likely that they enter the egg like the normal sperms. It also must be recalled that polyspermy is almost universal in the fertilization of insect eggs and that in many species the number of supernumerary sperms is very high. The en- trance of genetically inert sperms would probably offer no difficulties in the regular fertilization process and a union between two normal pronuclei would proceed as usual. It is clear from evidence in marine eggs that the breakdown of gamete nuclei releases substances which have a far reaching influence on the reactions in the egg. It is likely that such substances are present in abnormally large quantities in the giant sperms and that therefore their possibly beneficial effects are increased. Again, if the developing embryo utilizes the nucleo-proteins that are brought in by BRACHYSTETHUS RUBROMACULATUS DALLAS 31 supernumerary sperms, an advantage might well accrue to a species which has such substances available in the unusually large amounts that are represented in the giant sperms. In short, the latter may play a not inconsiderable role in the em- bryology. If that is the case, the harlequin lobe would in no sense be a useless organ. The development of such a lobe in the testes of two genera that are taxonomically as diverse as Brachystethus and Loxa may well have been due to a nongenetic but metabolic advantage that is thereby conferred on the developing embryos. SUMMARY 1. The fourth lobe in all testes of the pentatomid Brachystethus rubromaculatus shows an aberrant meiosis of a very definite character. 2. The first indication of this aberrancy is found in a chain formation and clumping of the autosomal tetrads just prior to metaphase. The two sex chromo- somes are not included in this aggregation of autosomes. 3. At metaphase the X and the Y take a normal position in the spindle but the aggregate of autosomes is shunted laterally out of the middle region of the cell, away from the polar axis. 4. The two sex chromosomes divide equationally in the first division while the autosomal aggregate passes unaltered to either pole. 5. In the second division the X and Y again behave as in a normal meiosis and separate to opposite poles. The autosomal aggregate, if present, passes to either pole apparently at random. 6. Since the clumped autosomal tetrads pass undivided through both divisions while the sex chromosomes behave normally, the following four main types of spermatids are produced : X ; Y ; X -f- autosomes ; Y -f- autosomes. 7. The mechanics and evolution of so constant an aberrancy are discussed. LITERATURE CITED BALTZER, F., 1910. Ueber die Beziehung zwischen dem Chromatin und der Entwicklung und Vererbungsrichtung bei Echinodermenbastarden. Arch. Zcllf., 5 : 498-612. COOPER, K. W., 1944. Analysis of meiotic pairing in Olfersia and consideration of the re- ciprocal chiasmata hypothesis of sex chromosome conjunction in male Drosophila. Genetics, 29: 537-568. HucHES-ScHRADER, S., 1935. The chromosome cycle of Phenacoccus (Coccidae). Biol. Bull., 69: 462-468. POLLISTER, A. W., AND P. F. PoLListER, 1943. The relation between centriole and centromere in atypical spermatogenesis of viviparid snails. Ann. N. Y. Acad. Sci., 45: 1-48. SCHRADER, F., 1931. The chromosome cycle of Protortonia primitiva (Coccidae) and a con- sideration of the meiotic division apparatus in the male. Z. i^'iss. Zoo!.. 134: 149-179. SCHRADER, F., 1941a. The spermatogenesis of the earwig Anisolabis maritima Bon. with refer- ence to the mechanism of chromosomal movement. Jour. Morph., 68: 123-148. SCHRADER, F., 1941b. Chromatin bridges and irregularity of mitotic coordination in the penta- tomid Peromatus notatus Am. and Serv. Biol. Bull., 81 : 149-162. SCHRADER, F., 1945a. Regular occurrence of heteroploidy in a group of Pentatomidae (Hemip- tera). Biol. Bull., 88: 63-70. SCHRADER, F., 1945b. The cytology of regular heteroploidy in the genus Loxa ( Pentatomidae- Hemiptera). Jour. Morph., 76: 157-177. TROEDSSON, P. H., 1944. The behavior of the compound sex chromosomes in the males of cer- tain Hemiptera Heteroptera. Jour. Morph., 75 : 103-147. AN INVESTIGATION OF CROSS STRIATIONS AND MYOSIN FILAMENTS IN MUSCLE* C. E. HALL, M. A. JAKUS, AND F. O. SCHMITT Department of Biology, Massachusetts Institute of Technology, Cambridge, Massach itsctts INTRODUCTION Present information on muscle structure is of two rather distinct types, that derived from histological studies through direct observation with the light micro- scope, and that deduced from indirect methods such as involve polarized light micros- copy, streaming double refraction of myosin solutions or x-ray diffraction. His- tological studies have always been limited by the inadequate resolving power of light microscopes ; the indirect methods have been limited by the fact that they can supply information only in proportion to the correctness of the necessary assump- tions. Although valuable knowledge has resulted from both types of approach, there is a lack of definite information on structures below the limit of light micro- scope resolution. The electron microscope provides the possibility of extending direct observation into a range of dimensions reached hitherto only by indirect methods and conjecture. Since success in biological electron microscopy depends to a large extent on the development of techniques suitable for the material, these initial observations may be expanded as new techniques are developed. It is impossible in the present space to summarize the previous research on muscle or the ideas put forth concerning its structure ; reference to the literature will be limited to those papers which provide a basis for the present study. A gen- eral review has been presented by Fenn (1945) while the histological aspects have been reviewed by Jordan (1933). Polarized light investigations have been re- viewed by Schmidt (1937) and by Weber (1934). Results of investigations of streaming double refraction in myosin solutions may be found in the papers of Edsall (1930, 1942), von Muralt and Edsall (1930), Mehl (1938) and Dainty, Kleinzeller, Lawrence, Miall, Needham, Needham and Shen (1944). Astbury and Dickinson (1940) and Astbury (1942) described wide-angle x-ray diagrams of muscle and myosin, and Bear (1944, 1945) described small-angle x-ray patterns of various muscles. A discussion of muscle physics has been presented in a review by Ramsey (1944). In the past, the electron microscope has been used very little in the study of muscle structure. Richards, Anderson, and Hance (1942) have shown electron micrographs of a wedge-cut section of cockroach striated muscle in which each darker band can be seen to consist of three components. Sections of muscle tissue cut by another method were examined by Sjostrand (1943) but the electron micro- graphs bear no obvious relation to known structures in muscle. In both these stud- * The substance of this paper was presented at the meetings of the Electron Microscope Society of America on December 1, 1945. 32 STRIATIONS AND MYOSIN IN MUSCLE 33 ies the object was primarily to investigate sectioning techniques for electron micros- copy rather than to study muscle itself. Ardenne and Weber (1941) published electron micrographs of filamentous myosin particles, which are undoubtedly to be identified with the asymmetrical par- ticles responsible for the streaming birefringence phenomena in myosin sols. Hall, Jakus, and Schmitt (1945) have studied another fibrous protein from molluscan muscles and correlated the structure with the x-ray diffraction results described by Bear (1944). ELECTRON MICROSCOPE OBSERVATIONS OF MYOFIBRILS Preparation of specimens A primary difficulty in the study of striated muscle with the electron microscope is the preparation of the tissue thin enough to be partially transparent to the elec- tron beam. In view of the difficulties connected with present microtome techniques, a preparative method was sought in another direction. One procedure which has been used previously depends on the tendency of many materials to fragment along natural cleavage boundaries. If such tendency to natural cleavage exists, it may be assisted by chemical or physical means or both. Thus collagenous tissue may be separated into fine fibrils by mechanical "teasing" and the separation is facilitated1 by weak acid (Schmitt, Hall, and Jakus, 1942). Such methods of fragmentation are applicable to many materials and are readily adapted to striated muscle. Fresh muscle is fixed in 10 per cent formalin, cut into small pieces and subjected to mechanical agitation in a Waring Blendor. When the resulting suspension is lightly centrifuged to throw down the larger fragments, the supernatant exhibits the characteristic sheen associated with fibrous suspensions. It consists mostly of thin fibrils which may be washed and applied directly to a conventional electron microscope specimen grid with collodion membrane. These fibrils are to be identi- fied with the myofibrils, or sarcostyles, of the muscle fiber. Most of the observations to be described were made on fibrils from the leg muscles of frog and rabbit. The muscles of lobster and scallop (striated portion) were found to contain relatively large fibrils which were usually quite opaque in the electron microscope. It was necessary to apply fixatives to all these muscles in order to obtain intact fibrils. Wing muscles of the fly, on the other hand, could be teased apart in weak saline solution without previous fixation. Most myofibrils from frog muscle have widths between 0.5 and 1.0 /x, although a few may be found as wide as 3.0/x. or as narrow as 0.2^. They appear to be ribbon-shaped on the electron microscope specimen holder but it is not possible to say whether this shape results from forces produced during drying or whether it represents the form of the myofibril in the intact tissue. Although most of the structural features of the myofibril can be observed with- out the use of stain, contrast in the image may be increased by the application of phosphotungstic acid to the specimen (0.1 per cent solution at pH 3-5 for about one minute). The staining procedure is found to be particularly useful in increas- ing the contrast of the myosin filaments. An RCA electron microscope Type B with accelerating voltage raised to 65 KV was used throughout the study. A higher voltage would probably be advantageous for this type of material but is not readily obtainable from the standard power supply. 34 C. E. HALL, M. A. JAKUS, AND K. O. SCHMITT Unstained fibrils Unstained myofibrils from striated muscle are characterized by a succession of transverse bands of varying density, the main features of which can readily be identified with the bands previously described in histological studies (Jordan, 1933). An electron micrograph of a typical fibril from a slightly stretched frog sartorius is shown in Figure 1, together with the histological designations. The repeating -unit, or sarcomere, is bounded terminally by a narrow dense band usually referred to as the Z membrane or telophragma. This band is in the center of a light region, the / (or /) band, so called because it is relatively isotropic. Contiguous with 7 is the sharply-defined A (or Q) band which is optically anisotropic and has a higher scat- tering power for the electron beam than does /. In some myofibrils the H disc (or median disc of Hensen) appears as a lighter region in the middle of the A band (Fig. 1). Bisecting the A, or the H disc if it is present, is the narrow dark M band (or mesophragma). There may also appear, in either half of the I band, a relatively dark band designated as N in Figure 2. These bands are not always present and have been considered by some cytologists to be artifacts resulting from the lining up of granules. In the present study they have been noted in rabbit muscles but are either faint or absent in frog muscles. The clarity of their appear- ance in rabbit fibrils leaves no doubt as to their existence in this formalin-fixed material. Myofibrils from rabbit muscle usually show two to four fine cross bands sym- metrically disposed with respect to the M band and near it. They are beyond light microscope resolution, being about 0.1 ^ from center to center, and represent a periodic variation in the dense material of the A band. In favorable cases fibrils can be seen to consist of longitudinal myosin filaments which are difficult to resolve in unstained fibrils except where they fray out at the edges. They are visible in various parts of Figures 1 and 2. Stained fibrils Phosphotungstic acid combines with the cross bands roughly in proportion to their intrinsic density and serves to increase the contrast of the structures described above. Segments of stretched myofibrils stained with phosphotungstic acid are shown in Figure 3. The A band is stained more heavily than is the / band, and the M and Z bands appear quite opaque. When the H disc is present, it absorbs less stain than does the remainder of A. Besides accentuating the transverse bands, phosphotungstic acid enhances the contrast of the longitudinal myosin filaments. These range in width from 50 to 250 A and extend continuously, and in relatively straight lines, through both A and / bands. Although the filaments are usually indistinguishable within the dense Z bands, they can be traced through several successive sarcomeres when the Z bands are partially disintegrated. In the A band the filaments are relatively dense, sharply defined, and almost parallel in orientation. In the 7 band they are less per- fectly aligned and of lower density. The higher density of the A band appears to result from the higher density of the component filaments. FIGURE 1. Myofibrils from frog sartorius, stretched about 30 per cent, unstained. X 25,000. All myofibrils shown in this and subsequent figures were fixed in 10 per cent formalin. FIGURE 2. Myofibrils from rabbit leg muscle, unstained. X 25,000. STRIATIONS AND MYOSIN IN MUSCLE PLATE I 35 N 2b 36 Q K. HALL, M. A. JAKL'S, AND F. O. SCHMITT In both bands, the myosin filaments present a knotted or beaded appearance with Irequent constrictions and variations in density. The nodes are often almost equidistant along the filament and, in some regions of the fibril, they may be aligned laterally to produce a line cross striation with a period of about 400 A in the direc- tion of the fibril axis. However, this feature has not been found sufficiently repro- ducible or regular to be designated as a periodic spacing in the nature of those found in collagen and in clam muscle fibrils. The higher density of the ./ band in electron micrographs of unstained fibrils is in accord with the fact that the refractive index of A is higher than that of /. Since the myosin filaments pass continuously through both bands, it may be con- cluded that some other substances occur in much higher concentration in A than in /. This conclusion is in agreement with the observations of Scott ( 1932). who found a higher concentration of salts in the A band, and of Macallum (1905), who dem- onstrated a distribution of potassium within the A band which is remarkably like the distribution of unstained density as seen in Figure 1. Thus it seems quite evident that the A band contains, besides myosin filaments, some substance char- acterized by a relatively high concentration of salts. This material will be desig- nated as the ".-/ substance". Phosphotungstic acid is apparently absorbed in pro- portion to the density of the A substance. It should be noted that phosphotungstic acid forms an insoluble complex with potassium and has been used in the quantita- tive determination of potassium by Kieben and Van Slyke (1944). This does not mean that the phosphotungstic acid locates potassium specificallv for it may react with numerous substances; nevertheless, the observed absorption of this stain in the A band is consistent with the conclusion that the A substance does contain a rela- tively high concentration of potassium. That the A substance is closely associated with the myosin filaments is indicated by the lack of any observable quantity of interfilamentary material in frayed A bands. The sharp boundary between the A and / bands is noteworthy since there is no apparent membrane or other structure to confine the A substance. Owing to the limited resolution of the light microscope it had been concluded (Schmidt. 1937) that the transition between the A and / bands is not abrupt but gradual. Likewise there is no evidence in electron micrographs for any envelope or limiting membrane around the myofibril. One <>f the most prominent histological features of striated muscle is the Z membrane, the nature of which has been the subject of much discussion in the past. In electron micrographs the 7. membrane appears to be amorphous material of high ^taining affinity which cements the myosin filaments together in this region. Fre- quentlv in frayed myofibrils the filaments separate laterally but adhere to one an- other in the region of 7. and also at M. Furthermore, there is a tendency lor the filaments to break at Z and sometimes at M. It has been postulated that the 7. membrane takes the form of an annular ring about the fibril (see Liang. 1936). Although this could be true, there is no unequivocal electron microscope evidence for such a conclusion. The Z membrane appears to consist of interfilamentary material present throughout the fibril and not limited to the periphery. It is defi- nitely not collagenous as was suggested by Iliiggqvist (1931 ). FH.I KI. .1. Myofibrils from frog sartorius. Wretched about 30 per cent, stained with phosphotungstic arid; (r) shows collation fibrils. X 40,000. STRIATIONS AND MYOSIN IN MUSCLE 37 PLATE II 3c C. E. HALL, M. A. JAKUS. AND F. O. SCHMITT Collagen In muscle In tin- preparation of specimens no effort was made to separate the collagen, which occurs in skeletal muscle in appreciable quantity. However, there is no dif- ficulty in identifying collagen since it has a regular spacing of about 640 A and displays a characteristic fine structure after staining with phosphotungstic acid ( Schmitt, Hall, and Jakus, 1945). Several stained collagen fibrils are visible in Figure 3c. It has been observed that collagen fibrils from rabbit and frog muscle are quite uniform in width (about 500 A), but the relation of this protein to the muscle structure is not evident in present electron micrographs because of the ex- tensive fragmentation of the muscle. Structural alterations in c.rtcnsion and contraction Although evidence is not yet sufficient to permit a complete description of the structural alterations associated with extension and contraction, some observations have been made and ma}- be described. Extended sarcomeres were obtained by stretching an excised Irog sartorious and fixing it at the stretched length. In some instances the muscle was stimulated electrically during fixation to produce isometric contraction. In the absence of the electrical stimulus, the fixative itself provided a weakly stimulating effect. In other experiments, excised muscles were placed in formalin without restraint, thus producing a state of weak isotonic contraction. Strong isotonic contraction was produced by electrical stimulation ol excised muscles before and during fixation. Myofibrils from these muscles were prepared for elec- tron microscope observation in the manner described. The high concentration of , / substance toward the ends of the .1 band, resulting in the appearance of the H disc, occurs consistently in myofibrils from muscles which have been stimulated (electrically or by the fixative) while held in a state of exten- sion ( Figs. 1 and 3). Muscles which are free from tension and are stimulated to contract during fixation do not in general show a division of the A band. Myo- fibrils from such muscles are shown in Figure 2. In the myolibrils of muscles which were stretched up to 130 per cent of rest length and fixed at this length, the A bands are about the same length as those in muscles which had contracted slightly as a result of fixation. Thus the increase in sarcomere length in stretched and fixed muscles is due mainly to an increase in the length of the / band. This observed relative constancy in the length of the A band during extension is in agreement with the findings of Buchtal, Knappeis, and Lind- hard ( 1936) who concluded that the / band has a relatively low modulus of elasticity. Electron micrographs of myofibrils from muscles fixed in a state of strong con- traction show sarcomere lengths down to about 1 /t as compared with about 2 /JL in relaxed or weakly contracted muscles. In general the observed sarcomere structure is of two types. The first type, shown in Figure 4, consists of a very narrow / band and a quite solid A band in which the median // disc is absent or very faint. The sarcomere length in Figure 4 is about 1.5 /<„ which is about 25 per cent less than the average length in relaxed muscles. The second type of sarcomere structure K 4. Myofihril from frog sartorius contracted by electrical stimulation, stained \vitli phosphotungstic arid. Y- 25,000. Fi<;ri case are about \ .2 p., which is about 60 per cent of the average relaxed length. The / band is no longer visible and 7. appears somewhat wider and more poorly defined than in less strongly contracted fibrils. The region between M and Z is of uniform density and presumably contains a uniform concentration of A substance. In occasional fibrils the 7. and .17 are distinguishable only with difficulty. Such fibrils may represent a transition stage between the two types of fibrils described. It this is the case, the second type of fibril might result from the accumulation, around the 7., of A substance which had migrated away from M. This stage ap- parently corresponds to the striation reversal described by Jordan (1933) and earlier by Rollett (1891 ). who refers to this densely staining Z as the C, or con- traction band. One aspect of particular significance in strongly contracted sarcomeres is the relative straightness of the myosin filaments, as seen in Figure 5. Since the con- tracted sarcomeres are as little as 50 per cent of the relaxed length, this can only mean that the filaments themselves shorten in contraction. Similarly, in extension the fila- ments must individually lengthen. Changes in sarcomere length are not to be associated with any gross spiralling or folding of filaments in the order of dimen- sions visible in the electron microscope. The myosin filaments are the smallest visible contractile units. ISOLATED MYOSIN FILAMENTS Myosin may be extracted from muscle in weakly alkaline salt solutions and these extracts have been shown to contain filaments visible in the electron microscope (Ardenne and Weber, 1941 ). In order to determine whether visible differences exist between different myosins, a quantitative study of filaments from various muscles was undertaken. Extracts of myosin were prepared by a method essentially similar to that de- scribed by Greenstein and Edsall (1940). Muscles were removed from the animal immediately after death, trimmed, cut into small pieces and blended in a Waring Blendor with about ten parts of a cold solution of KC1 (0.5 M) and NaHCCX (0.03 M). The suspension was stirred mechanically, at about 4° C., for periods of time varying from seven to twenty-four hours and strained through several layers of closely-woven cheese-cloth. The filtrate, showing strong double refraction of flow, was poured into 8 to 10 volumes of cold distilled water, with constant stirring, and the precipitate which formed was allowed to settle overnight in the cold. Further concentration was accomplished by centrifugation, after removal of the supernatant. The precipitated myosin was washed with cold distilled water, and redissolved by adding powdered KG crystals to a concentration of about 0.45 M. To further purify the mvosin the precipitation and solution was repeated once or twice. For electron microscope examination, the myosin preparation was diluted to the desired concentration with the KC'l-Xal !('(), solution used in the extraction and centrifuged to remove any undissolved protein. A droplet of the solution was !•"[(, TKK n. Myosin filaments from rabbit leg muscle. •; 30,000. I- 'n.i i'i 7. Myosin filaments from lobster abdominal muscle. X 30,000. I-'K.I KI X. Myosin filaments from clam ( Mya arenaria) adductor muscle. X 30,000. STRIATIONS AND MYOSIN IN MUSCLE PLATE IV 41 42 C. E. HALL, M. A. JAKUS, AX I) V. O. SCHMITT placed on the supporting film of tin- specimen grid, the excess removed by blotting, and the film washed with Kdsall's solution. The adhering filaments were then fixed and stained, for about one minute, with 0.1 per cent phospho-12-tungstic acid (pH 3 to 5) and washed briefly with water to remove the tmcombined phosphotungstic acid. Unstained filaments .are poorly defined because of low contrast and, in the absence of the stabilizing effect of the heavy metal ion, appear to be adversely affected by the drying process. Myosin extracts were made from the muscles of rabbit leg, frog leg, lobster abdomen, scallop adductor (striated part) and clam (Mya) adductor (classified as smooth muscle). Electron micrographs of such myosin preparations show slender filaments similar to those observed in intact myofibrils and resembling those shown by Ardenne and Weber (1941). The filaments have varying lengths and widths and display no tendency to branch or split longitudinally. They are rough in ap- pearance and show fluctuations in density along their length. However, there is no evidence of any regular structural variation which could be interpreted as a significant periodicity. There are no observable changes in density which can lie correlated with the A and / bands, which indicates that the A substance has prob- ably been washed out during extraction. This general appearance is very much the same for myosin filaments from the five different animal forms. Typical elec- tron micrographs of myosin filaments from rabbit, lobster, and clam are shown in Figures 6, 7, and 8 respectively. A statistical study of lengths and widths of myosin filaments from the five selected forms was made. Measurements were taken directly from enlarged prints and in any given print all filaments were measured which could be discerned as complete and individual. Widths can be measured only approximately because of the smallness of this dimension and also because of the roughness of contour. The uncertainty of measurement, however, is significantly smaller than the spread in widths. Results are shown in Figure 9. Each interval in the plot of widths con- tains the same number of measurable increments and each set of measurements represents about 300 filaments. TABLE I Dimensions of myosin filaments Mn« lc source Average width Average length Rabbit lesj 120 A 4100 A Frog le- 140 A 4100 A Lobster abdomen 140 A 6800 A Scallop adductor, striated 130 A 5000 A Clam (Mya) adductor, smooth 150 A 3100 A iXearlv all measured widths fall into a narrow range between 50 and 300 A. There may be some filaments having widths below 50 A which were not observed because of the resolution and contrast limitations of the electron microscope. However, the quality of the background in the electron micrographs and the lact that all distributions fall off toward small widths indicate that no great quantity of the myosin occurs in this range in electron microscope specimens as prepared. STRIATIONS AND MYOSIN IN MUSCLE 43 Between one animal form and another there is very little variation in magnitudes or distribution of widths. There is a slight difference in the averages as shown in Table I, hut it is so small as to be scarcely significant. Filament widths were also measured from electron micrographs of intact myo- fibrils from frog muscles. The distribution plot resembles very closely the corre- 60 40 20 60 40 T >- o o 111 20 60 4O 20 60 40 20 60 40 20 LOBSTER PECTEN (STRIATED) MYA ADDUCTOR O 100 200 30O WIDTHS IN & o apoo opoo ispoo 20,000 2spoo sopoo 35000 LENGTHS IN & FIGURE 9. Distribution of lengths and widths of myosin filaments from muscles of various animal forms. 44 Q E. HALL. M. A. JAKUS, AND F. O. SCHMITT >ponding plot for frog shown in Figure (>. thus indicating that there is no great difference in the width of filaments in electron micrographs of intact fibrils and mvoxin extracts. In comparison to the fair constancy of widths there is a wide distribution of lengths. Must of the measured lengths are below 15. 000 A. Filaments from lobster muscle are in general significantly longer than the others, while filaments from Mya aductor, which is the only smooth muscle studied, are significantly shorter. As a result of the wide range of lengths there is only rough significance in the number-averages given in Table 1. The distribution plots are reasonably reproducible under similar conditions. It is evident from the statistical results that myosin extracts of this type do not constitute a monodisperse system and there is no justification for referring to the filaments as molecules. Widths of filaments apparently do not alter appreciably during extraction. The lengths, however, bear no relation to any observable dimensions in the intact myofibrils. Since most of the lengths represented in the plots of Figure 9 are considerably shorter than are the corresponding sarcomeres, it is apparent that the myosin filaments have been broken at random into shorter segments during the extraction procedure. Since the myosin filaments seen with the electron microscope are to be identified with the asymmetrical particles responsible for the streaming double refraction of myosin solutions, it is pertinent to compare the results of the methods where possi- ble. Mehl (1938) reported the length of rabbit myosin as 8,500 A while Edsall (1942) gives a figure of 12.600A . The second figure is near the upper limit of the distribution plot for rabbit in Figure 9 while the first falls within the range of lengths representing the bulk of the protein. Although no figures have been re- ported on the other muscles used here, Edsall and Mehl ( 1940) have described lobster myosin as being more viscous than rabbit myosin at equal concentration and strongly birefringent, which is in qualitative agreement with the electron micro- scope observation that lobster preparations contain significantly longer fibrils than do those from rabbit. Ardenne and Weber ( 1941 ) do not give the source of the myosin extract from which they made electron micrographs. They state that the filament widths are from 50 to 100 A although no measurements are tabulated and the micrographs used as illustration contain filaments at least 200 A in width. The myosin fila- ments within the limited field of the micrographs which these authors have shown are apparently similar to those described here. Paramyosin In a recent publication llall. |akus, and Schmitt (1945) described the structure of a fibrous muscle globulin which is present in appreciable amount in molluscan smooth muscles, and correlated the observed structure with the x-ray patterns obtained bv Rear (1944) from the same material. Since this protein can be iden- tified by electron microscope observation and x-ray diffraction it merits a distin- guishing name and is therefore designated as paramyosin. Although paramyosin I-IMKK 10. Regular structures from clam adductor muscle after dispersion in Edsall's solution. X 40,000. I'K.ruK 11. Fragments from dispersed rlam muscle. Darkly staining particles are adher- ing to tine filaments, presumably myosin. X 40,000. STRIATIONS AND MYOSIN IN MUSCLE PLATE V 45 •>., 10 €* •• .* r 46 C. E. HALL, M. A. JAKUS, \XI> F. O. SCHMITT resembles inytisin in sonic respects, there are definite differences between the two proteins. Paramyosin contains a characteristic axis spacing of about 145 A. Para- nivosin has not been detected in striated muscle while nivosin is apparently common to all muscles. Furthermore, the typical needle-shaped fibrils of paramyosin dis- integrate in Edsall's solution (0.5 M" KC1, 0.03 M NaHCO,) while the myosin fila- ments remain essentially the same as in fixed intact muscle except as they mav be broken into shorter lengths during extraction. ( 'lain adductor muscles The adductor muscle of the clam differs from the other muscles investigated in that it is classified histologically as "smooth" muscle and contains a large quantity of the fibrous protein, paramyosin. Electron microscope observation has failed to reveal any large periodicity resembling the characteristically banded sarcomere of striated muscle. However, if clam muscle is dispersed in Edsall's solution and centrifuged. there are thrown down fibrous aggregates to which large particles adhere. A typical micrograph of this type of material is shown in Figure 10. The structure consists essentially of bundles of fine filaments (presumably myosin) and darkly staining nodules producing a cross striation with a period of about 1.100 A. In Figure 11 the bundles have been dispersed, revealing the individual dark nodules which adhere to the fine filaments at more or less regular intervals. The attachment of the dense component at regular intervals along nivosin bundles (Fig. 10) is suggestive of a rudimentary structure analogous to the / membrane and sarcomere of striated muscle. As yet. the significance of this structure is m it km >wn. DISCUSSION With respect to the existence and disposition of the cross striations, the electron microscope observations on fixed muscle, stained and unstained, are in close agree- ment with the results obtained with the light microscope (see Jordan. 1933). Con- firmatory evidence is presented concerning the A7 bands, H discs and the nature of ">triation reversal" in strong contraction. Additional information has been added by virtue of the high resolution of the electron microscope. Thus it is possible to say definitely that the 7. band is not collagenous as suggested by Haggqvist (1931) and to observe directly the myosin filaments and their relation to previously de- termined structures. ( >f fundamental importance is the observation that the myosin filaments extend continuously through the fibril in relatively parallel straight lines. There is no marked disorientation of the filaments in either ./ or / bands of fibrils in any state of contraction. Although the optical anisotropy oi the A band led to the postulate that it contains asymmetric myosin particles, there has always been some doubt as to whether the relatively isotropic / band consists of mvosin in nnoriented state or ot -Mine other protein ( Weber, 1934). It is now clear that the / band consists mainly of well oriented myosin filaments and it is not possible to account for the low birefringence on the basis of gross disorientation. I 'ossibly the difference in anisotropy between the A and / bands is to be attributed to differences in orienta- tion within the filaments. The only obvious alternative explanation is that the low birefringence in the / band is due to partial compensation of the birefringence of the myosin filaments by other components. STRIATIONS AND MYOSIN IN MUSCLE 47 No evidence has been found in this investigation for the existence, either in myofibrils or in extracted myosin, of "rodlets" of the specific dimensions postu- lated by Weber (1934) from polarized light and diffusion experiments. Theo- retical difficulties underlying Weber's calculations have been pointed out by Frey- Wyssling (1940) and Schmitt (1944). Electron micrographs of isolated myosin filaments show that in width and gen- eral appearance the filaments are similar to those seen in fixed intact myofibrils. Although myosin filaments from whatever source are essentially the same, the statistical study indicates that they occur in various lengths and are not to be designated as discrete "myosin molecules". However, the length of rabbit myosin reported by Mehl (1938) from streaming double refraction studies is reasonably close to the weight-average length calculated from the distribution curve in Figure 9. No reliable measurement of width is available for comparison with the electron microscope results. Since myosin filaments in intact fibrils are continuous, the wide range of lengths found in myosin suspensions is noteworthy. It appears that the filaments, while little changed in width, are broken more or less at random during the extraction procedure. The longest filaments approach the sarcomere length and there may be significance in the fact that lobster muscle yields longer filaments than does frog muscle and also has the longer sarcomere. The absence of filaments longer than their corresponding sarcomere is consistent with the observed tendency of the filaments to break at Z. Ziff and Moore (1944), following an extraction procedure similar to that used here, state that their myosin solutions contain a homogeneous substance which forms sharp boundaries in electrophoresis and sedimentation. If homogeneity is meant to denote constancy in particle length, this conclusion is in disagreement with the electron microscope results. In considering the apparent inconsistency it should be noted that, in the electrophoresis and sedimentation of rod-shaped particles of nearly constant diameter, a sharp boundary does not necessarily indicate a constant length. Schramm and Weber (1942) reported that a small fraction of the myosin in extracts has a much higher sedimentation constant than that of the predominant component. No evidence for such a distinct heavy component has been found in the present study unless it represents aggregates of myosin filaments. The "molecular weight" of myosin has been estimated by Weber and Stover (1933) to be 0.6 - 1.2 X 10° and by Ziff and Moore (1944) to be 3.9 X 10°. If a particle weight is calculated from the average dimensions of the filaments in rabbit myosin from Table I, using a density of 1.3 and assuming a circular cross section, the result is 36 X 10G. The results differ by at least a whole order of magnitude. In view of the fair agreement as to filament lengths from streaming birefringence and electron microscope observations, the discrepancy in particle weight is difficult to understand. It may be that the cross section of filaments dried on the supporting film is not circular as assumed in the electron microscope calcula- tion, but it is doubtful whether this assumption could introduce an error large enough to account for the discrepancy. In any event, the significance of all such calculations is questionable inasmuch as the extracts contain particles of widely differing dimensions, in no sense to be considered as molecular entities. Bear has reported small-angle x-ray diffractions from various muscles. One set of diffractions (Bear, 1944) has been correlated with the structure of para- 48 C. K. HALL. M. A. JAKUS, AND F. O. SCHMITT myosin fibrils, as determined from electron micrographs (Hall, Jakus, and Schmitt, 1945). This set of diffractions has not been obtained from any striated muscle. In addition Bear (1945) has reported a second set of small-angle diffractions obtained from a variety of muscles, both smooth and striated. The wide occurrence of this pattern is strongly suggestive that the diffractions originate in the myosin component but it has not been possible to identify in the electron microscope the structure responsible for the diffractions. Furthermore, since the meridional dif- fractions are orders of 27 A, it is doubtful whether the structure can be observed directly. Another feature of the diffractions is the occurrence of a periodicity estimated by Bear to be between 350 and 420 A. Although this dimension is quite large enough for electron microscope resolution, it appears that the pattern is some- what like the paramyosin pattern in that the large periodicity cannot be discerned unless the axis spacing (27 A) is well resolved. The fine banded appearance fre- quently observed in myofibrils is of about the same order of magnitude as the large periodicity, but this may be fortuitous. The x-ray data also indicate a lateral periodicity of about 115 A which should be large enough for electron miscroscope observation. Although the average width of myosin filaments is quite close to this figure, the significance, if any, of this coincidence remains to be determined. Since the myosin filaments are observed to pursue a straight course through the A and I bands in fibrils from contracted as well as relaxed muscles, it may be con- cluded that contractility is a property of the individual filaments in their normal environment in muscle. It seems probable that alterations in length and tension depend on changes within the filaments in response to changes in the chemical environment. In seeking a description of the contractile mechanism, due considera- tion must be given to the role played by the A substance and the adenosine triphos- phate. The electron microscope technique provides a promising method for study- ing these structures and processes. It may be expected that further correlation of the x-ray, electron microscope, polarization optical and chemical evidence will con- tribute greatly to an understanding of the nature of contractility. SUMMARY 1. Electron micrographs were made of myofibrils isolated from frog and other skeletal muscles fixed in formalin. The structure with respect to the location and disposition of the principal cross striations is in good agreement with that previ- ously determined from histological studies. 2. The myofibrils are composed of bundles of myosin filaments ranging in width from about 50 to 250 A and extending continuously and in relatively straight lines through the isotropic and anisotropic bands in both the extended and contracted states. The anisotropic bands also contain material of high electron scattering power and affinity for phosphotungstic acid. The distribution of this "A substance" changes with contraction in characteristic fashion. The evidence indicates that the myosin filaments are the contractile units. 3. While the myosin filaments have an indefinite length in the intact fibril, they are fragmented extensively during extraction in weakly alkaline salt solutions (method of Greenstein and Edsall). Filaments from such extracts have fairly uni- form widths (50 to 250 A) but highly variable lengths, in general below 15,000 A. Filaments from rabbit, frog, lobster, scallop and clam muscles are similar in appear- STRIATIONS AND MYOSIN IN MUSCLE 49 ance; widths are fairly uniform but lengths vary significantly from one form to another. 4. The relation of these findings to physical chemical data previously obtained by others on similar myosin extracts is discussed. 5. In the one smooth muscle examined (clam adductor) no striations comparable to those of skeletal muscle fibrils were found. However, a regular structure with a period of about 1,100 A was observed. LITERATURE CITED ARDENNE, M. VON, AND H. H. WEBER, 1941. Elektronenmikroskopische Untersuchung des Muskeleiweisskorpers Myosin. Kolloid-Z., 97 : 322-325. ASTBURY, W. T., 1942. X-rays and the stoichiomctry of the proteins, with special reference to the structure of the keratin-myosin group. Jour. Chcm. Soc., 337-347. ASTBURY, W. T., and S. DICKINSON, 1940. X-ray studies of the molecular structure of myosin. Proc. Roy. Soc., Ser. B, 129 : 307-332. BEAR, R. S., 1944. X-ray diffraction studies on protein fibers. II. Feather rachis, porcupine quill tip and clam muscle. J our. Am. Chcm. Soc., 66 : 2043-2050. BEAR, R. S., 1945. Small-angle x-ray diffraction studies on muscle. Jour. Am. Chcm. Soc., 67: 1625-1626. BUCHTAL, F., G. G. KNAPPEIS AND J. LINDHARD, 1936. Die Struktur der quergestreiften, lebenden Muskelfaser des Frosches in Ruhe und wahrend der Kontraktion. Skand. Arch. f. Physiol., 73 : 163-198. DAINTY, M., A. KLEINZELLER, A. S. C. LAWRENCE, M. MIALL, J. NEEDHAM, D. NEEDHAM, AND S-C. SHEN, 1944. Studies on the anomalous viscosity and flow -birefringence of protein solutions. III. Changes in these properties of myosin solutions in relation to adenosinetriphosphate and muscular contraction. Jour. Gen. Physiol., 27 : 355-399. EDSALL, J. T., 1930. Studies in the physical chemistry of muscle globulin. II. On some physicochemical properties of muscle globulin (myosin). Jour. Biol. Chcm., 89: 289-313. EDSALL, J. T., 1942. Streaming birefringence and its relation to particle size and shape. Ad- vances in Colloid Science, 1 : 269-316. EDSALL, J. T., and J. W. MEHL, 1940. The effect of denaturing agents on myosin. II. Vis- cosity and double refraction of flow. Jour. Biol. Chcm., 133 : 409-429. FENN, W. O., 1945. Contractility. Physical Chemistry of Cells and Tissues. Blakiston, Philadelphia. 447-522. FREY-WYSSLING, A., 1940. Analyse der Formdoppelbrechungskurven. Kolloid-Z., 90 : 33-40. GREENSTEIN, J. P., AND J. T. EDSALL, 1940. The effect of denaturing agents on myosin. I. Sulfhydryl groups as estimated by porphyrindin titration. Jour. Biol. Chem., 133 : 397- 408. HAGGQVIST, G., 1931. Gewebe und Systeme der Muskulatur. Hdb. der mik. Anat. des Menschcn (Springer), 2: 142. HALL, C. E., M. A. JAKUS AND F. O. SCHMITT, 1945. The structure of certain muscle fibrils as revealed by the use of electron stains. Jour. App. Phys., 16 : 459-465. JORDAN, H. E., 1933. The structural changes in striped muscle during contraction. Physiol. Rev., 13 : 301-324. LIANG, T-Y., 1936. Histophysiologische Untersuchungen iiber die Beziehungen der Muskel- kontraktion zu Doppelbrechung und Querstreifung. Chinese Jour. Physiol., 10 : 327-354. MACALLUM, A. B., 1905. On the distribution of potassium in animal and vegetable cells. Jour. Physiol., 32 : 95-128. MEHL, J. W., 1938. Double refraction of flow of protein solutions. Cold Spring Harbor Symp. on Quant. Biol, 6 : 218-227. MURALT, A. L. VON, AND J. T. EDSALL, 1930. Studies in the physical chemistry of muscle globulin. III. The anisotropy of myosin and the angle of isocline. IV. The anisotropy of myosin and double refraction of flow. Jour. Biol. Chem., 89: 315-386. RAMSEY, R. W., 1944. Muscle: physics. Medical Physics. Year Book Publishers, Chicago. 784-798. 50 C. E. HALL, M. A. JAKUS, AND F. O. SCHMITT RICHARDS, A. G., JK., T. F. ANDERSON, AND R. T. HANCE, 1942. A microtome sectioning tech- nique for electron microscopy illustrated with sections of striated muscle. Proc. Soc. Exp. Biol. Veil., 51 : 148-152. RIEBEN, W. K., AND D. D. VAN SLYKE, 1944. Gravimetric determination of potassium as phospho-12-tungstate. Jour. Biol. Chan., 156: 765-776. ROLLETT, A., 1891. Ueber die Streifen N (Nebenscheiben), das Sarkoplasma und die Con- traction der quergestreiften Aluskelfasern. Arch. f. mik. Anat., 37 : 654-684. SCHMIDT, W. J., 1937. Die Doppclbrechung von Karyoplasma. Zytoplasma und Metaplasma. Borntrager, Berlin. SCHMITT, F. O., 1944. Structural proteins of cells and tissues. Advances in Protein Chcm., 1 : 25-68. SCHMITT, F. O., C. E. HALL, AND M. A. JAKUS, 1942. Electron microscope investigations of the structure of collagen. Jour. Cell. Comp. Physiol., 20: 11-33. SCHMITT, F. O., C. E. HALL, AND M. A. JAKUS, 1945. Fine structure in the fiber axis macro- period of collagen fibrils (Abstract). Jour. App. Phys., 16: 263. SCHRAMM, G., AND H. H. WEBER, 1942. liber monodisperse Myosinlosungcn. Kolloid-Z., 100: 242-247. SCOTT, G. H., 1932. Distribution of mineral ash in striated muscle cells. Proc. Soc. Exp. Biol. Med.,29: 349-351. SJOSTRAND, F., 1943. Electron microscopic examination of tissues. Nature, 151 : 725-726. WEBER, H. H., 1934. Die Muskeleiweisskorper und der Feinbau des Skeletmuskels. Ergcb. Physiol., 36 : 109-150. WEBER, H. H., AND R. STOVER, 1933. Das kolloidale Verhalten der Muskeleiweisskorper. Biochem. Z., 259 : 269-284. ZIFF, M., AND D. H. MOORE, 1944. Electrophoresis, sedimentation and adenosinetriphosphatase activity of myosin. Jour. Biol. Chcm., 153: 653-657. EFFECT OF WATER CURRENTS UPON THE ATTACHMENT AND GROWTH OF BARNACLES x F. G. WALTON SMITH University of Afiami Marine Laboratory - INTRODUCTION There exists a considerable body of published work dealing with peculiarities in the distribution of sessile marine invertebrates in general, and, although experimental evidence is incomplete some speculation has been offered as to the manner in which their distribution is limited by various factors of the environment which influence their attachment and growth. The object of the present studies was to investigate the action of the water current factor by experimental methods, with particular reference to the species of barnacles which are most abundant in the Miami area. Much of the earlier work on sessile marine animals has been reviewed by Me Dougall (1943), and need not be considered here in detail. A bibliography of .investigations dealing with sessile organisms from the point of view of ship fouling is given by Neu (1933), and an excellent general account of marine fouling by Visscher (1928) who does not, however, include more than a brief reference to the effect of water currents. The effects of environmental factors have been deduced from the results of eco- logical surveys by a number of investigators. Stephenson and his co-workers, in a series of publications entitled The South African Zone and Its Relation to Water Currents (see bibliography), Fischer-Piette (1928, 1928a), Pierron and Huang (1926), and Prennant and Teissier (1924) have described the populations of rocky shores and the distribution of sessile organisms in general, but with scanty reference only to the relation between barnacle attachment and the velocity of water currents. Stephenson and Bright (1938) and Stephenson and du Toit (1937), however, note the absence of certain species of barnacles from surfaces exposed to heavy wave action. The possibility of water movement influencing barnacle attachment is suggested by Fischer-Piette (1932) in his reports on surveys of the shores of the English Channel. Shelford and Towler (1925), Towler (1930), and Newcombe (1935) have paid particular attention to barnacles in associations of Balanus-Littorina and Balanus-Mytilus and draw attention to the more luxuriant growth of barnacles ex- posed to strong water currents, particularly in shallow water. The role of water currents is also mentioned briefly in accounts of ecological surveys by Rice (1930), Moore (1935a), and Moore and Kitching (1939), who stress the part played by 1 These experiments were conducted while the author was engaged by the Woods Hole Oceanographic Institution in investigations under contract with the Bureau of Ships, Navy Department, which has given permission to publish the results. The opinions contained herein are those of the author and do not necessarily reflect the official opinion of the Navy Department or of the Naval service at large. " Contribution number 301 from the Woods Hole Oceanographic Institution. 51 F. f;. WALTON SMITH them in the provision of an adequate food supply, as well as the adverse effects of wave action. Experimental investigation has yielded more direct information. Moore (1933) studied the effect of currents upon the attachment of cyprid larvae with particular regard to orientation, but gives no estimate of current velocities which limit attach- ment. The related phenomena of wave action are considered by Moore in a sepa- rate paper (1939). Evidence as to the effect of water currents is also afforded by the study of ships and other objects moving through the water. The economic importance of ship- bottom fouling has stimulated this kind of observation. Heutschel (1934) and Neu (1932) state that vessels collect barnacles most readily when in harbor and when subjected to relatively less water movement. Visscher (1930) further observes that barnacles are usually killed during ocean passages of about 500 miles. Hiraga (1934) showed experimentally that, in the laboratory, 10 to 17 day old barnacles were lost during a five day exposure of test planks to water currents, whereas twenty- four day old barnacles remained alive and intact. In a later paper of Visscher (1938) he concludes that barnacles are not distributed by ships. This is contrary to the views of Fischer-Piette (1929, 1935) who expresses the opinion that Balaims amphitrite found by him on the coast of France, was brought from the tropics on the hulls on ships. Later observations by Phelps ( 1942) on test panels support the conclusion that excessive water currents inhibit the attachment and growth of barnacles. The extensive observations of McDougall on the pile fauna at Beaufort include the results of exposing various types of experimental collecting apparatus. His experiments on the effect of water currents indicate that fewer barnacles settle and that growth is slower in more rapid currents. His results did not, however, give an accurate indication of limiting or optimum velocities to which his collectors were exposed. The experiments which form the 'subject of the present account are divided into a study of the effects of water currents upon the attachment of cyprid larvae, and into a consideration of the effects of water currents upon the organisms following attachment. Under natural conditions of fluctuating tidal currents and wind drifts it is conceivable that a cyprid might attach during a period of minimum water flow in a particular region and that, once attached, the maximum flow would be insuf- ficient to detach it or to inhibit its growth. The aim of the investigation was to establish the limiting velocities of current for initial and continued attachment and growth. The advice and suggestions of Dr. Alfred C. Redlield of the Woods I lolc ( Vran- ographic Institution are gratefully acknowledged. The writer is also indebted to Mr. Frank L. LaOue of the Development and Research Division, International Nickel Company, Inc. and to Dr. William F. Clapp for permission to include in this paper data from their observations at Kure Beach. N. C. Acknowledgments are also due Mr. D. S. Reynolds, Mr. James Gregg, Mr. Alexander Frue, and Mr. Charles Weiss for their assistance at various times. METHODS In order to study the effects of water currents two types of apparatus were de- signed which would permit the movement of sea-water relative to the experimental WATER CURRENTS AND GROWTH OF BARNACLES 53 surface at predetermined and variable velocities. The first consisted of a rotating disc, immersed in the sea. The second type of apparatus consisted of a si-ric-s «>{ glass tubes of varying cross-sectional diameter. The rotating disc was employed both in the study of initial attachment of barnacles and in the study of growth sub- sequent to attachment. The glass tubes of graded diameter were employed in the observations on initial attachment only, as a check against results obtained with the rotating disc. \\~ATKR CURRENTS AND ATTACHMENT ON THE ROTATING Disc The rotating disc apparatus as used in the experiments here described is essen- tially similar to a machine described by LaOue (1943) for the purpose of studying the effect of sea-water currents upon corrosion of metal samples, with certain changes in design appropriate to the particular use to which it was put. The modified machine consists of a vertical shaft rotated by means of an electric motor and a system of belts and pulleys. The disc is attached to the shaft by means of a flange and small brass screws and is maintained in a position several inches below low water spring tide. At velocities of rotation used in the experiment no cavitation occurred at this depth. By changing the arrangement of pulleys the speed of rotation may be varied up to approximately 1750 r.p.m. The rate of movement in knots of any portion of the disc relative to an imaginary stationary body of water is readily calculated as ap- proximately R X D/370, where R is the number of revolutions per minute and D the diameter in inches. This velocity is nominal, however, and does not accurately represent the rate of relative movement between the disc and the water close to its surface. Frictional drag and centrifugal forces produce vortex movements which cause water to flow in a close spiral roughly parallel to the plane of the disc. As a result, the actual flow of water at any part of the disc's surface is less than the nominal rate calculated from the velocity of rotation. Size of the discs used varied up to 28 inches in diameter. Experimental procedure consisted of bolting a new rotating disc on the shaft and setting the machinery in continuous operation. The disc was removed and examined at suitable intervals. At the same time a similar stationary disc was im- mersed at a point nearby the rotating disc, at the same depth and resting similarly in a horizontal plane. Observations on the second disc acted as a control for such factors as abundance of organisms, nature of the surface of the disc. etc. The apparatus was used at the edge of a covered slip at the Miami Beach Boat Corpora- tion, where fouling is usually severe. The predominant organism at most times of the year is Balanns aniphitritc niveus Darwin and this wras accordingly selected for observation. Results The results of the first rotating disc experiment indicated that the approximate minimum current velocity required to prevent barnacle attachment is 1.1 knots. The disc used was 13 inches in diameter and was rotated at 540 r.p.m. The disc, together with its stationary control, are shown in Figure 1 as they appeared at the end of this experiment. Five barnacles attached to the center of the rotated disc during 23 days but did not grow beyond 2 mm. in greatest width. The limiting velocity was calculated from the 34 inch diameter of the circle to which the barnacle attachments were restricted (Table I). 54 F. t rotation rotation attachment attachment start occurred 0( curred (Days) (r.p.m.) i Inclii-' (Knots) 1 B 8/20/43 23 540 34 1.1 4 A 11/4/43 16 192 2 1.0 o A 12/7/43 14 134 2y2 0.9 7 A 1/5/44 19 60 8 1.3 Average — — — 1.1 FlGi KI-: 1. Surfarr of disc rotated at 540 r.]).in. for a period of 23 days, tot-rtlu-r with stationary control. Diaiiu-trr of disc, 13 inches. WATER CURRENTS AND GROWTH OF BARNACLES 55 In order to establish the limiting velocity with greater accuracy the experiment was repeated on three other occasions at slower speeds of rotation, with the results shown in Table I and Figure 2. The nominal velocity limiting barnacle attachment as shown by these experiments varied from 0.9 kimt> to 1.3 knots with an average of 1.1 knots. In addition to the adverse effect upon barnacle attachment it was noticed that slime film development was greatly reduced on the rotated disc. ATTACHMENT IN GLASS TUBES Owing to the difficulty of accurately measuring the rate of water How along the surface of the rotating disc, an independent method of investigating the relation be- tween water currents and fouling incidence was used. In this case sea water was passed through sections of glass tubing of varying diameter (Fig. 3). Comparison of fouling incidence in each sector was made with the varying linear velocity of the water, calculated on the assumption that linear rate of flow through tubes varies inversely as the square of the diameter of cross-section. In order to estimate errors which might be introduced by virtue of turbulent flow, carmine \vas intro- duced into the water during a preliminary test run. It was observed that flow in the main portion of each section was smooth. Areas near to the joints, where turbulence was noted, were not included in the experimental observations of fouling incidence. The tubes were approximately eight inches long and of 5 cm., 3.7 cm., 2.8 cm., 2.2. cm., and 1.4 cm. internal diameter respectively and were joined by means of fitted rubber stoppers. Attachment of barnacle larvae was encouraged by exclud- ing direct sunlight. The water flow was provided by means of a centrifugal pump, with the experimental tubes arranged on the inlet side in order to avoid possible mechanical damage to the larvae. While the apparatus was in operation the rate of flow was checked daily by volumetric measurement at the outlet. At the conclusion of each test run the tubes were carefully examined for signs of fouling. Results Observations on the glass tubes showed that limiting velocities for attachment of BahniHS amphitritc were between 0.5 and 1.0 knot. During the first experiment with this apparatus, sea water was run through it at a rate between 13 and 16 liters per minute, on a catwalk extending over the water at the University of Miami Marine Laboratory. Fluctuations in the height of tide and consequently the suc- tion head of the seawater pump gave rise to this fluctuation in current velocity. Barnacles attached in the two larger tubes within ten days after beginning the ex- periment. From the calculated linear velocity of the narrower of these tubes it appeared that barnacles are not prevented from attachment by currents varying between 0.5 and 0.6 knot (Table II'). In the next smaller tube where no barnacles attached the rate of flow varied from 0.8 to 1.0 knot. The limiting velocity there- fore lies between 0.5 knot and 1.0 knot. The appearance of the tubes at the end of a similar experiment is illustrated in Figures 3 and 4. The upper figure of the limits is lower than the critical velocity as measured on the rotating disc. However, in view of the difference in hydrodynamical conditions involved in the two methods, which was discussed previously, the results appear to be reasonably consistent. I-'. G. WALTON SMITH FIGURE 2. Portion of disc rotated at 192 r.p.m. for a period of 16 days, showing barnacles attached at center. Diameter of disc, 24 inches. KM. i !\ and place Katr <>t flow. Liters per minute Equivalent velocity in glass tubes (Knots) 1.4 cm. diameter 2.2 cm. diameter 2.8 cm. diameter .3.7 cm. diameter 5.0 cm. diameter 11/17/43 13-16 3.0-3.7 1.2-1.6 0.8-1.0 0.5-0.6 0.2-0.3 to 11/27 13 Miami — — • — B. aniphitrite B. aniphitrite Beach 5/20/44 21 5.3 2.3 1.4 0.8 0.4 to 7 12/44 Kure Beach* Balanus sp. Balanus sp. 7/12/44 12.7 3.2 1.3 0.8 0.5 0.2 to 9/18 44 Kure Beach* Balanus sp. Balanus sp. Balanus sp. 9/18/44 10.8 2.7 1.1 0.7 0.4 0.2 to 11/25/44 Kure Beach* ft. impro- msus B. inipro- visus B. eburneus B. improvisus B. ebnrn/'ns B. improvisus * Observations by \Y. F. Clapp and F. L. LaQue. GROWTH ON THE ROTATING Disc For the purpose of studying the effect of water currents upon barnacles already attached the rotating disc procedure was modified. Both rotating and stationary discs were allowed a preliminary period of stationary immersion during which barnacles became attached. Equal sectors of the disc were temporarily protected bv means of cloth attached "with thumb tacks. The cloth was removed from each sector after increasing intervals of time with the result that at the end of the pre- liminary period of immersion the maximum age of the organisms attached to the different parts of the disc varied from 16 days to 6 hours. The barnacles on each sector were then counted and the maximum diameter of the largest barnacle in each sector was measured. Following the period of stationary immersion the control disc was allowed to remain in a stationary position and the other was placed on the rotating shatt lor a further period. Both discs were examined and the number of barnacles per square inch and their greatest diameter were recorded separately lor each sector and for area:- at successive distances from the center of the disc. Results The 1ir>t experiment of this series provided an initial 16 days' period of station- ary immersion during which both experimental and control discs became seeded WATER CURRENTS AND GROWTH OF BARNACLES 59 with barnacles. At the end of this period barnacles had been attached on the five sectors for maximum periods of 6 hours. 1 day. 2 days. 5 days, and 16 days re- spectively (Fig. 4). Immediately following the period of stationary growth, the experimental disc was rotated at an angular velocity of 60 r.p.m. for \{) consecutive clays, while the control disc was allowed to remain in a stationary position for the same period. The appearance of both discs at the end of the experiment i> shown in Figures 5 and 6. FIGURE 5. Same disc as in Figure 4 after further period of 19 da>> -tationary immersion. Control to disc shown in Figure 6. Examination of the discs at the conclusion of the test showed that while increas- ing current velocities diminished the growth rate and finally brought about loss of attachment, the weaker currents appeared to enhance growth. Thus, when rotated on Sector I, where growth had continued for 6 hours before rotation, B. amphitrite 60 F. G. WALTON SMITH IMI.CRK <>. Similar disc to that of Figure 4 after period of \() days rotation at 6(1 p. p.m. Age of barnacles before rotation is marked on each sector. neither continued development nor remained attached at diameters beyond 16 inches, corresponding to a current of 2.7 knots (Table III). Within Sector VI, however, with barnacles originally 16 days old, a small number remained attached even at the outer edge of the disc, where the water flow approximated 4.7 knots. On the in- tervening sectors with barnacles of intermediate ages, the current velocities bring- ing about complete loss of attachment showed values between these extremes. I'robably because of further attachments on the stationary control, the density of attachments on all sectors of the rotated disc was below that of the control. The etlects upon growth rate are illustrated by the observations in Table IV. On Sector I. within a -1 inch diameter, equivalent to 0.7 knot, barnacles reached a maximum width of X mm., compared with growth on the control of only 7 mm. At 2.3 knots, growth was reduced to onl 3 mm. and at 2.7 knots, no barnacles re- WATER CURRENTS AND GROWTH OF BARNACLES 61 TABLE III Effect of -water currents upon adhesion of barnacles of different ages when subjected to 19 days rotation upon a submerged disc, as shown by density of attachment in number per square inch. Experiment initiated January 8, 1944 Position on disc Age before rotation Diameter (Inches) Nominal water velocity (Knots) Sector I (6 hours) Sector II (1 day) Sector III (2 days) Sector IV (5 days) Sector V (16 days) 0-28* 0 Dense Dense Dense Dense Dense 4 0.7 1.7 1.7 1.7 2.0 Too crowded to count Few 6 1.0 0 0.4 0.9 1.7 8 1.3 0 0.4 1.0 1.3 10 1.7 0.1 0.4 1.0 1.0 12 2.0 0.2 0.4 0.6 1.9 14 2.3 0.1 0.4 0.9 1.3 16 2.7 0 0.4 0.6 0.9 18 3.0 0 0.1 0.2 0.6 20 3.3 0 0 0.1 0.4 22 4.0 0 0 0 0 28 4.7 Few in cracks 0 Few in cracks Few in cracks * Stationary control. mained on the sector. At intermediate current velocities conditions of growth lay between these two extremes, with normal growth occurring between 0.7 and 1.7 knots. Similar observations were made upon barnacles of greater initial age within the remaining sectors. Thus, on Sector IV, where development had continued for a maximum period of 5 days before rotation, barnacles showed enhanced growth rate in currents up to one knot, normal growth similar to that of the control up to 2 knots, and almost complete cessation of growth at 3.3 knots. No barnacles remained on this sector at 4 knots. Similar conditions of enhanced growth at low velocities, normal growth at in- termediate velocities and cessation of growth with loss of attachment at higher velocities were observed on the remaining sectors, with barnacles of different initial ages. The anomalous absence of enhanced growth on Sector V is probably due to the crowded condition of the barnacles which had become attached during a 16 day stationary period. The one day old barnacles were apparently less affected by the 62 F. G. WALTON SMITH TABLE IV Effect of water currents upon growth of barnacles of different ages when subjected to 19 days rotation upon a submerged disc, as shown by diameter of largest barnacle in millimeters. Experiment initiated January 8, 1944 Position on disc Age and size before rotation Diameter (Inches) Nominal water velocity (Knots) Sector I 6 hours «1 mm.) Sector II 1 day ( <1 mm.) Sector III 2 days ( <1 mm.) Sector IV 5 days (2 mm.) Sector V 16 days (5 mm.) 0-28* 0 7 mm. 7 mm. 8 mm. 8 mm. 9 mm. 4 0.7 8 mm. 8 mm. 8 mm. 9 mm. None more than 8 mm. 6 1.0 — 8 mm. 9 mm. 9 mm. 8 1.3 — - 8 mm. 8 mm. 8 mm. 10 1.7 6 mm. 7 mm. 6 mm. 8 mm. 12 2.0 5 mm. 5 mm. 6 mm. 8 mm. 14 2.3 3 mm. 6 mm. 4 mm. 6 mm. 16 2.7 — • 5 nun. 3 mm. 5 mm. 18 3.0 — 1.5 mm. 1.5 mm. 4 mm. 7 mm. 20 3.3 — — 1.5 mm. 2.5 mm. 6 mm. 24 4.0 — • — — — 6 mm. 28 4.7 1.5 mm. — 1.5 mm. 4 mm. 5 mm. * Stationary control. inhibitory action of water currents than the two day old barnacles. Otherwise, the adverse effects of water currents increased with their velocity, and decreased with the initial age of the barnacles. On all sectors a few barnacles continued to grow at the perimeter of the disc where the presence of cracks provided local flow pockets with relatively still water. The first experiment demonstrated an actual loss of barnacles at current veloci- ties having critical values for the different initial ages of barnacles. In order to determine the time required for the loss of attachment to occur, a second experi- ment \vas carried out in which the disc was examined, not only at the beginning and end of the rotation period, but also at intervening times. The density of attach- ment and maximum size of barnacles were observed at each examination, as in the previous experiment. The experiment was carried out during a period of heavy barnacle set. At the eiit exposed, was very much greater than in the previous experiment. WATER CURRENTS AND GROWTH OF BARNACLES 63 TABLE Y Density of barnacle attachment compared with rate of flow of water relative to the surface of a rotating disc, expressed in number per square inch * (Figures in brackets indicate density upon a stationary panel serving as control; D, over 75 per square inch) Stationary growth period Diameter of disc (Inches) Velocity (Knots) Period of rotation Ohrs. 1 hr. 12 hrs. 3 days 5 days 7 days 9 days 11 days Sector I Control (0) (30) (30) (30) (30) (30) (60) (D) (D) 4 1.5 30 25 25 20 15 10 8 5 8 3 30 30 20 20 15 10 8 6 6 hrs. 12 5 30 30 20 20 15 6 6 6 16 6.5 30 30 20 20 15 6 6 5 20 8 30 30 20 20 15 4 4 2 24 10 35 30 20 20 12 4 4 2 Sector II Control (0) (30) (30) (30) (30) (30) (60) (D) (D) 4 1.5 30 30 30 30 20 15 15 8 8 3 30 30 30 30 20 15 15 12 12 hrs. 12 5 30 30 30 30 20 15 15 6 16 1.5 30 30 30 30 20 15 15 6 20 8 30 30 30 30 20 10 10 5 24 10 30 30 30 10 ' 7 7 7 4 Sector III Control (0) (30) (30) (30) (30) (30) (60) (D) (D) 4 1.5 30 30 30 20 20 12 12 10 8 3 30 30 30 20 20 12 12 12 1 day 12 5 30 30 30 20 12 10 10 10 16 6.5 30 30 30 20 12 10 10 6 20 8 30 30 30 20 12 7 7 5 24 10 30 30 30 7 5 5 4 4 Sector IV Control (0) (30) (30) (30) (40) (40) (50) (D) (D) 4 1.5 30 30 30 20 20 17 17 15 8 3 30 30 30 24 20 20 20 19 2 days 12 5 30 30 30 24 20 20 15 9 16 6.5 30 30 30 24 20 20 15 8 20 8 30 30 30 15 10 10 10 7 24 10 30 30 30 10 7 7 6 6 Sector V Control (0) (30) (30) (30) (50) (50) (D) (D) (D) 4 1.5 35 35 35 32 32 32 30 30 8 3 35 35 30 30 30 30 30 30 12 5 35 35 30 30 30 30 30 24 5 days 16 6.5 35 35 30 30 30 30 30 15 20 8 35 35 30 30 30 30 30 15 24 10 35 35 25 25 25 25 25 15 Sector VI Control (0) (70) (70) (70) (70) (D) (D) (D) (D) 4 1.5 75 75 75 75 75 75 75 50 8 3 75 75 75 70 70 70 70 30 10 days 12 5 75 75 75 70 70 70 70 30 16 6.5 75 75 75 70 70 70 70 25 20 8 75 75 75 70 70 70 70 20 24 10 75 75 75 70 70 70 70 20 Experiment initiated 2/19/45. 64 F. (;. WALTON SMITH Observations showed that, as in the previous experiment, losses in attached barnacles occurred roughly in proportion to the current velocity, although at no point of the disc was the entire set lost. Similarly, the limiting rate of flow neces- sary for inhibition of growth or for loss of attachment increased generally with the initial age of the barnacles. Since observations were made at intervals during the experiment it was also possible to note the period of rotation necessary to bring about the first losses in attachment. This also increased generally with increasing age of initial growth. The observations are set forth in detail in Tables V and VI and summarized in Table VII. In Table V the number of barnacles per square inch is recorded for various dis- tances from the center of the disc, and for each individual sector. These observa- tions are repeated for various intervals of time during the experiment. An increase took place in numbers on the control panels due to continued at- tachment of cyprids, whereas on the rotated disc barnacles became progressively reduced in number as the experiment continued. After one hour of rotation the only losses were among the six hour old barnacles on Sector I. These losses were very small, however, and although observed at current velocities of 1.5 knots and 10 knots, they did not occur at intervening velocities. The first significant losses took place twelve hours after the start of rotation among barnacles less than twelve hours old. Of the older ones, only those 5 days old showed losses, amounting to a change at most velocities from 35 to 30 per unit area. Barnacles on the remaining sectors, varying in age from twelve hours upwards, were not observed to diminish in number until the third day of rotation. The smallest loss over this period took place among the twelve hour barnacles, which were only lost at speeds of ten knots, whereas, after a similar period of rotation, other barnacles of greater ages had been lost even at 1.5 knot current velocity. Ten day old barnacles were not greatly disturbed by 1.5 knot currents until after eleven days' rotation. These observations, summarized in Table VII, indicate that in general the velocity of current necessary to dislodge barnacles increased with the age they had reached before the experiment. Barnacles twelve hours old, however, showed a marked increase in resistance to being dislodged, even compared to those ten days old, whereas those twro days old, on the other hand, seemed to be less re- sistant than younger or older ones. It is also noteworthy that in general the dura- tion of rotation necessary to bring about the first losses depends much more upon the initial age of the barnacle than upon the current velocities. In a similar manner the percentage of the original number of barnacles remain- ing following an eleven day period of rotation increased with increasing age before rotation. The ten day old barnacles at the end of the experiment, however, showed poorer resistance and had a lower percentage of continued attachment than in the case of the five day barnacles. It must be noted, however, that the ten day bar- nacles were densely crowded and it is possible that this factor decreased their resistance to dislodgement. Further examination of the results in Table V discloses that losses continued to occur to some extent on all sectors at the end of eleven days, although they were less on those sectors where initial age of the barnacle was greater. Observations of the si/.e of barnacles recorded during the experiment are given in Table VI. On Sector I the six hour old barnacles within a diameter of 8 inches, equivalent to a three knot water current, continued to grow, although at a slow rate. WATER CURRENTS AND GROWTH OF BARNACLES 65 TABLE VI Growth of barnacles in relation to flow of water over the surface of a rotating disc, expressed as greatest diameter in millimeters * (Figures in brackets refer to stationary control) Stationary growth period Diameter of disc (Inches) Velocity (Knots) Period of rotation 0 hrs. 1 hr. 12 hrs. 3 days S days 7 days 9 days 11 days Sector I Control (0) (0.3) (0.3) (0.3) (0.5) (0.8) (1.3) (2.5) (3.0) 4 1.5 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.5 8 3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 6 hrs. 12 5 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 16 6.5 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 20 8 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 24 10 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 Sector II Control (0) (0.3) (0.3) (0.3) (0.6) (0.8) (2.5) (3.0) (3.5) 4 1.5 0.3 0.3 0.3 0.3 0.3 0.3 1.0 1.5 8 3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 12 hrs. 12 5 0.3 0.3 0.3 0.3 0.3 0.3. 0.3 0.3 16 6.5 0.3 0.3 • 0.3 0.3 0.3 0.3 0.3 0.3 20 8 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 24 10 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 Sector III Control (0) (0.3) (0.3) (0.3) (0.6) (0.8) (2.0) (3.0) (3.5) 4 1.5 0.3 0.3 0.3 0.5 0.5 1.0 2.0 3.0 8 3 0.3 0.3 0.3 0.5 0.5 0.5 0.5 0.5 1 flay 12 5 0.3 0.3 0.3 0.5 0.5 0.5 0.5 0.5 16 6.5 0.3 0.3 0.3 0.5 0.5 0.5 0.5 0.5 20 8 0.3 0.3 0.3 0.5 0.5 0.5 0.5 0.5 24 10 0.3 0.3 0.3 0.5 0.5 0.5 0.5 0.5 Sector IV Control (0) (0.3) (0.3) (0.5) (0.8) (1-0) (3.0) (3.5) (4.0) 4 1.5 0.3 0.3 0.5 1.0 1.0 1.3 1.3 1.5 8 3 0.3 0.3 0.3 0.8 0.8 0.8 0.8 1.0 2 days 12 5 0.3 0.3 0.3 0.8 0.8 0.8 0.8 1.0 16 6.5 0.3 0.3 0.3 0.3 0.8 0.8 0.8 0.8 20 8 0.3 0.3 0.3 0.3 0.8 0.8 0.8 0.8 24 10 0.3 0.3 0.3 0.3 0.8 0.8 0.8 0.8 Sector V Control (0) (0.8) (1.0) (1.5) (2.5) (3.0) (4-0) (5.0) (6.0) 4 1.5 0.5 0.8 1.0 2.5 2.5 3.0 3.0 3.5 8 3 0.5 0.8 0.8 1.5 1.5 1.5 1.5 1.5 5 days 12 5 0.5 0.5 0.8 1.5 1.5 1.0 1.5 1.5 16 6.5 0.5 0.5 0.8 1.5 1.5 1.0 1.5 1.5 20 8 0.5 0.5 0.8 1.5 1.5 1.0 1.5 1.5 24 10 0.5 0.5 0.8 0.8 0.8 0.8 0.8 0.8 Sector VI Control (0) (2.5) (2.0) (2.5) (4.0) (5.0) (7.0) (9.0) (9.0) 4 1.5 2.0 2.0 2.0 3.5 3.5 4.5 4.5 5.0 8 3 2.0 2.0 2.5 3.0 3.0 3.5 3.5 3.5 10 days 12 5 2.0 2.0 2.0 3.0 3.0 3.5 3.5 3.5 16 6.5 2.0 2.0 2.0 3.0 3.0 3.5 3.5 3.5 20 8 2.0 2.0 2.0 3.0 3.0 3.5 3.5 3.5 24 10 2.0 2.0 2.0 3.0 3.0 3.5 3.5 3.5 Experiment initiated 2/19/45. 66 F. G. WALTON SMITH TABLE VII Summary of observations on the effect of water currents produced by rotation upon attached barnacles. Expressed as multiples of normal growth, period before loss in number first occurred, and percentage remaining at the end of eleven days * Current velocity (Knots) Period of stationary growth 6 Hrs. 12 Hrs. 1 Day 2 Days 5 Days 10 Days 0.1 X 0.4 X 0.8 X 0.3 X 0.6 X 0.6 X 1.5 (1 Hour) 20% 5 Days 30% 3 Davs 30% 3 Days 50% 3 Days 90% 11 Days 70% Inhibition Inhibition Inhibition 0.3 X 0.2 X 0.3 X 3 12 Hours 30', 5 Davs 40% 3 Days 40% 3 Days 60% 2 Days 90% 3 Days 40% 5 — — 0.1 X 0.2 X 0.3 X 12 Hours 20% 5 Davs 20% 3 Days 30% 3 Days 30% 2 Days 70% 3 Days 40% — — — Inhibition 0.2 X 0.3 X 6.5 12 Hours 20% 5 Days 20% 3 Days 20% 3 Days 30% 2 Davs 60% 3 Days 30% Inhibition 0.3 X 8 12 Hours 20% 5 Days 20% 3 Days 20% 3 Days 20% 2 Davs 60% 3 Days 30% 0.3 X 10 (1 Hour) 10% 3 Davs 2<>'; 3 Days 20% 3 Days 20% 2 Days 60% 3 Days 30% * Experiment initiated 2/19/45. Even at velocities below 1.5 knots the growth was considerably less than that of the stationary control. The ten day old barnacles on Sector VI continued to grow near the circumference, although subjected to ten knot currents. Barnacles of inter- mediate age showed a reduction of growth inversely dependent upon their initial age, and directly dependent upon the current. Barnacles less than two days old ceased growth at water currents over three knots, and only the ten day old barnacles continued growth at ten knots. At the end of this experiment the remaining barnacles were examined to de- termine whether any considerable portion remaining attached was dead. A few of the older barnacles had lost their soft parts, but it was difficult to obtain a reliable criterion as to the condition of the remainder. It was not definitely established, therefore, what proportion had died. Differences between the results of this and the previous experiment lie mainly in the higher current velocities required to bring about growth inhibition and loss of attachment and the absence of complete loss at any part of the disc in the later experiment. This apparent anomaly may be partly due to the much greater density of initial attachment, which may have reduced the effective velocity of the currents by providing a much more irregular surface. It is also conceivable that had the WATER CURRENTS AND GROWTH OF BARNACLES 67 experiment continued for the full 19 day period of the first experiment rather than 11 days only, the growth inhibition effect might have appeared at lower velocities and might have resulted in complete loss at some parts of the disc. DISCUSSION The experimental results indicated that current velocities limiting the attach- ment of barnacles lie in the neighborhood of one knot and that the limits for the three species, Balanus cburncus, B. amphitrite, and B. iniprovisus are in ascending order of magnitude. That a one knot current should limit attachment is in accord- ance with the conclusions of Visscher and earlier workers that attachment of barnacles does not readily take place upon ships in motion. It further appears that growth rates of previously attached barnacles are reduced by water currents in inverse proportion to the age of the barnacle and directly in proportion to the current velocity. The effect of crowding together is further to reduce the growth rate. The effect of very low current velocities is to increase the rate of growth. A slight loss of attachment occurs at all current velocities, but this may be partly due to overcrowding. Large and significant losses take place at velocities which are sufficient to reduce the growth rate. These losses begin some hours after the start of rotation, and continue for a period of at least eleven days. Since losses due to dislodgement do not occur during the first few hours of rota- tion, except in the case of barnacles attached within six hours or less, and since they continue to take place during the entire period of rotation, it seems probable that they are not directly due to mechanical action. Possibly the mechanism is one of interference with feeding processes, followed by reduced growth rate, death and diminished adhesion. A further point arising out of the experimental data is the fact that at equivalent current velocities 12 hour and 24 hour old barnacles show greater growth and less loss of attachment than those somewhat older. It is possible that at this stage of development, with metamorphosis incomplete, an orientation to the current may occur which facilitates feeding, although this was not actually observed. Once attached the barnacles are able to withstand current velocities of increasing magnitude, although for the first two or three weeks the growth rate is considerably reduced. Loss of attachment appears to follow reduced growth rate and may in- volve actual death from lack of food. The current velocities required to bring this about are sufficiently low to provide support for Visscher's conclusion that barnacles are killed by ocean passages of 500 miles or more, if vessels of more than 10 knots performance are considered, or if the age of attached barnacles is no greater than a 10 day stay in port would allow. On the other hand, current velocities lying below this figure, such as would be produced by slow sailing vessels, would appear from the experimental data to be insufficient to bring about a loss of the attached bar- nacles. In the case of faster ships, stays in port of longer than a few days might also permit the accumulated barnacles to survive an ocean passage. This type of ship service would readily provide a means of distribution for barnacles. It would explain Fischer-Piette's observations of the appearance of Balanus amphitrite on Mediterranian shores, and would support his hypothesis of distribution by ships. The enhanced growth rate in current velocities of 1.3 knots and less, depending on the age of the barnacle, provides a quantitative expression for the observations 68 F. G. WALTON SMITH of Kice. Moore. Kitching. and others upon the increased growth of barnacles sub- jected to water currents. The conditions in a harbor or an embayed or open coast are rarely such that tidal currents in excess of one knot are continuously present. Under these condi- tions, therefore, there would be no obstacle to the attachment of barnacles. Once attached, however, the barnacles would usually be subjected to intermittent tidal currents. No data are available as to the effects of intermittent currents but it would be reasonable to assume that these would be, if anything, less than the effects of con- tinuous currents. The data presented here would therefore make it appear unlikely that any but the strongest of tidal currents would prevent the attachment and con- tinued development of barnacles on rocks or stone, and wooden piling. Where fairly continuous currents of less than 1.5 knots were present the growth would be encouraged. A different situation occurs in estuaries where continuous currents might occur. Here, however, the limiting factor, according to the work of previous authors, is probably lowered salinity, since continuous currents would scarcely permit of the estuary remaining salt. Where wave action is strong the effects might be considered as equivalent to strong currents of frequently changing direction, which might conceivably inhibit attachment or growth, even below tide levels. Even where attachment occurred during temporary lulls, the strong recurrent wave action might be expected to dislodge the barnacles. This may explain the observations of Stephenson and his co-workers that some species of barnacles do not occur in a zone of heavy wave action. The effect of water currents under natural conditions may therefore be sum- marized as one of enhanced growth with velocities below 1.5 knots, and one of re- duced growth above this. It would rarely happen under natural conditions that water currents alone would prevent barnacle colonization. Strong wave action, however, might be expected to have this effect. Certain qualifications should be added to the above discussion which may ex- plain some anomalous observations. The arguments above are based on the cal- culated rates of flow of water over the rotating disc, and the results of experi- ments have indicated that these rates are slightly higher than the actual rates of flow taking place. The error, however, does not appear to be greater than 10 per cent to 20 per cent, and does not qualitatively affect the validity of the general con- clusions. Since the observations were made upon the relatively smooth surface of a wooden disc, they would not be applicable to very rough surfaces where local pockets of relatively still water might develop, to the areas immediately adjacent to lapped plates of ships' hulls, or to portions of piling or other submerged surfaces where the contours and configuration might produce local stagnation effects. SUMMARY 1. The work of previous authors, dealing with the effect of water currents upon barnacle attachment, growth and distribution is briefly reviewed. 2. Experiments were conducted to determine the effects of water currents upon the attachment and growth of barnacles, and particularly of Balanus amphitrite. WATER CURRENTS AND GROWTH OF BARNACLES 69 Submerged rotating discs and glass tubes of graded cross-sectional diameter were employed to provide variations in current velocity. 3. Tbe velocity of water current limiting attachment appears to lie between 0.5 and 0.9 knot for Balanus amphitritc, between 0.4 and 0.7 knot for B. cburncus, and above 1.1 knots for B. hnprovisus. 4. Following attachment the growth rate of barnacles was found to be increased by water currents of velocity less than 1 .5 knots and to be decreased by currents with velocities in excess of this. The adverse effects of water currents were found to decrease with increasing age of the barnacles subsequent to attachment. Six hours after attachment, growth rate was reduced to one-third of normal by a 1.5 knot current and completely stopped by a 3 knot current. Five days after attachment, growth was prevented by currents ranging between 3.3 and 8 knots. Loss of attachment appeared to some extent among all barnacles in which growth rate was reduced. This loss was greatest at velocities bringing about complete cessation of growth. 5. It is suggested that loss of attachment is due to interference with the feeding process, followed by reduction of growth rate, death, and diminished adherence. Possibly due to an orientation to the current which facilitates feeding, barnacles attached for one day 6r less show less retardation of growth rate and loss of ad- herence than barnacles two days old. 6. Since tidal currents are almost invariably intermittent it appears from the data presented that they are not sufficient to prevent the colonization of suitable surfaces by barnacles, except where the velocities are unusually high. It also follows from the numerical results obtained that on vessels making short stays in port and relatively long voyages, little permanent barnacle fouling will occur, since those organisms which attach will be killed and at least a portion of them dislodged. The evidence does not preclude, however, the continued growth of barnacles upon slow vessels making longer stays in port, and their geographical distribution by this means. LITERATURE CITED BRIGHT, K. M. F., 1938. The South African intertidal zone and its relation to ocean currents. II, III. Areas of the west coast. Trans. Roy. Soc. S. Africa, 26 : 49-88. BOKENHAM, N. A. H., AND T. A. STEPHENSON, 1938. The colonization of denuded rock sur- faces in the intertidal region of the Cape Peninsula. Ann. Natal Mus., 9: 47-82. EYRE, J., 1939. The South African intertidal zone and its relation to ocean currents. VII. An area in False Bay. Ann. Natal Mus., 9 : 283-306. EYRE, J., AND T. A. STEPHENSON, 1938. The South African intertidal zone and its relation to ocean currents. V. A sub-tropical Indian shore. Ann. Natal Mus., 9: 21-46. EYRE, J., BROCK RUYSEN, G. J., AND M. I. CRICHTON, 1938. The South African intertidal zone and its relation to ocean currents. VI. The East London district. Ann. Natal Mus., 9: 83-112. FISCHER-PIETTE, E., 1928. Recherchcs de bionomie et d'oceanographie littorales sur la Ranee et le littoral de la Manche. Ann. Inst. Oceanogr. Monaco, n. s. 5: 201-429. FISCHER-PIETTE, E., 1928 (a) Sur la distribution geographique de quelques organismes de rocher, le long des cotes de la Manche. Maritime Mus. Nat. Hist. Natur. Saint-Servan, II. Trav. du Lab. FISCHER-PIETTE, E., 1929. Le cirripede Balanus amphitrite Darwin a Saint-Servan. Bull. Lab. Maritime Saint-Servan, 4: 10-11. FISCHER-PIETTE, E., 1932. Repartition des principals especes fixees sur les rochers battus des cotes et des iles de la Manche, de Lannion a Fecamp. Ann. Inst. oceanogr. Monaco, n. s. 12 : 105-213. 70 F. G. WALTON SMITH I-isi. HKK-PiETTE, E., 1935. Quelques rcmarques bionomiques sur la cote basque franchise et espagnole. Bull. Lab. Maritime Saint-Sert'an, 14: 1-14. HENTSCHEL, E., 1924. Das Werden und Vegehen des Bewuches an Schiffen. Mitt. Zool. Mus. Hamburg, 41 : 1-51. HIRAGA, YUZURU, 1934. Experimental investigations on the resistance of long planks and ships. Zoscn Kiokai (J. Soc. Naral Arch. Japan), 55: 159-199. L.\QuE, F. L., 1943. Proving ground of Marine Corrosion. Inco Magazine, 19: 4-7. McDouGALL, K. D., 1943. Sessile marine invertebrates at Beaufort, North Carolina. Eco- logical Monographs, 13 : 321-374. , MOORE, H. B., 1933. Change of orientation of a barnacle after metamorphosis. Nature, 132: 969-970. MOORE, H. B., 1934. The biology of Balanus balanoides. I. Growth rate and its relation to size, season, and tidal level. Jour. Marine Biol. Ass. U. K., n. s. 19: 851-868. MOOKE, H. B., 1935. The biology of Balanus balanoides. IV. Relation to environmental factors. Jour. Marine Biol. Ass. U. K., n. s. 20 : 277-307. MOORE, H. B., 1936. The biology of Balanus balanoides. V. Distribution in the Plymouth area. Jour. Marine Biol. Ass. U. K., n. s. 20 : 701-716. MOORE, H. B., 1939. The colonization of a new rocky shore at Plymouth. Jour. Aniin. Ecol., 8 : 29-38. MOORE, H. B., AND J. A. KITCHING, 1939. The biology of Chthamalus stellatus (Poli). /. Marine Biol. Ass. U. K., n. s. 23: 521-541. NEC, WOLFGANG, 1932. Untersuchungen iiber den Schiffobewuchs. Int. Rev. Hydrobiol. Hydrogr., 27: 105-119. NEU, WOLFGANG, 1933. Biologisches Arbeiten iiber den Schiffsbewuchs. Int. Rev. Hydrobiol. Hydrogr., 29 : 455-458. XEWCOMBE, C. L., 1935. A study of the community relationships of the sea mussel, Mytilus edulis, L. Ecology, 16 : 234-243. PHELPS, A., 1942. Observations on reactions of barnacle larvae and growth of metamorphosed forms at Beaufort, N. C., June, 1941 to Sept. 1941. Ponrth semi-annual Report from Woods Hole Oceano graphic Institution to Bureau of Ships. I., Paper VII. (Un- published.) PIERROX, R. P., AND Y. C. HUANG, 1926. Animal succession of denuded rocks. Publ. Pugct Snd. Biol. Sta., 5: 149-157. PRENNANT, M., AND G. TEISSIER, 1924. Notes ethologiques sur la faune marine sessile des environs de Roscoff. Cirripedes, bryozoaires, hydraires. Trar. Sta. Biol. Roscoff, 2 : 1-24. RICE, L., 1930. Peculiarities in the distribution of barnacles in communities and their probable causes. Publ. Pugct Snd. Biol. Sta., 7 : 249-257. SHELFORD, V. E., AND E. D. TOWLER. 1925. Animal Communities of the San Juan Channel and adjacent areas. Publ. Puget Snd. Biol. Sta., 5: 33-73. STEPHENSON, T. A., STEPHENSON, A., AND K. M. F. BRIGHT, 1938. The South African inter- tidal zone and its relation to ocean currents. IV. The Port Elizabeth district. Ann. Natal Mus., 9 : 1-20. STEPHENSON, T. A., STEPHENSON, A., AND C. A. DU TOIT, 1937. The South African intertidal zone and its relation to ocean currents. I. A temperate Indian Ocean shore. Trans. Roy. Soc. S. Africa, 24 : 341-382. TOWLER, E. D., 1930. An analysis of the intertidal barnacle communities of the San Juan Archipelago. Puhl. Puget Snd. Biol. Sta., 7 : 225-232. TURNER, H. J. JR. The concentrations of chlorine and sodium pentachlorphenate required for the prevention of fouling in sea water pipe systems. Interim Report No. XI from the Woods Hole Oceonographic Institution to the Bureau of Ships, Navy Department. Nov. 8, 1945. (Unpublished.) VISSCHER, J. P., 1928. Nature and extent of fouling of ships' bottoms. Bull. Bur. Fisheries, 43 (2) : 193-252. VISSCHER, J. P., 1930. Fouling of ships' bottoms. 2. Factors causing fouling. 3. Methods preventing fouling. Paint and I 'arn. Prod. Mgr., 35, 36. VISSCHER, J. P., 1938. Some recent studies on barnacles. Biol. Bull., 75 : 341-342. MICRURGICAL STUDIES ON CHIRONOMUS SALIVARY GLAND CHROMOSOMES l ETHEL CLANCY D'ANGELO Department of Biology, Washington Square College, New York University, and Queens College,2 Flushing, New York Particular interest in the giant salivary gland chromosomes of clipteran larvae was awakened by Painter (1933) and Heitz and Bauer (1933) who demonstrated the existence of a close relationship between the genetic significance of the chromo- somes and their structure. Regarding the gross structure of these chromosomes there is general agreement, but not for fine details. Bauer (1936) and Painter and Griffen (1937) offered cytological evidence for a polytene concept of chromosome structure from observation, in fixed and stained preparations, of numerous longi- tudinal striations which they considered to be chromonemata. Metz and Lawrence (1937) and Metz (1941) maintained that the longitudinal striations appearing in these fixed and stained preparations were merely the result of drawing out the walls of alveoli. Buck (1942) afforded support for this view by showing in osmic vapor- treated chromosomes that an original alveolar structure could be reversibly trans- formed into a fibrillar-like one by micromanipulation. Kodani (1942), using cytolytic methods with alkali and urea, found evidence for only four chromonemata in each chromosome, a condition which Painter and Griffen (1937) and Buck (1937) observed in the early larval stages, but not in the giant chromosomes of later stages. More recently, Frolova (1944) has given experimental support to the polytene concept by demonstrating the presence of numerous longitudinal strands in fixed chromosomes after partial digestion with nuclease preparations. Aside from their cytogenetic interest, the salivary gland chromosomes, because of their large size and visibility in vivo, make especially suitable material for studies on physical properties. Using the micrurgical technique, various investigators (Vonwiller and Audova, 1933; Barigozzi, 1938; Stefanelli, 1939; and Pfeiffer, 1940) have found the freshly isolated salivary gland chromosomes to be tough, viscid, elastic gels. These observations are in essential agreement with those first made by Chambers and Sands (1923) on the chromosomes of Tradescentia pollen mother cells and by Chambers (1924a) on those of Dissosteria spermatocytes. Buck (1942), using osmic vapor-treated chromosomes, reported more detailed findings on elasticity. That the physical properties of the chromosomes may be easily modified by the presence of torn cytoplasm or calcium ions was shown by Duryee (1941) in his studies on amphibian chromosomes. The observations of Chambers (1924b) that torn cytoplasm is acidic, and of Chambers and Reznikoff (1926) that the micro-injection of various salts into protoplasm may have detrimental effects, 1 This work, done at Washington Square College, New York University, and at the Marine Biological Laboratory, Woods Hole, Mass., is presented in partial fulfilment of the requirements for the degree of doctor of philosophy at New York University. - Present address. 71 KTHEL GLANCY D'ANGELO further point out the need for precaution in studying the physical properties and structure of the chromosomes. This investigation is concerned with the structure and physical properties of the salivary gland chromosomes as observed within the intact fresh cell and after isola- tion into media designed to maintain them in the fresh condition. Grateful acknowledgment is made to Prof. Robert Chambers, under whose guidance this work was done, for his constant aid and invaluable criticism. MATERIALS AND METHODS The species selected for study were Cliironoinns phiiuosus. C. teutons, and cer- tain smaller unidentified species of Chironomus. Chironoinus larvae were chosen in preference to those of other Diptera because of their larger chromosomes (maxi- mum size, 20 /A in diameter and 150^, long) and the flattened shape of the salivary gland, factors which facilitate micromanipulation. Moreover, the chromosomes were always plainly visible in the living cell, the bands being especially well defined. Incidental observations were made on Drosophila chromosomes and they were found to agree in all important respects. The glands were carefully removed from mature larvae under a dissecting micro- scope and immediately immersed in either a drop of Chironomus hemolymph or amphibian Ringer's solution which was found to be isotonic. Care was taken to avoid the presence of any tissue but the gland itself since otherwise the medium became unduly acid. The hemolymph was obtained by inserting a pipette through the body wall of a larva into the hemocoele. In order to be satisfactory the hemo- lymph had to be freshly obtained and immediately covered with paraffin oil to pre- vent the development of alkalinity which occurred on exposure to air. The micromanipulation was carried out immediate!}' after mounting a single in- tact gland in the medium on a thoroughly cleaned coverslip which was then inverted over the moist chamber of a Chambers micromanipulator. Structural changes within the chromosomes of the intact cells do not appear for many hours in hemo- lymph or Ringer's solution, but extra precautions were taken by making observa- tions within 15 minutes after mounting the glands. The micropipettes routinely used had an inside bore of never over one /*, and the microneedles used for fine dis- section of the chromosomes tapered rapidly to points approaching the limit of visibility. Xuclei and chromosomes were removed from disintegrating cytoplasm as quickly as possible. This was necessary in order to preserve their normal appearance, an observation in agreement with that of Duryee (1937) on amphibian germinal vesicles. To remove the chromosomes, an entire gland was mounted under a dis- secting microscope in a medium, to be described later, in which the chromosomes were to be isolated. By means of fine steel needles a cell was quickly torn and the nucleus ruptured. The torn nucleus remained in situ, and, by exerting a gentle pressure on the torn cell, the chromosomes could be forced out of the nucleus and floated away without coming in contact with the injured cytoplasm. The loosely adhering clump of chromosomes was immediately transferred by means of a lip pipette to a fresh drop of the medium to be used, then mounted in the usual manner in the moist chamber. In this way the chromosomes were found to retain most nearly their normal appearance and physical properties after isolation, their sub- SALIVARY CHROMOSOME MANIPULATION 73 sequent condition depending on the medium in which they remained. No medium was found which would maintain the chromosomes in their normal condition indefi- nitely nor were extensive attempts made to find such a medium. Among the media tried and found to be unsatisfactory were hemolymph, Ringer's solution, paraffin oil, 0.1 M NaCl-0.01 M KC1 solution used by Duryee (1941), 0.25 per cent egg al- bumen solution used by Melland (1938), and 0.3 M sucrose solution suggested by the work of Chambers and Sands (1933). These special solutions were used with- out regard to pH conditions. From determinations on a great variety of cells including the Chironomu-> salivary gland nucleus, Chambers (1929) found the intranuclear pH to be within 7.6-7.8. It would be expected therefore that the medium in which the isolation of the chromosomes is to be done should have a pH within the aforementioned limits. However, the most satisfactory mixture, when buffered to pH 7.6-7.8, caused the isolated chromosomes to swell and become excessively sticky. Moreover, the swell- ing was accompanied by a fading of internal structure. However, at pH 7.0, the chromosomes closely approached, both in appearance and physical properties, those within the nucleus. It should be noted that the consistency was slightly greater than that of the intact chromosomes. The structure and properties of the chromo- somes were maintained throughout the experimental period. The medium used consisted of 0.09 M KC1, 0.06 M NaCl, and 0.005 M phosphate buffer, pH 7.0. Observations and micromanipulation experiments were routinely made with a Leitz microscope using a 1.8 mm. oil immersion objective, N.A. 1.25, and a 10 X ocular. A micrometer ocular was used for measurements. OBSERVATIONS The chromosomes in the fresh cell in Ringer's solution or hcmolymph The chromosomes and nucleolus in the fresh cell were always plainly visible, the bands being sharply defined and either beaded or homogeneous in appearance. A hyaline material, variously referred to as nuclear sap, nucleoplasm, or karyolymph, separated the chromosomes from each other. It has been shown (Glancy, 1940) that this hyaline material is differentiated into a central portion of jelly-like con- sistency in which the chromosomes are embedded, and a more fluid peripheral zone. The evidence for this regional differentiation was based on experiments involving the injecting of oil drops and carbon particles as well as the manipulation of the chromosomes. The hyaline material is referred to in this paper as the nuclear matrix. Stickiness. It was found possible to insert microneedles into the nucleus and to force the chromosomes against each other as well as against other structures in the nucleus, and also to push them against injected oil drops and carbon particles. Two chromosomes pushed together by means of microneedles could easily be sepa- rated again without sticking to each other. It should be noted, however, that a narrow hyaline zone, presumably the jellied nuclear matrix, always remained be- tween them, thus probably preventing actual contact of the two chromosomes. Sim- ilar results were obtained when a chromosome was forced against the nucleolus or the nuclear membrane. 74 ETHEL CLANCY D'ANGELO A micropipette containing an aqueous carbon suspension was inserted as close as possible to a chromosome without actually touching it, and a small quantity of the Mi^pension injected. In none of the attempts were carbon particles seen to adhere to the chromosomes, nor did they ever make contact with a chromosome. How- ever, injection of a large amount of fluid caused liquefaction of the jelly matrix, thus permitting the carbon particles to make contact with the chromosomes. In a simi- lar manner, various oils ; namely, Nujol, olive, almond, and peanut oils, were in- jected into the nucleus. None of the oil drops adhered to the chromosomes. When the cytoplasm was torn without tearing the nucleus, a change occurred in the nucleus. This was made evident by the marked shrinkage of the chromosomes. Two chromosomes pushed together in the nucleus of such a torn cell invariably stuck to each other, and thick viscid strands were pulled out when attempts were made to separate them. The chromosomes also adhered to the nucleolus, the nuclear membrane, injected oil drops, carbon particles, and the microneedles. In every case, separation could be achieved only when accompanied by thick viscid strands pulled out from the chromosomes. In summary, these experiments show that the chromosomes, because they are prevented from coming in contact by the jelly-like nuclear matrix, appear to be non- sticky in the intact cell. Liquefaction of this matrix occurs when the cytoplasm is torn, thus permitting the chromosomes to make contact. They are then found to be sticky. Consistency. The micrurgical experiments to be described indicate that, compared to the cytoplasm and the nuclear matrix, the chromosomes are relatively soft, easily deformable gels, and that there are differences in consistency between the bands and the interband regions. A very finely tapered microneedle was pushed up through the cytoplasm and through the nuclear membrane, into the interior of the nucleus. The needle was pushed in from the side, and even before it was brought into actual contact with a chromosome, the surface of the chromosome became indented on the side next to the needle. The needle was pushed still farther and when its tip went into the interior of the chromosome, the indentation disappeared and the original contour of the chromosome was restored. This deformation of the chromosome must be due to the central jelly-like matrix since insertion of the needle into the more fluid peripheral region of the nucleus had no such deforming action on the chromosomes. The indentation of the chromosome, when the needle tip was being brought into its vicinity, and the readiness with which the chromosome returned to its original shape after the needle had penetrated indicated that the chromosome was of a softer con- sistency than the surrounding central nuclear matrix. The relatively low consistency of the chromosomes compared to the central nuclear matrix was also shown by experiments in which oil drops were injected into the nucleus. The effect of the oil drops in causing deformation of an adjacent chromosome differed according to the site of injection. Drops of Nujol, olive, almond, and peanut oils, 1-20 /x in diameter, were injected close to a chromosome. Large oil drops carefully injected into the central jelly-like nuclear matrix produced a rounded depression in the nearest chromosome. However, this did not occur when the matrix was agitated by the needle, thus showing the thixotropic nature of the jelly matrix. Peripherally injected oil drops were never observed to cause de- SALIVARY CHROMOSOME MANIPULATION 75 formation. These results indicate that the consistency of the central nuclear matrix is greater than that of the chromosomes, whereas the reverse is true in the peripheral region of the nucleus. The nucleolus was found to be the most easily deformed body in the nucleus. This was shown by forcing the nucleolus against the nuclear membrane, the chromo- somes, and the microneedles. In all cases the nucleolus was deformed in the man- ner of a fluid drop. When the nucleolus was separated by a microneedle from the small fourth chromosome to which it is normally attached in Chironomus, the nucleolus became spherical. Evidence for differences in consistency between the bands and the interband regions was obtained in several ways. A microneedle inserted into a chromosome and then withdrawn produced a space which closed over more rapidly in an inter- band region than in a band. The thin bands were intermediate between the inter- band regions and the thick bands in this respect. The longer persistence of the hole in the bands is indicative of a higher consistency. The greater resistance of the bands, especially the thicker ones, to deformation during manipulation of the chromosomes, as in the stretching experiments, also is in accord with their higher consistency. The results of injecting oil drops into the chromosomes led to similar conclusions. Of the various oils used, Nujol, olive, almond, and peanut oils, only the last two could be readily injected into a chromosome. This was easily done in an interband region, but only with difficulty or not at all in a band. The injected oil drops became spherical, showing the relatively low consistency of the interband regions, and caused the displacement of the relatively viscid bands on either side. Extensibility and elasticity. Observations on extensibility and elasticity were made by stretching a small portion of a chromosome between two microneedles. The degree of stretching which could be performed, however, was limited since the nuclear membrane was torn when the microneedles were moved more than approxi- mately 20-30 /z apart. In all cases in which a part of a chromosome was stretched within the nucleus, release from the needles resulted in a complete return to the original appearance. A short region of a chromosome was seized lengthwise between two microneedles and the needles were slowly drawn apart. As the chromosome was stretched, there was a decrease in its diameter, and its structure became less distinct. The thick bands were more resistant to stretching than the thinner ones, whereas the inter- band regions stretched most readily. During the stretching the broad bands often separated into several narrow bands, thus demonstrating their compound nature. Faint longitudinal striations which followed the lines of tension generally appeared in the interband regions when a chromosome was stretched more than twice its length. These stress lines became more pronounced as stretching was increased. They did not appear to be identical with the delicate longitudinal striations occa- sionally seen in unstretched fresh chromosomes. On release of tension, the stria- tions disappeared and the chromosome, even after being stretched five times its length, returned to its original condition. It was not found possible to stretch a given region more than this without tearing the nuclear membrane. In every case, the chromosomes recovered completely after the microneedles were removed. Lateral stretching of the chromosomes was also performed. Two microneedles were inserted into a chromosome close to the margins, then slowly drawn apart. 76 ETHEL CLANCY D'ANGELO The chromosome cylinder flattened during the process and hands which had ap- peared homogeneous became beaded (Fig. 1). As tension was increased, stress lines, radiating from the point of each microneedle to the chromosome, became visi- ble. On release of tension these lines disappeared, and the beaded appearance of the bands was lost. A chromosome could be stretched approximately three times its original width and recover fully. FIGURE la. Sketch of a portion of a chromosome showing two microneedles inserted directly into a band prior to lateral stretching of the chromosome within the intact cell. All sketches are of Chironomus chromosomes. FIGURE Ib. Beaded appearance of the band when the chromosome in (a) is stretched laterally. Note tension lines in the interband regions. The chromosome membrane. The micrurgical evidence to be presented in- dicates that the salivary chromosome possesses a delicate elastic membrane. A microneedle was inserted into the margin of a chromosome within the intact nucleus, in an attempt to remove any existing membrane. When the needle was slowly and carefully withdrawn, faint lines were seen radiating from it to the chromosome (Fig. 2a). To test whether or not these lines were actually folds in FIGURE 2a. Lifting by a microneedle of a membrane-like material from a portion of a chromosome within the intact cell. FIGURE 2b. Micro-injection of a carbon suspension into the area subjacent to the lifted membrane shown in (a). FIGURE 2c. Concentration of the carbon particles into a narrow zone as a result of gradual contraction of the membrane. a membrane which had been lifted off from the chromosome the same operation was repeated and an aqueous carbon suspension was injected into the triangular area in question. The carbon particles converted the region showing the radiating lines into a sac-like structure which remained connected to the chromosome (Fig. 2b). When the micropipette was carefully withdrawn, the fluid evidently escaped from the area, as the sac slowly contracted and the carbon particles came closer to the chromo- some where they became packed against its surface (Fig. 2c). Repetition of the SALIVARY CHROMOSOME MANIPULATION 77 experiment injecting a larger quantity of the suspension into the triangular area first converted it into a sac ; then as injection was continued, the sac burst, and the car- bon particles scattered in the matrix where they briefly exhibited Brownian move- ment. Oil drops were also injected in a similar manner into the triangular area pro- duced by pulling on the margin of the chromosome. After the withdrawal of the micropipette, the oil was gradually drawn to the chromosome surface where it re- mained as a slight protuberance. Larger oil drops were injected and could be seen to be held against the chromosome by what appeared to be a membrane. No de- formation of the chromosome occurred, presumably because there was no jelly- like matrix between the oil drop and the chromosome. Various salt solutions (0.01-1.0 M NaCl, KG, and CaCL) and also distilled water were injected into an interband region in an amount which caused a swelling of the region without breaking its contour. Figures 3a, b, and c show the condi- FIGURE 3a. Micropipette inserted into an interband regiun of a portion of a chromosome within the intact cell showing the initial stage of injection. FIGURE 3b. Localized swelling and loss of structure in the injected portion of the chromo- some. Results are similar with solutions of NaCl, KC1, GaCL, and distilled water. FIGURE 3c. The appearance of the same chromosome after injection of sufficient CaCl. solution to rupture the injected region. The bands have reappeared in their original positions, but the distinct boundary (membrane) shown in (a) is no longer evident. Note the granular region of the nuclear matrix. tion of the chromosome before, during, and after the injection. With the progress of the injection, the lesser bands in the vicinity faded out and the zone underwent swelling until a heavy band on each side of the injected zone was reached. The injection was continued within the limit of bursting. The pipette was then re- moved, and the swollen zone returned to its original dimensions in which the final appearance was that of the region prior to injection. The surface of the injected region could be ruptured when the amount of fluid injected was sufficient to cause it to swell to approximately two and a half times its original diameter. Thereupon the membrane-like boundary suddenly disappeared irrespective of the fluid injected. When CaCL. was used a granulation appeared in the nuclear matrix around the injected zone (Fig. 3c). This did not occur with water, NaCl, or KC1 solutions. It may be concluded, therefore, that the salivary chromosome possesses a delicate elastic membrane. Internal Structure. The visible structure of the chromosomes was studied by careful and extended observations of the chromosomes within the intact nucleus KTHEL GLANCY D'ANGELO both of the isolated gland and of glands within the intact larvae.2 For this purpose the observations were made with a Leitz 1.4 N.A. apochromatic oil objective and a triple lens condenser. The equipment was sufficient to resolve the hexagonal aper- tures of the Pleurosignia angulatum. The distinct and sharply defined bands were beaded in some regions, and homogeneous in appearance in others. The interband regions, which were much less refractile, generally appeared either homogeneous or finely granular. In unusually large chromosomes the granulation was coarse and often the arrangement gave the effect of longitudinal striations. These granular striations, which were seen at all levels in the chromosomes, were individually con- nected to the beads in the bands. During these observations the chromosomes were not under tension and in the case of the isolated gland were free from compression. That the salivary gland chromosome is made up of numerous longitudinal fibrils was further indicated by microdissection. The chromosome membrane was first removed by a very fine microneedle over a length of several bands from the region to be studied. The needle was then inserted into the margin of an exposed inter- band region. This adhered to the needle and, when the needle was carefully moved to one side, a strip of very thin fibril with one end on the tip of the needle could be seen being carried away from the chromosome. As the fibril was being carried still farther it was separated from the chromosome over a length of several bands (Fig. 4) . FIGURE 4. A single fibril with a chromoneina-like appearance being removed by a microneedle from a chromosome within the intact cell. The isolated fibril showed nodular swellings which corresponded in position with the bands from which it was being separated. This was done over and over again with the same chromosome and with different chromosomes. The same phenomenon could be observed whether the bands on the operated chromosome were homogeneous or beaded. The fibrils could be stretched and, when tension was released, they returned to their original lengths. One case was observed in which a fraying of the chromo- some resulted when a pipette had been inserted to inject water. The injection was made close to the periphery and the pipette then removed. These partially isolated fibrils closely resembled that in Figure 4. The microdissection experiments were performed more readily with large chromosomes in which longitudinal striations were visible. They were also done with the chromosomes in which the interband regions appeared homogeneous. The dissection evidence strongly suggests that longitudinal fibrils possessing nodular swellings really exist in the chromosomes. The state of the chromosomes in nuclei rendered optically homogeneous by alkaline or hypotonic Ringer's solution. The question has frequently been 2 As Buck (1938) has already described, the intact larvae can be mounted between slide and covcrslip and because of the transparency of its body wall, the cells in the salivary glands arc easily visible under high power, and can be satisfactorily obserxed. Their appearance was identical with that of the isolated glands mounted in Ringer's solution SALIVARY CHROMOSOME MANIPULATION 79 raised whether chromosomes in so-called hyaline nuclei maintain their morphological integrity or whether they lose their identity through dispersion in the hyaline phase. Naturally occurring hyaline salivary gland nuclei were never found, but could be readily produced by immersion of the gland in slightly hypotonic or in mildly alka- line (pH 8.4) Ringer's solution. \Yithin a few minutes after immersion the nucleus swelled, chromosomes faded, the interbands disappearing first, and finally only the nucleolus remained visible. When the gland was returned after several minutes to isotonic Ringer's solution at pH 7.0, the chromosomes reappeared. A microneedle could be inserted into one of these hyaline nuclei without causing its collapse and usually without producing any visible effect. In some cases, how- ever, the nuclear membrane wrinkled slightly when the microneedle was withdrawn, indicating that fluid was lost. When a microneedle was moved slowly back and forth within the nucleus no regional resistance to its movement was encountered. Similarly no resistance was offered to movement of the nucleolus throughout the entire nucleus. Return of the gland to Ringer's solution caused the chromosomes to reappear. It is noteworthy that the chromosomes which reappeared were fre- quently distorted in shape. Carbon particles injected into the center of an untreated nucleus remained at the site of injection presumably because they were caught in the central jelly matrix in which the chromosomes are embedded. However, an injection made just under the nuclear membrane of an untreated nucleus, where the nuclear matrix is relatively fluid, resulted in- a scattering of the carbon particles between and around the parts of the chromosomes in the peripheral region. None of the particles penetrated the central region of the nucleus. Exposure of such an injected nucleus to an alkaline or hypotonic Ringer's solution or to distilled water, caused the chromosomes to disappear. The fact that the chromosomes still retained their identity and did not go into solution was shown by the alignment of the carbon particles which kept their relative positions outlining the contiguous borders of what must be the swollen chromosomes (Fig. 5). The carbon particles failed to become completely dispersed FIGURE 5. Alignment of carbon particles outlining the contiguous borders of adjacent chromosomes in an optically homogeneous nucleus. The nucleus was first injected with a carbon suspension, then hyalinized by immersion of the gland in hypotonic Ringer's solution. as would be expected if the chromosomes had gone into solution. In nuclei which were first made hyaline, then injected with the carbon suspension in the peripheral region, a similar alignment of the carbon particles resulted. In both cases the chromosomes reappeared in the expected positions when the glands were placed in Ringer's solution. 80 ETHEL CLANCY D'ANGELO It may be concluded, therefore, that although moving microneedles back and forth within the hyaline nucleus fails to reveal the presence of the swollen chromo- somes, nevertheless the chromosomes have maintained their morphological integrity as shown by the injection of carbon particles. Isolated chromosomes The physical properties of the isolated chromosomes may be seriously affected by the methods used in their isolation and by the medium in which they are maintained. In the following experiments, attempts were made to maintain the properties which had previously been found for the chromosomes within the nucleus. It was found that rapid isolation of the chromosomes without contact with the torn cytoplasm was essential to prevent changes in structure and physical properties. Chromosomes which remained in the torn cell or came in contact with the cytoplasm during the isolation were shrunken, often distorted, and highly sticky. Of the various media tried, the one which caused the least change in the chromosomes after their isola- tion was the neutral KCl-NaCl solution described in the section on methods. In the experiments which follow on the isolated chromosomes, it is understood, unless otherwise stated, that the chromosomes were isolated into the neutral KCl- NaCl solution without coming in contact with the torn cytoplasm, transferred to a fresh drop of the same solution, and then manipulated. The duration of an experi- ment was not more than 10 minutes. Stickiness. Chromosomes isolated into the neutral KCl-NaCl solution. Ringer's solution or hemolymph, were not unduly sticky. It was found that alka- linizing the medium (pH 8.0) caused the chromosomes to become sticky, whereas acidifying the medium (pH 6.5) reduced stickiness. Consistency. The consistency of the isolated chromosomes in slightly alka- line KCl-NaCl solutions, pH 7.6-7.8, was found to resemble most closely that of the chromosomes in the intact cell, but structure was indistinct and the chromo- somes were sticky. As the pH was decreased, the chromosomes shrank and the consistency increased. In neutral solutions little shrinkage occurred and the con- sistency was but slightly greater than that of the chromosomes within the intact cell. In acidic solutions (pH 6.0-6.5), the effects were more marked. At corresponding pH values, the consistency was always somewhat greater in Ringer's solution and in hemolymph than in the KCl-NaCl medium. Chromosomes maintained in the neutral KCl-NaCl solutions were, like those in the intact nucleus, soft easily deformable gels which could be compressed without buckling. When they were cut or pulled apart by microneedles the cut or broken ends assumed smooth, rounded contours. This is in marked contrast to the results obtained with chromosomes which have been isolated after exposure of the glands to osmic acid vapors, formaldehyde, or acetic acid, following the methods of Buck (1942). Even brief treatment has a pronounced effect on the consistency of the chromosomes, causing them to become relatively rigid bodies which break squarely across when stretched. It may be concluded that, although the consistency of isolated chromosomes maintained in a slightly alkaline KCl-NaCl solution most closely resembles that of the chromosomes within the cell, a neutral medium is preferable for manipulation since structure is still maintained and stickiness is lessened. SALIVARY CHROMOSOME MANIPULATION 81 Extensibility and elasticity. The use of isolated chromosomes permits un- restricted stretching which obviously is impossible in the intact cell. It was found possible to stretch an isolated chromosome 10 times its original length and obtain complete recovery on release of tension. During the process of stretching, the interbands stretched somewhat more readily and so underwent a greater decrease in diameter than the bands. The wide bands separated into two or more narrow bands as occurred in the intact cell operations. As stretching was increased the diameter of the chromosome decreased so that at 10 fold elongation it appeared to be a mere thread with occasional nodular swellings. If this degree of stretching was not repeated more than 2 or 3 times, it was completely reversible. Repeated stretch- ing caused permanent deformation, particularly of the interband regions, which re- mained in a partially stretched condition even after tension was released. Much greater stretching was possible without causing a chromosome to break in two. In some cases, a chromosome could not be broken within the limits of stretching of the micromanipulator. Thus one chromosome was stretched 25 fold without breaking. When tension was released, it returned to 1.3 X its original length. Other chromo- somes broke in twro when stretched 12 fold. The break always occurred in an inter- band region, although this was not apparent until the chromosome had partially recovered. Fibrillae were occasionally seen at the broken ends. It should be pointed out that greatly stretched chromosomes developed a coarse fibrillar structure and rapidly deteriorated. This fibrillar appearance was different from that ob- served in unstretched chromosomes in that these fibrils were relatively coarse, and always followed the lines of tension. Deterioration also occurred, although much more slowly, in chromosomes which were not manipulated, indicating that the medium is not all that is to be desired. It was found that on the addition of acids (pH 6.0-6.5) or calcium ions to the KCl-NaCl medium, the consistency of the bands was increased more than that of the interband regions. This effect was especially marked in the acid solutions, the bands becoming curved or bent rather than stretching or separating into narrow bands when placed under tension. The interband regions, although having a higher consistency than in neutral solutions, nevertheless were readily extensible, and when stretched more than twice their length appeared coarsely fibrous. At a 5-10 fold elongation, a chromosome broke in an interband region as before with fibrillae being evident at the broken ends. In slightly alkaline solutions, pH 8.0, the chromo- somes became extremely ductile and showed almost no elasticity. The loss of elasticity was partially prevented by the addition of calcium salts to the alkaline solutions. Isolated chromosomes were also stretched laterally. These gave results quali- tatively similar to those observed in the intact cell. The isolated chromosomes, however, could not be stretched more than approximately double their width with- out deteriorating. In some instances, at least, the deterioration was associated with tearing the chromosome membrane. In some of the experiments Buck's (1942) method was used of exposing glands to osmic acid vapors for 18-24 hours at 5° C, then isolating the chromosomes in distilled water. The chromosomes could be isolated readily by this method, and, in agreement with Buck, were tough, relatively rigid bodies, breaking squarely across in an interband region when stretched more than four times their length. Occa- sional fibrillae appeared extending from the broken ends of the chromosomes. This KTHKI. (il.AXCV D'ANGELO was also observed In Buck, although he attached no particular significance to it. Stretching was reversible only up to 2 fold elongation (or 100 per cent as stated by Buck). Thus these chromosomes were far less extensible and elastic than those in the fresh state. Structure. The chromosomes isolated in the KCl-NaCl solution at pH 7.0 gen- erally showed no structure either in the interband regions or in the bands. Occa- sionally fibrillae were visible in the interband regions. On the other hand, in an acid solution (pH 6.5), fibrillar structure was more generally evident. In those cases where the fibrillar structure was already apparent, it became more distinct when the medium was slightly acidified. Isolated chromosomes were dissected in the same manner as the chromosomes in the intact cell. The fibrils removed from the isolated chromosomes, behaved, except for a slightly increased stiffness, similarly to those removed from the chromosomes within the nucleus. Occasionally it was found possible to produce a breakdown of the chromosomes into numerous longitudinal fibrils. This was done as follows : an intact gland was immersed in a hanging drop of neutral KCl-NaCl or in amphibian Ringer's solu- tion, then a large pipette, about 15/j. in diameter, was inserted into the nucleus of an intact cell and a single chromosome withdrawn. This could be done readily, indicating the ease with which the nuclear matrix becomes solated, and the relative non-adhesiveness of the chromosomes to each other. The chromosome was then expelled into the drop to one side of the gland, whereupon it sometimes became shredded into numerous longitudinal fibrils which resembled the fibrils obtained by dissection of an intact chromosome. This operation was successful only if the bore of the pipette was close to the diameter of the chromosome, thus suggesting the possibility that the chromosome membrane may have been removed while the chromosome was being either sucked into or expelled from the pipette. If the pipette was much larger than the chromosome, then the chromosome was expelled intact. The fibrils of the shredded chromosome were very numerous, sometimes appearing as a jumbled mass, and sometimes as a bundle of essentially parallel longi- tudinal fibrils. This experiment was* repeated several times on the same day and on different days. The operation wras not successful every time but the fault could be ascribed to the critical bore of the pipette and the speed with which ejection was made. The fraying of these chromosomes occurred irrespective of whether the medium into which they were expelled was NaCl-KCl or amphibian Ringer's solution. The operation was not done a sufficient number of times to discover an optimum medium for the fraying. The fact that it can be made to occur even very occasionally is highly significant. DISCUSSION The results of these micrurgical experiments indicate that the normal structure and physical properties of salivary gland chromosomes can be maintained only if due precaution is taken to avoid the deleterious effects of torn cytoplasm. The "acid of injury" reaction of torn cytoplasm, originally described by Chambers (1924b), is significant in this regard. More recently, Duryee (1941) has reported the in- jurious effects of torn cytoplasm and calcium ions on the isolated chromosomes and germinal vesicles of amphibian oocytes. The changes in structure and physical SALIVARY CHROMOSOME MANIPULATION 83 properties of the salivary gland chromosomes which occur in a torn cell appear to be similar to those effected by acid or calcium ions. This suggests that calcium and hydrogen ions may be released from the injured cytoplasm. The general finding that chromosomes isolated in media of a physiological pH are sticky raises the question as to why they do not adhere to each other within the normal nucleus. It has previously been shown (Clancy, 1940) that the chromo- somes within the salivary gland nucleus are normally held apart by a thixotropic jelly matrix which becomes solated when the cell is torn. Although the chromo- somes in the torn cell tend to become progressively sticky, this change is minimized if they are immediately removed from the cell so that contact with the torn cyto- plasm is avoided. With prolonged contact, however, the chromosomes become so sticky that it is virtually impossible to manipulate them. The latter observations may indicate that the stickiness of isolated chromosomes is caused in part by chemi- cal substances liberated from the disintegrating cytoplasm. The reports of various investigators (Vonwiller and Audova, 1933; Barigozzi, 1938; Stefanelli, 1939; and Pfeiffer, 1940) to the effect that isolated chromosomes are tough, viscid gels, are in contrast to the observations in this paper which in- dicate them to be soft, easily deformable gels. The discrepancy may be explained in part by the failure of these investigators to observe precautions with respect to injured cytoplasm and the medium. My observations show that the consistency of isolated chromosomes tends to be greater than that of those within the nucleus, especially in the presence of injured cytoplasm. This increase in consistency seems to be associated with a simultaneous shrinkage of the chromosomes. Since the above investigators manipulated the chromosomes in the same drop of hemolymph or Ringer's solution in which the isolation had been performed, it is likely that the drop may have been acid because of the presence of the injured tissues. Even slight acidification of the medium has been shown to cause an increase in consistency. A similar effect results from the presence of calcium ions. Osmic acid vapors, used by Buck (1942) in treating the salivary glands prior to isolation of the chromo- somes, cause far greater changes in physical properties although no shrinkage occurs. The findings of Buck, which we have confirmed using his technique, to the effect that the chromosomes are tough, cannot be compressed without buckling, and that they break squarely across when stretched, are at variance with those obtained on fresh chromosomes as shown in this investigation. It is thus evident that the consistency of the chromosomes is influenced by many factors. The relative consistency of the bands and the interband regions has been dis- cussed ever since Balbiani (1881) first suggested the bands to be gel and the inter- band regions, sol. Guareschi (1939) has observed that in dark field only the thicker bands are visible, whereas the thin bands and interband regions are not visible, from which he inferred that the former are in a gel state, and the latter in a weak gel, or more probably a sol state. Our microdissection experiments confirm Guareschi's inferences as to these differences in consistency. The thin bands and the interband regions must be weak gels rather than sols, however, for a puncture made in these regions leaves a temporary opening, a condition which does not obtain in a sol. This observation is further strengthened by the high elasticity of the chromosomes for, while gels are generally very elastic, a sol has negligible elasticity. The extensibility and elasticity of the chromosomes have been studied by numer- ous investigators. Since it is well known that the properties of a colloidal structure 84 KTHKL GLANCY IVAXGELO depend to a large extent on the medium, it is to be expected that the conditions under which the chromosomes are examined will influence their degree of elasticity. It is to be noted that the stretching experiments of Vonwiller and Audova (1933), Barigozzi (1938), Stefanelli (1939) and Pfeiffer (1940) were performed in 0.6 per cent NaCl, Ringer's solution, or hemolymph, without regard for the effects of the torn cytoplasm. These investigators found that the chromosomes could be stretched many times their length, though not reversihly. Pfeiffer and Stefanelli found stretching was not reversible beyond a 2 fold elongation as did Buck (1942) using osmicated chromosomes. It has been shown in this investigation that the elastic limit of fresh chromosomes in the intact cell is at least 5 fold, and when isolated in a KC1 /NaCl medium, as much as 10 fold. This unusually high elasticity indicates the presence of long fibrous molecules (Mark, 1941) of which nucleo- proteins are a characteristic type (Mirsky and Pollister, 1942). The presence of a membrane on the salivary chromosomes has been described heretofore from the appearance of fixed and stained chromosomes (Schultz, 1941 ; Kodani, 1942). Metz (1934) believed the presence of a sheath or membrane to be necessary on theoretical grounds. Painter (1941) originally considered a pellicle to be present, but finally concluded, after studying the action of alkaline solutions on the salivary chromosomes, that the chromonemata are held together by a matrix, and that a pellicle is lacking. That the fresh chromosome possesses a membrane is strongly suggested by my microdissection experiments and observa- tions that the outer boundary of the chromosome ruptures at a critical diameter when stretched by fluid injections. The membrane may aid in preventing contact between the chromosomes, since they tend to become sticky if it is torn, or it may serve to hold the chromonemata together. The question as to whether or not the chromosomes in hyaline nuclei maintain their morphological integrity has frequently been raised. Stefanelli ( 1939) and Guareschi (1939), studying salivary gland nuclei rendered hyaline by immersion oi the gland in hypotonic solutions, concluded that the chromosomes are completely dispersed. Stefanelli's conclusions, based primarily on the observations that micro- needles can be moved back and forth within the hyaline nucleus without meeting resistance and that the nucleolus can be displaced without rebounding, do not con- clusively demonstrate that the chromosomes are dispersed, however. The failure to find evidence for chromosomes by this technique might be explained in several ways: first, movement of the microneedles within the thixotropic nucleus might well induce solation ; secondly, the consistency of the nuclear matrix in such nuclei may he close to that of the chromosomes; and thirdly, the swollen chromosomes are closely pressed against the nuclear membrane, leaving the center of the nucleus almost devoid of any resisting structure. To Guareschi's conclusions, it may be objected that a dilute gel cannot necessarily be distinguished from a sol under dark field illumination, and secondly, that the refractive indices of the colloidal particles in the chromosomes and the surrounding nuclear matrix may have been changed by the hydration which occurs in hypotonic solutions. Micro-injection results on salivary gland nuclei, likewise made hyaline by hypotonic solutions, indicate that the chromosomes, though invisible, maintain their morphological integrity, since injected carbon particles clearly indicated the outlines of the chromosome cylinders. If the chromosomes were completely dispersed, the injected carbon particles should be distributed at random in the nucleus. SALIVARY CHROMOSOME MANIPULATION 85 The polytene concept of chromosome structure rather than the alveolar is sup- ported by the various observations made in this investigation. The controversy re- garding the structure of the salivary chromosomes results in part from the fact that fixed and stained chromosomes do not always present a similar appearance with re- gard to the finer details. Since the striations are especially evident in stretched chromosomes, Metz (1941) believes them to be artifacts produced by drawing out the walls of alveoli. Buck (1942) has provided experimental support for the alveolar hypothesis. It is agreed that a fibrillar-like appearance in homogeneous- appearing chromosomes can be produced by stretching. However, I have never observed a pre-existing alveolar structure in fresh chromosomes, nor does Buck state that the alveolar condition was present before he treated the glands with osmic acid vapors, although he does state that the treated chromosomes closely resemble those in the living animal. Most of his photomicrographs fail to show an alveolar structure. It is conceivable that stretching the chromosomes could cause aggrega- tion of delicate, scarcely visible or non-visible fibrils, into the relatively coarse fibrils which are commonly seen in stretched chromosomes. Buck tried unsuccessfully to shred the chromosomes into fibrillae. This may easily be accounted for by the fact that he used osmicated chromosomes which are of such high consistency that shredding is impossible. A fibrillar appearance was occasionally noted in the fresh salivary chromosomes studied in this investigation. Buck (1942) also reported that fresh chromosomes may occasionally show longitudinal striations, but he attached no particular sig- nificance to this observation. In chromosomes which show no structural details in the interband regions, a slight shrinkage produced by mild acidification (pH 6.5) results in the appearance of delicate longitudinal striations but never alveoli. Since the chromosomes have not been stretched, it does not seem likely that the fibrillar appearance can be explained on a basis of drawn out alveoli. The observation that fibrillae may appear at the broken ends of greatly stretched chromosomes may be interpreted in two ways, as Buck (1942) pointed out when he made a similar ob- servation on osmicated chromosomes. The fibrillae may have been pre-existing, or they may be artifacts produced by the stretching. More significant is the micro- dissection from fresh unstretched chromosomes, either within the intact nucleus or after isolation, of delicate longitudinal fibrils which give the appearance of chromo- nemata. It is difficult to see how such fibrils could be removed from an alveolar mesh. Attempts to remove transverse fibrils, which should be equally possible if an alveolar structure exists, were uniformly unsuccessful. The finding that a fresh chromosome, expelled from a micropipette into Ringer's or KCl-NaCl solution, may break up into numerous delicate longitudinal fibrils, but never an alveolar mesh, is also difficult to explain except on a polytene basis. SUMMARY AND CONCLUSIONS Micrurgical experiments performed on the fresh salivary gland chromosomes of mature Chironomus larvae within the nucleus of the intact cells and after their isola- tion into a specially designed medium indicate the following : 1. Puncture of the cell with a fine microneedle has little or no effect on the structure and physical properties of the chromosomes provided tearing of the cyto- plasm is prevented. 86 ETHEL CLANCY D'ANGELO 2. Tearing of the cytoplasm causes the chromosomes to shrink markedly and become highly viscid. 3. The structure and physical properties of chromosomes isolated into a medium consisting of 0.09 M KC1, 0.06 M NaCl, and 0.005 M phosphate buffer at pH 7.0, are not appreciably altered, except for a slight increase in consistency, from those of chromosomes in the intact cell if precaution is taken to prevent contact of the chromosomes with the torn cytoplasm. Physical properties 4. Evidence for stickiness of the chromosomes can be obtained only when the thixotropic jelly matrix surrounding them is solated. 5. The chromosomes are soft, easily deformable gels with interband regions of lower consistency than the bands. 6. The chromosomes are highly extensible and elastic. They can be stretched 5 fold within the intact cell without permanent deformation. Greater stretching cannot be accomplished without tearing the nuclear membrane. Isolated chromo- somes regain their initial length after 10 fold elongation, but excessive stretching (12-25 X) causes the chromosomes to break in an interband region. 7. Modifications in the properties of the chromosomes are easily induced by chemical agents. Dilute alkalies decrease consistency and elasticity, but increase O J J ' extensibility. Dilute acids and calcium ions increase consistency, and decrease ex- tensibility and elasticity. Osmic acid and formalin have a similar but much greater effect. 8. Chromosomes in nuclei hyalinized by hypotonic or alkaline Ringer's solution retain their morphological integrity. Structure 9. The fresh chromosome possesses a delicate elastic membrane. 10. Chromosomes, either in the intact nucleus of the unoperated cell, or when isolated, occasionally display delicate longitudinal striations in the interband regions, but never show alveoli. 11. Slight acidification of the medium to pH 6.5 causes the appearance of similar striations in the interband regions of unstretched chromosomes. 12. Chromosomes broken in two by stretching frequently exhibit nbrillae at the broken ends. 13. Delicate fibrils may be dissected from the intact or isolated chromosome. The fibrils possess nodular swellings at intervals corresponding to the bands. 14. A chromosome drawn up into a micropipette of optimal size, then expelled into Ringer's or KCl-NaCl solution, may shred into numerous fibrillae. These observations are best interpreted as supporting a polytene concept of chromosome structure. LITERATURE CITED BALBIANI, E. G., 1881. Sur la structure du noyau des celles salivaires chez les larves de Chironomus. Zool. Anz., 4: 637-641. BARIGOZZI, C., 1938. Esperienze di microdissezione sui cromo somisalivari di Chironomus sp. Arch. f. exp. Zellf., 22 : 190-195. BAUER, H., 1936. Beitrage zur vergleichenden Morphologic der Speicheldriisen Chromosomen. Zool. Jahr., 56 : 239-276. SALIVARY CHROMOSOME MANIPULATION 87 BUCK, J. B., 1937. Growth and development of the salivary gland chromosomes in Sciara. Proc. Nat. Acad. Sci., 23 : 423^428. BUCK, J. B., 1938. Some properties of living chromosomes. Coll. Net., 13 : No. 8. BUCK, J. B., 1942. Micromanipulation of salivary gland chromosomes. Jour. Hered., 33 : 3-10. CHAMBERS, R., 1924a. Etudes de microdissection. IV. Les structures mitochondriales et nucleaires dans les cellules germinales males chez la Sauterelle. La Cellule, 35 : 107- 124. CHAMBERS, R., 1924b. Physical structures of protoplasm as determined by microdissection and injection. Cowdry's "General Cytology," Sec. V, pp. 235-309. Chicago Press. CHAMBERS, R., 1929. Hydrogen ion concentration of protoplasm. Bull. Nat. Res. Council, 69 : 37-50. CHAMBERS, R., AND P. REZNIKOFF, 1926. Micrurgical studies in cell physiology. I. The action of the chlorides of Na, K, Ca and Mg on the protoplasm of Amoeba proteus. Jour. Gen. Physiol, 8: 369-401. CHAMBERS, R., AND H. C. SANDS, 1923. A dissection of the chromosomes in the pollen mother cells of Tradescantia virginica. Jour. Gen. Physiol., 5: 815-821. DURYEE, W. R., 1937. Isolation of nuclei and non-mitotic chromosome pairs from frog eggs. Arch. exfi. Zellf., 19 : 171-176. DURYEE, W. R., 1941. The chromosomes of the amphibian nucleus. Cytology, Genetics, and Evolution, pp. 129-141. Univ. of Penna. Press. FROLOVA, S. L., 1944. Study of fine chromosome structure under enzyme treatment. Jour. Hered., 35 : 235-246. GLANCY, E. A., 1940. Micromanipulation studies on the nuclear matrix of Chironomus salivary glands. Biol. Bull, 79 : 372-373. GUARESCHI, C., 1929. La morfologia della cromatina delle ghiandole salivari di Chironomus plumosus in rapporto ad esperienze chimico-fisiche. Boll. Zool., 10: 109-114. HEITZ, E., AND H. BAUER, 1933. Beweise fur die Chromosomennatur der Kernschleifen in den Knauelkernen von Bibio hortulanus L. Zcit. Zcllf., 17 : 67-82. KODANI, M., 1942. The structure of salivary gland chromosomes of Drosophila melanogaster. Jour. Hered., 33: 115-133. MARK, H., 1941. Structure and behavior of high polymers. Cold Spring Harbor Symposia, 9 : 204-210. MELLAND, A. M., 1938. Isolation of salivary gland nuclei. Biol. Bull., 75 : 348. METZ, C. W., 1934. The role of the "chromosome sheath" in mitosis and its possible relation to phenomena of mutation. Proc. Nat. Acad. Sci., 20: 159-163. METZ, C. W., 1941. Structure of salivary gland chromosomes. Cold Spring Harbor Symposia, 9 : 23-39. METZ, C. W., AND E. GAY LAWRENCE, 1937. Studies on the organization of the giant gland chromosomes of Diptera. Quart. Rev, Biol., 12: 135-151. MIRSKY, A. E., AND A. W. POLLISTER, 1942. Nucleoproteins of cell nuclei. Proc. Nat. Acad. Sci., 28 : 344-352. PAINTER, T. S., 1933. A new method for the study of chromosome rearrangements and the plotting of chromosome maps. Sci., 78 : 585-586. PAINTER, T. S., 1941. An experimental study of salivary chromosomes. Cold Spring Harbor Symposia, 9 : 47-54. PAINTER, T. S., AND A. B. GRIFFEN, 1937. The structure and the development of the salivary gland chromosomes of Simulium. Genetics, 22: 612-633. PFEIFFER, H. H., 1940. Mikrurgische Versuche in polariseiertem Lichte zur Analyse des Fein- baues der Riesenchromosomen von Chironomus. Chromosoma, 1 : 26-30. SCHULTZ, J., 1941. The evidence of the nucleoprotein nature of the gene. Cold Spring Harbor Symposia, 9 : 55-65. STEFANELLI, A., 1939. Sul comportamento dei colloidi nucleari delle ghiandole salivari di Chironomus plumosus. Esperienze di micromanipolazione combinato con esperienze chimico-fisiche. Boll. Zool., 10: 149-162. VONWILLER, P., AND A. AuoovA, 1933. Mikrodissektion an der Speicheldriisen von Chironomus. Protoplasma, 19 : 228-241. SERIAL PUBLICATIONS ADDED TO THE MARINE BIOLOGICAL LABORATORY AND THE WOODS HOLE OCEANOGRAPHIC INSTITUTE LIBRARY SINCE FEBRUARY, 19431 EXPLANATION OF SYMBOLS Titles recorded alphabetically with initial dates. * ceased publication. [ ] incomplete. + receive currently. Abhandlungen der Medizinischen Fakultat der Sun Yatsen Universitat 1929: Canton. 2, no. 2; 3, no. 2 *Abhandlungen und Monographien aus dem Gebiete der Biologie und Medizin 1920: Bern and Leipzig. 1-3 Academy of Natural Sciences of Philadelphia ; Monographs 1935: 6-7 Acta Agralia Vadensia 1934: Wageningen. 1 Acta Horti Petropolitani (Trudy Imperator- skago S. — Peterburgskago Botanicheskago Sada) 1871: 16; 21-23, no. 2 Acta Physiologica Scandinavica (supersedes Skandinavisches Archiv fur Physiologic) 1940: 1, no. 1 Acta Zoologica Lilloana 1943: Tucuman, Argentina. 1 + Advancement of Science (supersedes Report of the British Association for the Advance- ment of Science) 1939: 1 + Advances in Protein Chemistry 1944: New York. 1 + American Journal of Clinical Pathology; official publication, American Society of Clinical Pathologists 1931: 1 + American Midland Naturalist; Monograph Series 1944: University of Notre Dame. 1 + American Review of Soviet Medicine 1943: American-Soviet Medical Society. 1 + Anales de la Sociedad de Biologia de Bogota 1943: 1 + Annales de 1'Universite de Paris 1926: 2, 12- 13, 14 (1939) Annals of Surgery; a monthly review of surgical science and practice 1885: 119 + *Annual of the National Academy of Sciences 1863: Cambridge. 1863-65 Annual Report of State Game Warden and Annual Report of Game and Inland Fish Commission 1938: Baltimore, Md. 1-5 (1943) Annual Report of the Division of Laboratories and Research; New York State Department of Health: 1933 + Annual Report of the Librarian of Congress; Supplement; Quarterly Journal of Current Acquisitions 1943: 1 + Arbeiten auf dem Gebiete der Chemischen Physiologic 1903: Bonn. 13 *Archives de Plasmologie Generale 1912: 1, no. 1 *Archives of Dermatology; a Quarterly Journal of Skin and Venereal Diseases 1874: 1-8 (1882) Archives of Surgery 1920: American Medical Association. 1-39, 41 + Aus der Natur; Zeitschrift fur alle Natur- freunde 1905: 5, no. 24 Beitrage zur Physiologischen Optik see Problemy Fiziologicheskoi Optiki Biologica; Trabajos del Institute de Biologia de la Facultad de Biologia y Ciencias Medicas de la Universidad de Chile 1944: 1 + Biological Review of the City College of the College of the City of New York 1938: 1 + Biometrics Bulletin; the Biometrics Section; American Statistical Association 1945: 1 + Boletin Biologico; organo de los Laboratories de la Universidad de Puebla 1942: Mexico. 1 + Boletin del Laboratorio de Estudios Medicos y Biologicos 1942: Mexico. 1 + 1 The Supplement to the Biological Bulletin, vol. 84, no. 1, 1943, gave a complete list to that date. 88 SERIAL PUBLICATIONS, MARINE BIOLOGICAL LABORATORIES 89 Botany Pamphlet; Carnegie Museum 1935: 1 + British Chemical and Physiological Abstracts, C. Analysis and Apparatus 1944: 1944+ Brittonia; a series of botanical papers 1931: New York Botanical Garden. 1 + *Bulletin de 1'Association Francaise pour 1'Avancement des Sciences (superseded by Sciences) 1896: 119-136 (1936) Bulletin(s) from the Institute for Medical Research, Federated Malay States 1924: 3 (1924) Bulletin of the Seismological Society of America 1911: 1 + Bulletin of the United States Army Medical Department 1943: 4, nos. 1-5 Calcutta University; Journal of the Depart- ment of Science 1919: n.s. 1 + Ceylon Journal of Science ; Section B, Zoology (1-22, pt. 1 as Ceylon Journal of Science) (for 1-22, pt. 1 see Spolia Zeylanica) Circular; New Jersey Agricultural Experiment Station: 33-35, 44-45, 48, 50-52, 55, 68, 71, 80-81, 84, 91-92, 100, 107 Communications on the Science and Practice of Brewing see Wallerstein Laboratories Communications Compte Rendu Annuel et Archives de la Societe Turque des Sciences Physiques et Naturelles see Turk Fiziki ve Tabii Ilimler Sosyetesi Yillik Bildirigleri ve Arsivi Congres International des Peches Maritimes, d'Ostreiculture et d'Aquiculture Marine 1896: 1896, Sables-d'Olonne, Rapports Contributions from the Biological Laboratory of the Catholic University of America 1916: 1,4-16, 18-19, 22-45 (1943) Drosophila Information Service; Material contributed by Drosophila workers 1934: Department of Genetics, Carnegie Institu- tion of Washington, Cold Spring Harbor. Fieldiana. Zoology (1-30 as Field Museum of Natural History Publications; Zoological Series) 1895: Chicago Natural History Museum. Duke University Marine Station; Bulletin 1943: 1 + Excerta Medica de la Secretaria de Comunica- ciones y Obras Publicas 1942: Mexico. 1 + Experimental Medicine and Surgery; a quarterly devoted to experimental investiga- tions of clinical problems 1943: 1 + Fieldiana. Geology (1-9 as Field Museum of Natural History Publications; Geological Series) 1895: Chicago Natural History Museum. 1 + Game Bulletin; State of California Depart- ment of Natural Resources; Division of Fish and Game 1913: 2 + Grenzfragen des Nerven- und Seelenlebens 1900: Wiesbaden. 3; 11 Helvetica Physiologica et Pharmacologica Acta 1943: 1 + Hospital Corps Quarterly; Supplement to the United States Naval Medical Bulletin 1917: 16+ Illinois Medical and Dental Monographs 1935: 3, nos. 1-2 (1940) Industrie Chimique see Przemysl Chemiczny Iowa State College Journal of Science; a quarterly of research 1926: 1 + Istanbul Universitesi Fen Fakiiltesi Mecmuasi (Revue de la Faculte" des Sciences de 1'Universite d'Istanbul) 1923: Seri A, Matematik, Fizik, Kimya. 8+; Seri B, Tabii Ilimler. 9 + lubileinoe Izdanie k 135-Letnemu lubilefu Moskovskogo Obshchestva Ispytatelei Pri- rody (Apergu Historique; Societe des Naturalistes de Moscou) 1940: 2 nos. Journal of Applied Mechanics 1933: American Society of Mechanical Engineers. 1 + Journal of Dairy Research 1929: London. 5-8; 9, nos. 1, 3 Journal of Dairy Science 1917: American Dairy Science Association. Baltimore. [8- 23] Journal of General Biology, Moscow, Aka- demiia Nauk SSSR see Zhurnal Obshchei Biologii Journal of Meteorology 1944: American Meteorological Society. 1 + Journal of the History of Ideas; a Quarterly Devoted to Intellectual History 1940: 1 + (State of) Maryland Board of Natural Re- sources; Department of Research and Education; Educational Series 1944: 1 + Mathematical Tables and Aids to Computa- tion; a quarterly journal edited on behalf of the Committee on Mathematical Tables and Aids to Computation by the Chairman, Raymond Clare Archibald 1943: National Research Council. 1 + Melanges Biologiques tires du Bulletin de 1'Academie Imperiale des Sciences de St.- Petersbourg 1849: 5, nos. 90 SERIAL PUBLICATIONS, MARINE BIOLOGICAL LABORATORIES *Memoir of the Thoreau Museum of Natural History 1914: Concord, Massachusetts. [2] Memoires de (presented a) 1'Institut d'Egypte 1862: 18 *Microscopical Bulletin and Science News 1883: Philadelphia. [3-18] Miscellaneous Reports; a Publication of the Institute of Meteorology of the University of Chicago 1942: 1 + Monographs from the Walter and Eliza Hall Institute of Research in Pathology and Medicine 1941: Melbourne. 1-2; 44- Monthly Science News: Toronto. 14-16, 19 26, 304- *Mycologisches Centralblatt; Zeitschrift fiir Allgemeine und Angewandte Mycologie 1912: Jena. 1, no. 1 National Research Council; Division of Geology and Geography; Report of the Committee on Marine Ecology as Related to Paleontology 1940: 1940-42 Northwestern University Studies in the Biological Sciences and Medicine 1942: 1 Nutrition Reviews 1942: 14- Obituary Notices of Fellows of the Royal Society 1932: London. 9, 11-13 Occasional Papers of the Marine Laboratory; Louisiana State University 1942: 1 + Occasional Papers of the Museum of Zoology; Louisiana State University 1938: 14- *Papers from the Mayo Foundation for Medical Education and Research and the Medical School 1915: 1-2 (1922) Pflanzenforschung 1922: 14- Philosophy and Phenomenological Research; a quarterly journal 1940: International Phenomenological Society. 1 4- Problemy Fiziologicheskoi Optiki (Beitrage zur Physiologischen Optik) (Problems of Physiological Optics) 1941: Leningrad, Akademiia Nauk SSSR. 14- Proceedings of the Florida Academy of Sciences 1936: 1 + Proceedings of the Louisiana Academy of Sciences 1932: 14- *Proceedings of the Meteorological Society 1861: London. 5 (1869-71) Proceedings of the Royal Entomological Society of London; series C. Journal of Meetings 1936. 14- Troceedings of the Society of Public Analysts 1875: London. 1 Proceedings of the Staff Meetings of the Mayo Clinic 1926: Rochester, Minnesota. [15-18], 194- Przemysl Chemiczny (Industrie Chimique). Warsaw. 22, nos. 11/12 (1938) *Pubblicazioni della Stazione Biologica in San Bartolomeo (1-23 as Pubblicazioni dell'Is- tituto di Biologia Marina del Tirreno in San Bartolomeo) 1928: 1-34 (1932) Publication; Department of Agriculture and Natural Resources; Bureau of Science; Manila: 13-14 Quarterly of Applied Mathematics 1943: 14- Records of Observations; Scripps Institution of Oceanography 1942: 14- *Report of the Council of the British Meteoro- logical Society 1850: 2-4; 6-10 Report of the Director to the Board of Trustees; Chicago Natural History Museum (a continuation of the Field Museum of Natural History Publications; Report Series) 1943: 19434- Revista de la Sociedad Malacologica "Carlos de la Torre"; Museo "Poey"; Universidad delaHabana 1943: 14- Revista Sudamericana de Morfologia 1943: 14- Revue de la Faculte des Sciences de 1'Uni- versite d'Istanbul see Istanbul Univer- sitesi Fen Fakultesi Mecmuasi Sbornik Rabot po Fiziologii Rastenii; Moscow, Akademiia Nauk SSSR, Institut Fiziologii Rastenii imeni K. A. Timiriazeva. 1941 Schriften der Konigsberger Gelehrten Gesell- schaf t ; Naturwissenschaftliche Klasse 1924: 1, no. 2 Science Reports of National Wuhan Uni- versity; Biological Science 1939: 1-3 Smithsonian Institution War Background Studies 1942: 14- Special Publications of the New York Academy of Sciences 1939: 14- Strukturbericht ; Zeitschrift fiir Kristallo- graphie, Kristallgeometrie und Kristall- physik; Erganzungsband 1931: 1-7 (1943) Supplement to the Journal of the Royal Statistical Society; being the organ of the Industrial and Agricultural Research Sec- tion of the Society 1934: London. 1, 34- *Supplementary Papers; Royal Geographical Society 1882: 1-3 Texas Reports on Biology and Medicine 1943: 14- Transactions of the American Neurological Association 1875: New York. 68-70 Transactions of the American Society of Mechanical Engineers 1880: New York. 674- SERIAL PUBLICATIONS, MARINE BIOLOGICAL LABORATORIES 91 Transactions (and studies) of the College of Physicians of Philadelphia 1841: ser. 3: 37-49, 52, 54; ser. 4: [1-8] Transactions of the Royal Society of Tropical Medicine and Hygiene 1907: 36+ Tropical Plant Research Foundation Scientific Contributions 1926: Washington, D. C. 1-8, 11-22 Trudy Imperatorskago S.-Peterburgskago Botanicheskago Sada see Ada. Horti Petropolitani Turk Fiziki ve Tabii Ilimler Sosyetesi Yillik Bildirigleri ve Arsivi: Istanbul. 5-6, 8+ United States Department of the Interior; Fish and Wildlife Service ; Special Scientific Report 1940: 1-9, 11 + United States Naval Medical Bulletin; Supplement see Hospital Corps Quarterly United States War Department ; Arctic Series of Publications issued in connection with the Signal Service, United States Army 1885: 1-3 (1887) United States War Department; Office of (the Chief) Engineers; Beach Erosion Board; Papers 1933: 2; Technical Memo- randum 1940: 1-2, 7; Technical Report 1941: 1-2 University of California Publications in Microbiology 1943: 1 + University of Iowa Studies in Engineering; Bulletin 1926: 2-5; 7-11; 13 + *University of Iowa Studies in Medicine 1916: 1-3 (1932) *University of Iowa Studies in Physics 1907: [2] Unsere Welt; Illustrierte Monatschrift zur Forderung der Naturerkenntnis 1909: Bonn. 1, no. 4 Uspekhi Khimii 1932: Akademiia Nauk SSSR. 12, no. 2 + *Veroffentlichungen der Zentralstelle fur Balneologie 1911: n.s. 10 Verstandliche Wissenschaft 1927: 1 Wallerstein Laboratories Communications ; (1-7 as Communications on the Science and Practice of Brewing) 1937: New York City. Yale Journal of Biology and Medicine 1928: 1 + Zeitschrift fur Kristallographie, Kristall- geometrie, Kristallphysik, Kristallchemie 1877: 105 + Zhurnal Obshchei Biologii (Journal of General Biology) (supersedes Biologicheskii Zhurnal) 1940: Akademiia Nauk SSSR. 4+ Vol. 90, No. 2 April, 1946 THE BIOLOGICAL BULLETIN PUBLISHED BY THE MARINE BIOLOGICAL LABORATORY THE SWIMMING OF LEPTOSYNAPTA DONALD PAUL COSTELLO Marine Biological Laboratory. Woods Hole and Zoology Department, University of North Carolina, Chapel Hill The fact that young adult leptosynaptids swim, or pass through a post-larval swimming stage, has not heen adequately recorded, and is not generally recognized by the students of the Holothuria. Adult leptosynaptids are usually considered to he exclusively bottom-dwelling forms, and the only reference to the contrary that I have found is a brief and incomplete statement by (."lark (1907) to the effect that: "Few, if any, synaptids swim, except in the larval state, but young ones, up to 2 cm. in length, are sometimes found floating, // not s-^'iininhitj, in the water." The italics are mine. I am therefore placing on record the following observations on the swimming habits of these animals, made during the summers of 1944 and 1945 at Woods Hole, Mass. I observed the same phenomenon a few years earlier (in 1939), but no detailed records were made at that time. On August 14th. 1944, between 9:15 and 9:30 P.M., while collecting the swarm- ing heteronereis form of Nereis li/nhata in Eel Pond, ten small, colorless, transparent holothurians were observed swimming more or less aimlessly near the surface of the water. They were being carried partly by their own movements and partly by the tidal currents, across the field of light cast on the water by the 100-watt bulb of the Nereis collecting lamp. All ten were collected. These swimming animals were 30 to 50 mm. in length and about 3 mm. in diameter. When touched in the water by the cheesecloth surface of the Nereis collecting net, they ceased their swimming movements and contracted, shrinking to about one-fourth of their ex- tended length by the expulsion of fluid, became opaque, and sank slowly toward the bottom. If collected by being dipped up with a quantity of sea water in a finger bowl without being disturbed, they continued their swimming movements in the dish, jarring of the dish caused extrusion of fluid, contraction, and sinking of the animal. The change from the glass-like transparency of the extended form to the white opacity of the contracted form is extremely striking. The animal swims only when in the extended state. The effective stroke occurs after the animal has assumed the shape of a wide U (see Fig. 1 ). The longitudinal muscles on one side of the body suddenly contract, drawing the posterior end of the body up to and past the tentacled anterior end, and then, as suddenly, relax. Re- laxation results in the resumption of the U-shaped form, and is followed by another contraction of the longitudinal muscles. A series of these contractions creates an effective swimming movement, comparable to the breast-stroke of human swimmers, 93 DONALD PAUL COSTELLO in the direction <>! the base <>l the U. It, <>n one stroke, the head crosses above the ])o.sterior region, on the next it usually crosses below, so that there is an alternation of movements, presumably brought about bv the contraction of sets of longitudinal muscles of opposite side's of the body. At the same time, the whole body turns frequently, so that locomotion is not in one direction, but is more or less aimless. N'evertheless, the swimming leptosynaptids can move across a ten-inch finger bowl in the course of four or live minutes. abed FIGURE 1, a, b, c, d. Successive stages of swimming movements of Lcptnsyiniptti inlnicrcus. The effective stroke, directed toward the base of the U, occurs between a and b, and between c and d. During the period between b and c, there has occurred a rotation of the entire body, on the axis of the U, through 180 degrees. On the following evening (August 15) nine more leptosynaptids were observed swimming in Eel Pond between 9:25 and 10:00 P.M., and were collected. At 10:00 P.M., on entering the darkened laboratory, I observed that four of the leptosynaptids collected the previous evening (and left in a bowl of standing sea water on the sea water table since that time) were in the extended condition and undergoing rapid swimming movements. These animals had remained in the contracted state at the bottom of the dish during the day (see Fig. 2). Similar observations were made for these and other animals on subsequent evenings. Swimming is undoubtedly a Photograph of live Leptosynapta inlnicrcns. living young adult ".swimming forms. are in the completely contracted, opaque state; the fifth almost completely con- tracted, and partially transparent. Centimeter scale indicated below. SWIMMING OF LEPTOSYNAPTA 95 dark-conditioned phenomenon, yet placing leptosynaptids in the darkroom during the day did not induce swimming. Presumably either a certain period of illumina- tion or of complete rest is first necessary. Quatrefages (1842) early called atten- tion to the greater general activity of Leptosynapta at night as compared with that during the day. On the morning of August 16th, at 11 :00 A.M.., one leptosynaptid in a large finger bowl (covered with cheesecloth, over which a slow stream of sea water was flowing) was still swimming, but did not show the alternate criss-cross- ing of the two ends as observed in nature. This is the only Leptosynapta observed swimming during the day. A complete record of Eel Pond collections is as follows: Date Time Number collected Aug. 14, 1944 9:15- 9:30 P.M. 10 15 9:25-10:00 9 16 2 17 9:10-9:40 5 18 none 22 10:00 P.M. 1 9:30 P.M. 2 28 none 29 30 31 Sept. 4 6 7 1944 Total 29 Aug. 25, 1945 10:30 P.M. 1 Sept. 6 9:00 P.M. 2 7 1 8 1 9 8:45 P.M. 1 1945^Total 6 The 1945 specimens were collected by Miss Catherine Henley and Miss Roberta Lovelace. Nereis collections were made almost every night from early June until the end of September, 1945, but no leptosynaptids were seen other than the six listed above. The last specimens taken in 1944 were larger than those first col- lected. Those kept for two or three weeks in the laboratory grew slightly and showed an increase in number of tentacular sensory cups. These facts indicate the probability that Leptosynapta swims only during a limited period of its young adulthood. An examination of the collected specimens revealed only a few grains of sand, and some diatoms and other debris in the intestine. Presumably a necessary con- dition for swimming is absence of intestinal sand which would prevent floating. When the animals swell to the extended condition, the intestine is not filled with water. The intestine can be seen within the body and is approximately of normal diameter. The water probably enters the perivisceral cavity. Tests were made to ascertain the places of exit of the fluid from the body by dropping extended animals into a suspension of Chinese ink in sea water. These tests were somewhat incon- 96 DONALD PAUL COSTELLO elusive, but indicated a general release of fluid along the major portion of the body surface, rather than any markedly localized release. A preliminary examination of the anchors, anchor plates, and other spicules indicated size characteristics somewhat intermediate between those of the two common Woods Hole species, Leptosynapta inliacrens and Leptosynapta roseola. The radial plates of the calcareous ring were seen to be perforated. Some of the material was then sent to Dr. Elisabeth Deichmann, who kindly identified the swim- ming forms as Leptosynapta iulwcrens (O. F. Miiller). This species, as found in the Woods Hole region, was described by Clark (1899). The deviations from the typical adult characteristics which were found in the swimming forms were deviations characteristic of young individuals. Leptosynaptas have apparently never before been seen swimming in Eel Pond, even by such a veteran collector as Mr. George M. Gray. Fish (1925), who studied the extensive plankton collections of the Woods Hole Fisheries Labora- tories, records the presence of only one leptosynaptid — as follows : "A specimen of Leptosynapta inhacrcns (Miiller), 20 mm. long, was taken on Sept. 19 after a hard northeast storm. This was not a free-swimming form and would not normally occur in surface collections." Dr. George L. Clarke, of the Woods Hole Oceano- graphic Institution (personal communication), has never taken a holothurian in his plankton hauls. Swimming of certain other holothurians is not unknown. There is a pelagic form, Pelagothitria natatri.r, /rom the Eastern Pacific, described by Ludwig (1893). For Sticlwpus natans, swimming movements were observed by Sars (quoted by Ludwig, 1889-92). Prior to the present work these constituted the rar.e excep- tions to the general rule that adult holothurians do not swim. LITERATURE CITED CLARK, H. L., 1899. The synaptas of the New England coast. Hull. l~. S. l:isli Coiiiiuiss'nui for 1SW, pp. 21-31. CLARK, H. L., 1907. The apodous holothurians. Smithsonian Contributions to Knowledge, 35: No. 1723, 1-231. FISH, C. J., 1925. Seasonal distribution of the plankton of the Woods Hole region. Bull. V. S. Bureau of fisheries, 41 : 91-197. LUDWIG, H., 1889-92. Die Seewalzen. Bronn's Klass. u. Ordnungen d. Tier-reichs. Echino- (lermen. Bd. 2, Abt. 3, Buch 1. S. 1-460. Lrrmir,, H., 1893. Vorlaufiger Bericht iiber die auf den Tiefsee-Fahrten des "Albatross" (Friihling 1891) in ostlichen Stillen Ocean erbeuteten Holothurien. Zool. AHZ., 16: 177-186. QUATREFAGES, A. DE, 1842. Mcmoire sur la Synapte de Duvernoy (Synapta Dnvernaea A. de Q.). Annales dcs Sciences Nat., 2e Serie, 17: 19-93. CORRELATION BETWEEN THE POSSESSION OF A CHITINOUS CUTICLE AND SENSITIVITY TO DDT * A. GLENN RICHARDS AND LAURENCE K. CUTKOMP 1 Zoology Department, University of Pennsylvania; Marine Biological Laboratory, and Division of Entomology, University of Minnesota INTRODUCTION The remarkable insecticidal qualities of 2, 2-bis (p-chlorophenyl) 1, 1, 1, trich- loroetbane, more commonly called "dichloro-diphenyl-tricbloroethane" or just "DDT," were discovered by Paul Miiller of the J. R. Geigy Company of Switzer- land a little more than five years ago (Lauger, Martin, and Miiller, 1944). The first samples reached this country in August, 1942. Researches in Switzerland, Germany, England, and the United States, accelerated by war-time urgency, have developed a voluminous amount of data. Most of these data are purely practical and will not be reviewed here ; 2 further, these economic reports and papers are appearing so rapidly that it is very difficult to separate fact from fancy in the con- fusing maze of seeming contradictions. Certainly numerous popular articles and stories are dangerously misleading (see, e.g., warning by Cox, 1945, and Anony- mous, 1945). The fact remains, however, that DDT is a remarkably potent in- secticide and that from the human point of view it can be used with relative safety for certain insect pests (flies, mosquitoes, lice, bedbugs, etc.). Since this new compound is so extremely toxic to many insects, one is naturally led to inquire what its effects are on other animals. Both a desire to find more favorable animals for studying the physiological effects of DDT, and a need to evaluate the effects of the compound on the entire biota of natural waters, led to determinations of the toxicity of DDT to representatives of various animal phyla. During the course of making these determinations it was noted that animals with exoskeletons, especially chitinous exoskeletons, were notably susceptible. Experi- ments were accordingly performed to test the validity of this correlation. From these experiments it was concluded that the correlation is valid. A hypothesis is developed in this paper that chitinous cuticles selectively concentrate DDT from the bathing media by adsorption and result in a higher concentration of the toxin inside the animal. There is good evidence to indicate that the relation of the chiti- nous cuticle to the selective action of DDT is no more than that of a concentrating mechanism. The actual lethal action of DDT seems to be another problem (see, e.g., Yeager and Munson, 1945). * Paper No. 2267, Scientific Journal Series, Minnesota Agricultural Experiment Station, St. Paul 8, Minnesota. 1 Present address of both authors : Division of Entomology and Economic Zoology, Univer- sity Farm, St. Paul 8, Minn. 2 Large numbers of papers can be found, especially in the February, 1944 and June, 1945 issues of the Journal of Economic Entomology. The review in Soap is a useful digest (Anony- mous, 1945). 97 A. GLENN RICHARDS AND LAURENCE K. CUTKOMP The work described in this paper was done under a contract, recommended by the Committee on Medical Research, between the Office of Scientific Research and Development and the University of Pennsylvania, and, later, the University of Minnesota. Most of the data were incorporated in a paper presented at the meet- ings of the Entomological Society of America in New York City, December 14, 1944. MATERIALS AND METHODS For the purpose of this study only aquatic species or aquatic stages were em- ployed. The toxin was applied as what seems to be a collodial suspension in the bathing medium because the solubility of DDT in water is very low, at room temperatures too low to show significant toxicity to susceptible animals. Mechani- cally shaking DDT in distilled water (68 hours), filtering, and then centrifuging with an angle centrifuge to remove minute particles (15,000 G for one hour),3 gave a solution that slowly killed about 50 per cent of the mosquito larvae tested at 15° C. but had no significant effect on larvae at 28° C. Several slightly different tests gave similar results. Comparing these data with data from bio-assay of known dilutions of collodial suspensions prepared from acetone solutions, we con- cluded that the solubility of DDT in water probably lies between 0.2 and 1.0 parts per billion when prepared by agitation in water.4 Presumably suspensions from acetone solutions would give much higher solubility values, perhaps one hundred times as high (Lewis and Randall, 1923). The exact physical state, then, of the preparations we are calling "suspensions" is uncertain but at least the preparations are sufficiently homogeneous to dilute satisfactorily. The suspensions were prepared by adding a known volume of an acetone solu- tion of DDT beneath the surface of a known volume of water (distilled, fresh, or sea water). A suspension of one part per million that will give reproducible results and dilute satisfactorily can be prepared by adding one ml. of a 0.1 per cent DDT solution beneath the surface of a liter of water. Stronger suspensions were usually made from more concentrated DDT solutions. Suspensions as strong as one part per hundred thousand are not stable and do not dilute satisfactorily. Suspensions of one part per million in sea water diluted with distilled water and tested by bio-assay with mosquito larvae indicated that the suspensions in sea water have essentially the same characteristics as those in fresh and distilled water. Since all suspensions contained a small amount of acetone, at least two controls were always used. One was a control in the water indicated ; the other, the same plus whatever was the highest per cent of acetone in any suspension. In general, most animals were indifferent to 0.1 per cent acetone. The few cases where 0.1 per cent acetone seemed to have a possibly deleterious effect (e.g., certain Coelenterata) , tests were repeated with lower acetone concentrations. The more dilute suspen- sions, prepared by dilution of a one part per million suspension containing 0.1 per cent acetone, had only negligible amounts of acetone. * This is a force sufficient to sediment bacteriophage particles of approximately 20 mM diam- eter but should not affect molecules several times larger than the molecule of DDT. 4 This is very much lower than the one part per million figure given by Neal ct al. (1944). Our bio-assay datn, a-, wi-11 as similar data from other laboratories, are not consistent with this high estimate. CUTICLE AND DDT SENSITIVITY The amount of fluid used varied with the size of the animals. Protozoa and certain other forms were tested in watch glasses containing 10 ml. of fluid. Other forms were tested in shell vials or finger bowls containing 25-200 ml. of fluid. Of course, single series and replicates were with uniform volumes. Continuously flowing suspensions were not feasible ; the more difficultly cultured marine animals were changed to fresh suspensions twice daily. For species that did not require frequent changing of the medium, subsequent assays with mosquito larvae or hermit crabs were commonly employed to verify the continued toxicity of test fluids. Specimens were obtained from various sources. Protozoa were from labora- tory cultures and were identified by Dr. D. H. Wenrich. Marine forms were collected at Woods Hole, mostly by the Supply Department, and the names used are those current at that laboratory. Other forms were collected or purchased, and identified by the authors. Chemical tests for the presence of chitin were routinely performed on exo- skeletons using the methods given by Campbell (1929). Positive results were obtained with Arthropods and the perisarc of Colenterates. The Bryozoan used (Biigula) gave atypical results suggesting a skeleton composed of something similar to, but not, chitin.5 For adsorption tests cuticle was removed from Horseshoe Crabs (Limit I us Polyphemus), manually cleaned of cells, rinsed thoroughly with distilled water, and dried. Part of this material was treated with 5 per cent NaOH at 100° C. until the color was removed and the pieces were negative to protein tests. This latter material is referred to herein as "chitin" although tests showed that a small percentage of the material had been changed to chitosan (see Campbell, 1929). Smaller amounts of cuticle were prepared from cockroaches and used in a few experiments, but most of the adsorption work was done with the more available cuticle of Limiilus. All tests were performed in replicate. Many of them were repeated. RESULTS The results obtained are summarized in the accompanying table. In all cases the toxin (DDT) was applied as a colloidal suspension in the bathing medium. The resistances or susceptibilities shown refer only to the indicated concentration in the surrounding medium. The figures are not comparable to median lethal doses and should not be interpreted as implying that "resistant" animals would be re- sistant to the injection of solutions or emulsions. 5 The cuticle of Bugula is variously referred to in the literature as "chitinous" or "material akin to chitin." We applied the various chemical tests used by Campbell (1929) and others. The cuticle is not dissolved in hot concentrated KOH. After prolonged heating (which changed control pieces of known chitin to chitosan completely) it only crinkled in 3 per cent acetic acid instead of dissolving. The addition of one per cent HLSO, gave no change and no precipitate. Another piece treated with I + KI in water turned brown but only very slowly. After the addi- tion of one per cent H,SO4 it remained brown for several minutes and then after 5-10 minutes slowly changed to a greenish-brown (instead of the clear violet given by chitosan). It dis- solved immediately in 75 per cent H,SO4 but after standing no crystals resembling chitosan sulfate crystals were found. These tests suggest a similarity to chitin but certainly, at least in this species, the bryozoan cuticle is not true chitin. 100 A. GLENN RICHARDS AND LAURENCE K. CUTKOMP I o u cu g- O II -O CD CU C n •** a •- ~ "2. cu _3 % ~ t ^ Qi r" * ** QJ ir", 3C * '^ « x u 12 W 3 w ca H '— C 'rations en •? | 3 o ^J Q, Q •^ gT3 ^j en 4-1 ^ 5 o) CU I1 o 'G C D O D. 4_, efi S 5 03 3 "y ^ ~ rf "5. u cu i! — O en _ ; O O° O ^ 6 E B E cfl S •go c - « •a &*-* o" t "— ' sell • • ^o CJ CJ CJ O O O CU 4) S***T* IQ CC3 CG •o icic tc *i *o *r ^ ft) CU CU fl> 4) Ol CU CU 01 O| ^ o^ o o o " 0 0 o o = ~' 33 Z Z Z W ZZ ZZ14 6 rt 6 d E •O E R 0 SS^J r^i 8 CJ ^ •a-o-d j. *-» 4-> CJ CJ o o-i° 01 Ol - £ *S -n *j *-» 4-1 CU CJ CJ CJ •a cu Ol CU 'S'S^o o o ^ *jttj to n^ o o 0 Op — c (73 £ ft i'f,'f,'f_ '^ zz o '35 ,1 c f Ol a ^o ^^ a •a *e]o O _>, en 3 —i a •oE aj H a n o o o V ~ ~ ^ « ^, -a u o u CJ CJ Ol •a t -0*1*1*! ^Cu CJ jy 01 Cu CU -o o •°T ^iA ej cj Cu Ol *:*! CJ Ol 4J W— fN U cu *: cu O C3 •T~l ^ 0 o:_; i Ol eu »j . 's'sK 0 X XX XXX rs te 5 O O o W < ^ZZZ sT o o ZZ ,° ° 0 0 z 3^ c _o 2 o 3 i 2 c o" u — JI'.S 4^ -> <-> , — , o a CJ CJ C O o o 0 Ol te cu •a S-^-oiStotS CU ^J •Zr CJ CJ QJ OJ o 6 u o> *1 cu a cu -z-v o U ^_ - •^ _ ^^H ^.^H^- p , CU E EE SEE o o o o o o lO ""> l/j U") IO U"j E E EE E E ,CU S 3(j O 0 O O O 0 o o o o o o o o o° O O 0 o o o o o 0 C3 d '- x ~ d X Q d Oft ") d jg ^ K O i_ x ^|| 1||§ a en s -s d §• fs cj .~ ~ a w ° a ^ 5 8 o d>H s d 1 c^ ~ ~C ~ n *•- "S 5.2 'S'^-S 5 a g 00 O l-s| «? s R w ^> l-s! ^ 2 ""2 C3 "S DO •c K <^^* a >s &,!?, i^ua. ft, -^^ ft,C)^ 0 0) en 01 J^ a ti ca a *^ ,J3 jr 3 c o fc "cu •b o 2 CJ IH ca O N o o l_, 1 cu o « E Ol astro o N O X a. U a! X * O « /. •a = tf a 3 '-3 »> en O _>^ u •3 be s> 4_» 2 (J # .c 4-> CT 3 CD c en Q, 'f en O en — ' . rt' *£ .CU O >,"n fj a— . ^^ cu 13 X "^ a en 3 S'3 5 "3 C ^Q O3 '-•-' CU O "O Q._ S u ^ o o _o g^ « CJ 4-* aj cu J3 • O -r-l S S^ o •^= -n « D .5 w J= -n cu en ^ §| c o S "o O O 3 cu ._, en 0 c cu J_, o3 o3 o 11 OJ CU U. c •s S t{ S *^- .a Seg * c ^ CD r- Q i_ !—••-* en C3 OJ fcX) CU *^4 G M s u S •- s s a tn 4S S rt •" io ag^ T T3 "° CD u CU £ ^,3 •^^ , _, c2 ^ T3 C n3 «T3 rt S 0-2 H t^* 4_> en ^ t/1 ^^* en cu" Q g^ CU 4-1 u ^•» J^ en ^i 03 C O T3 TD c cu c cu ^ y 03 'tj cu O cu tn "^ HjS- •^-« , ^^ en •±- "Zsfi < "3 rO 4> ^ ^ be 4- ^ §^ « cu - •c ^ f— i O - CU M U « W 4-1 r^ ) f -C .ti en .X f rr C .S c CUTICLE AND DDT SENSITIVITY 101 4-1 cn nl i- h j= 0 « •5 _> £ d d ca Clj W L^ § _5j _>> "t? •O •a •a •a cu g cn c^ o a .-> 0 w"0 7^ 1/5 V V «-l 4-* M " § •5*0 ^ii ^ et) CU CU te-o cu cu *** cu •ota 1^ CU 1 3 O CJ cu cu teto cu cu » -5 'g oS O "S m o S S rs O 0 j~| _« *-• S z2 2 w UJ zz S •£ 8 0 cd -O^-v ca •o ca ca •o ca •a cu ^ r* 4-1 •3 T3 rt o ca Cl'-n in ^ w IN CTJ M o" o s CJ CJ cu cu ss o- cu^"' o> V S 4-» bo — • cj cu 3 >. cu C •** "3 IB »• • te te *o "O «^ •a T3 te cu 1^ cu cu cu cu cu CJ (U l; 4^ cu cu cu cu CU 1- o 0 0 o~ ^) _-. 1C O^3 O — 12 •^ ~ o 42 TO — C Z zz ZL4 wz^^: t-4 w W w z t— i cu ~ C o 0 g C c ca ca V <* ca ca tt • ^ ctf § •a Is •O" •o T> **• -i-J C/) m 00 o" r*-. vn •a v"^ '7 •* CN *^j rt Q H o 55 CJ CJ cu cu CJ CJ cu cu U CJ cu cu c C ^' "^ d ^> cu *Q r; 'H ^ >^ Q te te cu cu tete cu cu te te cu cu "O V •OT cu c 3 -a J CU •o -d cu cu te cu ^ S 5^o Q — O 0 0 0 o o P2 s — r. :S ~ ~ o "^ Si ^ ^ d "o z z zz z z W W W- dU! w w z § 2 -S-l CO g o^ — ^ }r^ t* *^ O •*-> 8 C« •a • •a ca o -a • 3 1 a ca ca ca ca ca ^ ^ QJ M •S g > S.H "-2 O O ca IN o_ *) CN » c <• i ,j, .4. _ r^ ~, *-> e o" _ W «j «-• 4J "f 1 A-. i w rj -< v — — ^2 — ' o C Q cu CJ CJ cu CU CJ cu cu •^ cu" -'j^ ^^ ^_j W S c O 0 te cu cu cu te te cu cu 7, • O T3 ' O ten ?10 •O -3 -O -a-O -O -O ^ 4^ d^ ty CU CU QJ *^> o* 5 •*-* W ^ cu $ 4 k u ^^ o z 0 0 zz 0 0 z z -^ i ^ L4 ; S or t z;. rt _^ cu !~! 11 ' jaj rt _*4 cd •O T3 T3 "d 1 S^ E 3^ >> *3 »> ^ tu ^ , y cn * .^ lag J ™ i— - — - J2 -T-J cn O O cD >. « - '5 £ .n J= S ° g c a o c i >> *• ^— * ^ < ^~. ^-^ 3 .h "3 S S £ ^ X ca ca C g "3 J -S -> u ^ o 1-3 oo r*-. 3 « £ 15 "° 8 I In 1 OCN |i|s CJ OJ 1C W ^•ctc tu-Ste l~l 1 1 &,£ c -c'l SR g '* s.s M 2£ -— P x ^ ^ *-* j-j. 'o 1^ ' ^ *"* c^J "^ (^ 3 ^« S > cu .rt rt c J2J2 j cn cn 1 cu cu J k« L- ca ca cu cu j^ jn r/) , . C/l flj C3 Cy fli i- CU CU j- ca CU en co cu cu jr J3 to cn CU (U ca ca Cfl CO Cfi X 73 W 7) cu cu cu cu i> u cu ™ X CO cu cu cu '•S js u " S to SI cn 73 cu *j O •— >I- f"1 cn c CO CO '^^ cn t£ '*- en •—'_.— 'w O *O C3 4-1 "^ • 1 cu" g E E E E g EEE w lH «= ta "^Q He BU S S O O •>ir>m 0 0 GOOD 0 o o m oo 0 G •o o O 0 IO 1/1 o o o o o o o iO 00 »O O o O O o o o o o o rt w 'S '-u ^ S ^ ca O - e *-> c-3 3 "° g c m c <« S- s ^, (U c CU -n O. K C3 •3 13 •5 cn ^ "2 <£ _§ "3 _a O.". u, a K d x\ d '5. « " &« 'Hd2 .ao m u ( j D Q n Littorina sp Helisoma In Physa sp. Thyone sp. Aslerias sp. "> O.R tt CO «U «0 .t? 8 «•- 3 ^ | 2 »-ii^ 8-§-2J5£ •i^sst ^J O ^ S ^a,fe:ts., 32 p. RICHARDS, A. G., 1944. Notes and news in entomology (brief review of recent advances in study of insect cuticle). Entom. News, 55: 18-21. RICHARDS, A. G. AND J. L. WEYGANDT, 1945. The selective penetration of fat solvents into the nervous system of mosquito larvae. Jour. N. Y. Entom. Soc., 53: 153-165. SEAGREN, G. W., M. H. SMITH AND G. H. YOUNG, 1945. The comparative antifouling efficacy of DDT. Science, 102 : 425-426. SMITH, C. N. AND H. K. GOUCK, 1944. DDT, sulfur and other insecticides for the control of chiggers. Jour. Econ. Entom., 37: 131. SMITH, M. I. AND E. E. STOHLMAN, 1945. Further studies on the pharmacologic action of 2,2-bis (p-chlorophenyl) 1,1,1-trichlorethane (DDT). U. S. Public Health Rf>ts., 60: 289-301. VARGAS, L. AND R. C. IRIS, 1944. Accion del DDT sobre algunos arthropodos domesticos. Rev. Soc. Me.v. Hist. Nat., 5 : 229-235. WHITE, W. C. AND T. R. SWEENEY, 1945. Metabolism of 2,2-bis (p-chlorophenyl) 1,1,1-tri- chloroethane (DDT). I. A metabolite from rabbit urine, di (p-chlorophenyl) acetic acid; its isolation, identification, and synthesis. U. S. Public Health Rpts., 60: 66-71. WIGGLESVVORTH, V. B., 1941. Permeability of insect cuticle. Nature, 147: 116. YEAGER, J. F. AND S. C. MUNSON, 1945. Physiological evidence of a site of action of DDT. Science, 102: 305-307. PROTOPLASMIC VISCOSITY CHANGES IN DIFFERENT REGIONS OF THE GRASSHOPPER NEUROBLAST DURING MITOSIS J. GORDON CARLSON Department of Biology, University of Alabama l * Protoplasmic viscosity has been studied by a variety of methods. Each has certain advantages and certain limitations, with regard to the kind of living ma- terial to which it is best suited and the accuracy of the results that it will give. These methods and their uses have been reviewed critically by Heilbrunn (1928, 1943) and therefore will not be considered in detail here. Of the several that have been employed the brownian movement method is probably best suited to viscosity studies involving limited regions of the single cell. It may be applied in either of two ways: 1) to calculate the' absolute viscosity by measuring the displacement in one direction of the molecularly-bombarded particle and the time required to bring this about, certain other characteristics of the cell and its environment being known, or 2) to compare the viscosities of different regions of a cell at different mitotic stages or under different experimental conditions through observations of the relative speeds of brownian movement. The latter has been used in the present investigation. The neuroblasts of the grasshopper embryo possess several advantages in such a study. Since they are situated on the ventral side of the embryo, they can be brought next to the cover glass in hanging-drop preparations and microscopic observations of the mitotically active, living cells can extend over several hours. All the cell features that are visible in the usual fixed and stained preparation can be seen, and many phases of the mitotic cycle can be identified readily. The cell is relatively large, measuring about 30 p in diameter. It maintains a visible polarity from one mitotic division to the next, so that a given region can be located in any cell at any stage of division. The cytosome contains large numbers of mitochron- dria, which not infrequently find their way into the spindle during mitosis. By observing the brownian movement of these tiny bodies it is possible to compare the viscosities of the surrounding protoplasm of all non-nuclear parts of the cell at all stages of the mitotic cycle. MATERIAL AND METHODS Embryos of Chortophaga viridifasciata (De Geer) were prepared by the hang- ing-drop method previously described for this material (Carlson and Hollaender, 1 The preliminary observations on which this study is based were made as Rockefeller Fellow in the Natural Sciences at the Genetics Laboratory of the University of Missouri in the winter of 1940-41 and at the Biological Laboratory, Cold Spring Harbor, in the summer of 1941. The work has been completed with the aid of a 1944-45 grant from the University Re- search Committee of the University of Alabama. 109 110 J. 1,<>KIX)N CARLSON 1944). The culture medium, which must he isotonic with the emhryonic cells, is made up as follows : NaCl 0.70 gm. KC1 0.02 gm. CaCL 0.02 gm. MgCl2 0.01 gm. NaH2PO4 0.02 gm. NaHCO:; 0.005 gm. Glucose 0.80 gm. H,O (pyrcx redistilled ) '. . . . 100.0 cc. The pH of this solution is approximately 6.5, which is about the pH of the grass- hopper egg yolk. A small amount of yolk is added to each hanging-drop to pro- vide nitrogenous food materials for the cells. Preparations were studied in a constant temperature box enclosing all of the microscope except the upper part of the body tube and arm. All observations were made at 26 ± 0.5° C. The light used as a source of microscope illumination wa^ passed through copper sulfate solution to filter out the heat. Observations were made exclusively on neuroblasts. These cells are shown in representative mitotic stages in Figure 1, which is based on camera lucida sketches of living cells. It will be noted that the neuroblast deviates from the typical cell in two main respects. First, the nucleus is not spherical but roughly hemispherical with a central cytoplasmic core connecting the polar and apolar regions of the cell (Figs. I A and IB). Second, the cell divides unequally to form a small daughter ganglion cell and a large daughter neuroblast (Fig. 17.) The great advantage of these cells is their large size, which makes it possible to observe and study in the living, unstained cell many structural features that are very diffi- cult or even impossible to deal with in smaller cells. The mitochondria of the neuroblast are spheroidal in shape and measure ap- proximately 0.3-0.5 /.i in diameter. TERMINOLOGY Delation <>j viscosity to broivnian movement. The relation between the rate of brownian movement and viscosity is given by Einstein's equation RT( NSirrja in which Dx is the displacement of the particle along the x axis ; R, the gas constant ; T, the absolute temperature ; t, the time ; N, the Avogadro number ; a, the radius of the particle ; and 77 the viscosity expressed in poises. Accordingly, the viscosity of the dispersion medium varies inversely as the square of the displacement of a given particle along one axis. The rapidity of brownian movement of the mitochondria in selected regions of the cell was observed and classified according to the following system. The VISCOSITY CHANGES IN GRASSHOPPER MITOSIS 111 D G H FIGURE 1. Representative stages in the mitotic cycle of the grasshopper neuroblast, re- constructed from camera lucida sketches of living cells growing in culture medium. A and B, polar and side views, respectively, of interphase or prophase cells; C, late prophase a few min- utes before breakdown of the nuclear membrane, the cell having acquired a spherical form ; D, prometaphase ; E, metaphase ; 1~ , middle anaphase; G, late anaphase; H, early telophase; /, late telophase, the cell having lost its spherical form, a, polar cytoplasm ; c, equatorial cytoplasm ; g, ganglion cell; h, half spindle; i, interzonal region; n, nucleus; nc, nuclear core cytoplasm; p, polar cytoplasm. 112 J. GORDON CARLSON viscosity values and the relative speeds of brownian movement are designated as follows : Very high: Xo visible movement of mitochondria. High: Movements very limited, discernible only with prolonged observa- tion. Medium: Movements readily observable, but not rapid. Low: Movements quite free and rapid, but individual mitochondria can be followed with the eye. Very low: Movements so rapid that individual mitochondria cannot be followed with the eye. Even if the ability to make purely objective distinctions between adjacent members of this series be open to question, values separated by one class, viz., very high and medium, high and low, or medium and very low, can be distinguished readily and would give results essentially similar to those obtained. Mitotic stages The stages for which data were obtained and their distinguishing characteristics in the living, untreated state, when observed through 12.5 X compensating oculars and a 2 mm. oil immersion objective, are : Interphase: Nuclear threads appear in optical section like small, rounded, scattered granules in an otherwise homogeneous nuclear background. Period of growth. Very early prophase: Nuclear granules smaller in size. Threads have be- gun to appear in the previously homogeneous background. Early prophase: Nucleus filled with extremely fine chromatin threads. No nuclear granules. Middle prophase: Chromatin threads of appreciable thickness. Ends when about seven chromosomes in cross-section can be counted in one-fourth the nuclear circumference. Late prophase: Chromosomes well-formed. Toward the end of this pe- riod cell assumes spherical shape (Fig. IB, 1C). Ends with the simultaneous disappearance of the nuclear membrane and appearance of the spindle. Prometaphase (Fig. ID) : Chromosomes straighten and move into the equatorial plane. Spindle develops. Metaphase (Fig. IE) : Chromosomes in equatorial plane. Early anaphase : Begins with the initial separation of the chromatids and ends when the distal ends of the chromosomes leave the cell equator. Middle anaphase (Fig. IF) : This stage is terminated as the cleavage fur- row penetrates to the interzonal region. Late anaphase (Fig. IG) : Ends as the chromosomes lose their sharp out- lines and the cleavage furrow appears to be complete. Early telophase (Fig. IH) : This stage lasts until the nucleoli become visible. Middle telophase: The nucleoli increase in size while retaining their spherical shape, the nuclear membrane re-forms, the interzonal spindle remnant disappears, and the cell re-assumes the hemispherical form. VISCOSITY CHANGES IN GRASSHOPPER MITOSIS 113 Late telophase (Fig. I/) : Begins as the outlines of the nucleoli become irregular. Ends as linear arrangement of chromatin granules is lost. The relative duration of these stages is given in an earlier paper (Carlson, 1941). Regions o$ cell Data have been obtained for six different regions of the cell, which are labeled in representative stages of the mitotic cycle in Figure 1. Polar cytoplasm: This includes the region about the spindle pole of the daughter neuroblast when a spindle is present. At other stages it is the region within which the spindle pole of the daughter neuroblast will develop preparatory to the next division. Apolar cytoplasm : The cytoplasm situated near the place of formation of the preceding cleavage furrow. It is opposite the polar region. Equatorial cytoplasm: This is the portion of the cell that is situated at or near the cell equator in metaphase and anaphase and midway between the polar and apolar regions during the remainder of the mitotic cycle. Nuclear core cytoplasm: The cytoplasmic core that passes through the nucleus (see p. 110). Half spindle: This includes the developing spindle at prometaphase, the fully formed spindle at metaphase, and at anaphase the portion of the spindle lying between the pole and the plane in which the kinetochores of the adjacent chromo- some group are situated. Interzonal region: This is the part of the cell lying inside the ring of in- terzonal fibers and situated between the two separating groups of daughter chromo- somes. It makes its appearance as the chromatids separate at early anaphase and persists as a spindle remnant into middle telophase. Nucleus: No information that seemed reliable was secured for this part of the cell. Tiny nuclear granules comparable in size to the mitochondria showed very limited movement at various stages of the mitotic cycle. This might indicate a persistently high viscosity or it might mean merely that these granules are at- tached to the chromatin threads, which form a semi-rigid framework within the nucleus, and would therefore not be expected to exhibit much brownian movement. Belaf (1929a), however, has reported a rapid brownian movement of tiny granules of unknown nature situated in the ground substance of the nucleus of the grass- hopper spermatocyte during prophase and telophase, indicating, according to him, that this substance is quite fluid. OBSERVATIONS The curves shown in Figure 2 were constructed in the following way. A linear series of arbitrary numerical values was assigned to the five selected viscosity classes described previously. Four observations were made for each region of the cell on different days and with different preparations. These were averaged and plotted. Because the determination of each of these was to a certain extent sub- jective, in that it depended on the judgment of the observer, and because the vis- 114 J. GORDON CARLSON _ (Nl o >- a: ce o A1ISCOSIA VISCOSITY CHANGES IN GRASSHOPPER MITOSIS 115 cosity varied slightly from preparation to preparation and from cell to cell, certain of these curves showed irregularities. In order to discover whether these were the result of experimental error or of actual viscosity shifts, new cultures were pre- pared, checked, and any necessary corrections made. It is believed that the graph in its present form represents quite accurately the cycle of viscosity changes in the different regions of the cell during mitosis. It should be emphasized, however, that the five arbitrarily-chosen viscosity values represented by ordinates are not neces- sarily a linear series of absolute values, as their graphical treatment implies. During the greater portions of interphase and prophase the viscosity of the cytosome maintains a relatively high and constant value. In late prophase, a few minutes before the disappearance of the nuclear membrane, a rapid fall in viscosity is initiated. This change begins as the cell changes in shape from hemispherical (Fig. IB) to spherical (Fig. 1C). The increase in the rapidity of mitochondrial movement is distinguishable first at the juncture of the equatorial and apolar regions as this part of the cell begins to draw in coincidentally with the start of the round- ing-up process. From this region the lowered viscosity spreads into the apolar region, which thickens as the rounding-up proceeds, and subsequently into the equatorial and polar regions. Obviously, some of this is due to the actual move- ment of the cytoplasm, which must of necessity shift in position as the cell shape changes, but this can hardly be responsible for more than a small fraction of the viscosity change that takes place at this time. Since this alteration in shape occu- pies only about 5-10 minutes, this sequence of changes does not appear in the graph. The viscosity continues to fall through prometaphase and metaphase to reach a minimum in anaphase. The progress of the cell through telophase is ac- companied by a steady rise in viscosity until the interphase-prophase level is at- tained. During telophase the hemispherical form of the cell is re-assumed (Fig. 17). The half spindle appears to show a slight fall in viscosity from prometaphase to early telophase, by which time it is too small for further study. As the chromosome halves begin to separate at early anaphase, the interzonal region between the groups often contains one or more mitochondria showing a moderate amount of brownian movement. During middle and late anaphase large numbers of mitochondria move into the interzonal region, until they are as concen- trated as outside, except for a small region inside each of the separating chromo- some groups. Their movement is quite rapid, but less so than that of the mito- chondria outside in the equatorial region. As the cleavage furrow presses in against the interzonal fibers, the transverse diameter of the interzonal region de- creases and the motion of the mitochondria quickly slows down and then stops entirely. The fact that these stationary bodies are lined up in rows parallel with the long axis of the spindle makes it appear as if they had been caught among interzonal fibers of higher viscosity as these were pressed inward by the deepening cleavage furrow. By late telophase these fibers are no longer visible. The high viscosity level of the nuclear core cytoplasm immediately after the formation of the nucleus and core may be due to the fact that this is the region occupied earlier by the spindle remnant of the preceding division. 116 J. GORDON CARLSON DISCUSSION The brownian movement method Of the three most frequently used methods of making determinations of proto- plasmic viscosity changes, viz., centrifuge, microdissection, and brownian move- ment, the last has the advantage of producing no physical disorientation of the cell contents. It can he depended on to give reliable results, however, only if certain factors are taken into consideration : 1) The protoplasm surrounding the granules in brownian movement must be homogeneous, if the viscosity of the protoplasm as a whole, exclusive of the gran- ules, is being determined. If regions of different viscosity are present, observa- tions of the rate of brownian movement of a particle will indicate only the viscosity of the material immediately surrounding it. The polar, equatorial, apolar, and nuclear core regions of the neuroblast cytosome are visibly homogeneous except for the mitochondria, and the fact that these bodies appear to migrate about and pass one another in the course of their shorter zig-zag movements demonstrates that they are not enclosed in a substance of one viscosity that is in turn surrounded by material of another viscosity. The spindle, however, is probably an exception to this (see p. 118). 2) Alterations in the size of the mitochondria from one stage of mitosis to another would cause an apparent shift in the viscosity of the surrounding medium, even though no actual change took place. Such changes could conceivably result either from shrinking and swelling as a consequence of osmotic shifts or from division and growth. The latter can be ruled out in the case of the neuroblast because very few mitochondria would need to divide in each cell generation to make up for the loss to the daughter ganglion cell, which receives very little cyto- plasm and very few mitochondria. The resulting effect on brownian movement would be insufficient to alter the general viscosity picture. With regard to osmotic changes, Lewis and Lewis (1915) report that immersing cells in hypotonic solution causes swelling of the mitochondria. Belar (1929a), on the other hand, states that mitochondria are particularly insensitive to the usual swelling effects of hypo- tonic solution. Unfortunately, the neuroblast mitochondria are so small and move so rapidly at certain stages that exact comparisons of size by actual measurements are impossible. Careful visual comparisons of the mitochondria of adjoining cells in different mitotic stages, however, have failed to reveal any significant size dif- ferences. As far as the present study is concerned, swelling of the mitochondria of late prophase, prometaphase, and metaphase cells by intake of water (p. 119) would lend to reduce rather than augment the observed viscosity shift in these stages; therefore, the actual change would be even greater than that shown in Figure 2. 3) If the granules under observation are so crowded as to interfere with each other's movements, the viscosity values obtained at the lower levels will be too high ; for frequent collisions will retard their movements. The mitochondria of the neuroblast are definitely crowded, and this could easily result in appreciable errors in absolute viscosity values based on the migration of a given body a certain dis- tance in a certain time, according to the method described by Pekarak (1930). Much of this error is doubtless avoided in the present study, however, because viscosity determinations were based on the rapidity with which the particles as a group dance about rather than on the total distance they migrate in a given interval of time. VISCOSITY CHANGES IN GRASSHOPPER MITOSIS 117 Results of related studies Of the several studies made on viscosity changes of the whole cytoplasm during mitosis, the centrifuge determinations of Heilbrunn (1917, 1921) for the marine invertebrate egg (Arbacia, Cumingia, and Nereis) and of Kostoff (1930) for meiotic cells of Nicotiana and the brownian movement determinations of Kato (1933) for the meiotic cells and staminate hairs of Tradescantia have given results that are similar to those I have obtained for the polar, apolar, and equatorial cyto- plasm of the grasshopper neuroblast. All these studies show that the viscosity is relatively high during interphase, when interphase is referred to at all, high during all or most of prophase, falling during metaphase, lowest at anaphase, and rising at telophase.2 The similarities in these results are all the more striking in view of the great diversity of the cell types used. The marine invertebrate egg contains relatively large quantities of ergastic matter, such as oil globules and yolk and pig- ment granules, and develops during mitosis an astral system and spindle that in- volve a large portion of the protoplast. The pollen mother cell, though consisting of a more or less homogeneous cytosome with relatively small spindle and no astral system, nevertheless is atypical mitotically because of the extended prophase and the two successive meiotic divisions without an intervening interphase. The stami- nate hairs of the plant contain large vacuoles with only a peripheral film and a few central strands of cytoplasm. The grasshopper neuroblast contains a fairly homogeneous cytosome with no visible ergastic matter or vacuoles but an abundance of mitochondria. The spindle is moderate in size and no asters are visible. This suggests strongly that viscosity change is a fundamental factor in mitosis, and is largely independent of individual pecularities of the type of cell studied. Zimmermann (1923), who studied the viscosity of the alga, Sphacelaria, by means of the centrifuge and brownian movement, and Seifriz (1920), whose studies of the marine invertebrate egg are based on microdissection, place the viscosity fall and minimum somewhat in advance of this mitotically. According to Zimmer- mann the viscosity is lowest at metaphase, while Seifriz finds it lowest during late prophase and metaphase. Both describe the viscosity as rising at anaphase and high at telophase. In contrast with these findings are the results obtained by Chambers (1917, 1919) from studies of Arbacia and Echinarachnius eggs with the microdissection needle. He reported that the viscosity of the greater part of the cell is low during prophase, rising during metaphase, high during anaphase and early telophase, and falling in late telophase. He related the viscosity to the state of development of the amphiaster : low when no amphiaster is present, rising as the amphiaster forms, highest when it has reached its maximum development, and falling as the amphiaster disappears at the end of mitosis. He does, however, describe a liquefaction of the equatorial region previous .to anaphase and persisting through cleavage. Fry and Parkes (1934) duplicated as closely as possible the centrifuge studies of Heilbrunn on Arbacia, Cumingia, and Nereis eggs. The results they obtained for viscosity in relation to time after fertilization were identical with those reported earlier by Heilbrunn; in fact, in their paper they used Heilbrunn's curves for Arbacia and Nereis eggs. They interpreted this data, however, as supporting 2 Heilbrunn associates the pre-cleavage fall and the post-cleavage rise in viscosity of the Arbacia egg with spindle development and disappearance, respectively. 118 J. GORDON CARLSON Chambers' conclusions, claiming that Heilbrunn had misidentified certain of the mitotic stages. This would place viscosity changes of the marine invertebrate egg in entire disagreement with those of the grasshopper neuroblast, in which there can be no question of the correct identification of the different mitotic stages. Kostoff s description of viscosity changes in somatic cells of Nicotiana (1930) is not in accord with any of these results. Using the centrifuge technique, he re- ported two cycles of viscosity change for each mitotic cycle : high viscosity at prophase and anaphase, and low viscosity at metaphase and interphase. Observations on the viscosity of the spindle are complicated by the probable presence in the half spindle of two materials, spindle fibers and interfibrillar sub- stance. If there are two such materials present, the mitochondria or other cyto- plasmic granules that make their way into the spindle by brownian movement would be expected to occupy the region of lower viscosity, i.e., the interfibrillar region. Under such conditions their speed of movement would be determined by two factors : the viscosity of the interfibrillar substance and the amount of space available for movement between the fibers, especially if this were very limited. Belaf (1929b) found that most spindles in cells of the stamen hairs of Trades- cantia contained a few tiny granules in quite rapid brownian movement. This motion of the granules, which appeared first in the polar caps (clear regions that adjoin opposite sides of the nucleus at late prophase and in which the spindle later develops), was evident in the fully-formed spindle and continued undiminished up to the time of formation of the cell plate, when the granules disappeared from view. This contrasted with the situation he found in the spindles of animal cells, namely, nematode eggs, lepidopteran and grasshopper spermatocytes, and Actino- phrys, in which granules of a comparable size in the spindle were always relatively quiet (see Belaf, 1929a). He interpreted this difference to indicate that in the Tra- descantia cell there was a larger amount of the less viscous interfibrillar substance than in the animal cells he studied. I have not been able to confirm Belar's obser- vation that the granules show greater freedom of movement in the direction of the long axis of the spindle than at right angles to it, but this may mean only that the interfibrillar substance is more abundant and the fibers farther apart in my material than in his. Ris (1943) reports unrestricted brownian movement of the cytoplasmic gran- ules that make their way into the interzonal region during anaphase in embryonic cells of Tamalia. Chambers (1924) states that the spindle of the dividing sand- dollar egg has become "distinctly fluid" by the time the chromosomes have reached the poles. He doubtless refers to what I have termed the interzonal region. These observations agree well with the conditions in the grasshopper neuroblast. Possible causal factors in viscosity change With regard to the factors responsible for viscosity changes of the cytoplasm, two possibilities seem deserving of consideration: alterations in the water content and in the pentose nucleic (ribonucleic) acid content of the cytoplasm. It has already been pointed out (p. 115) that the first detectable viscosity drop coincides with the initiation of the "rounding-up" of the neuroblast in late prophase and that the viscosity is rising as the cell flattens on one side to assume the hemi- spherical form during telophase. Since it can be demonstrated that immersion of VISCOSITY CHANGES IN GRASSHOPPER MITOSIS 119 these cells in hypotonic culture medium causes a fall in viscosity owing to intake of water and that hypertonic medium causes a rise in viscosity through loss of water, it seems not unlikely that the late prophase viscosity fall and "rounding-up" of the cell might result from the intake of water, while the telophase viscosity rise and the accompanying flattening of the cell at one side might be due to loss of water. Unfortunately, the irregularities in the hemispherical-shaped cell make it impossible to determine whether there is any actual change in cell volume during the alteration in shape. Chambers (1919) relates the change in form of the Arbacia egg from spherical to hemispherical following the first cleavage division to a viscosity shift, but he associates it with a lowering of the viscosity. Water exchange seems inadequate, however, as a complete explanation of the observed viscosity changes for two reasons. First, it seems doubtful that the intake of water would occur in sufficient amounts to account entirely for the observed viscosity fall. Second, the viscosity continues to rise for some time after the cell has returned to its hemispherical shape, when the water loss might be supposed to have been completed. Another factor, therefore, would appear to be involved, either in place of or in addition to water intake and loss. In support of the possibility that changes in the pentosenucleic acid content of the cytoplasm may be at least partly responsible for the observed viscosity changes is the fact that pentosenucleic acid, which is known to be present in large quantities in the cytoplasm of rapidly growing tissues (Brachet, 1933; Caspersson and Schultz, 1940) and which has a strikingly high viscosity (17 = 62.4, according to Cohen and Stanley, 1942), undergoes a change in amount per cell during mitosis (Brachet, 1940; Caspersson, 1940; Painter, 1943). Both Brachet, from studies of the eggs and early developmental stages of different animals, and Painter, using Rheo meiotic cells, reached the conclusion that cytoplasmic pentosenucleic acid is abundant in mitotically active cells at early prophase, less abundant or entirely absent from late prophase through anaphase, and increasing in amount following division. They believe that the late prophase decrease is due to transformation of pentosenucleic acid into the desoxypentosenucleic acid of the developing chromo- somes and that the increase following division is due to the loss of desoxypento- senucleic acid from the chromosomes and its transformation into pentosenucleic acid. These changes in the nucleic acid content of the cytoplasm during mitosis bear a very close resemblance to my viscosity curve for the neuroblast cytoplasm. If there is any discrepancy, it would seem to be the exact time during late prophase that the nucleic acid leaves the cytoplasm in detectable amounts, and neither the studies of Brachet nor Painter furnish information on this point. The statement of White (1942), however, that "the nucleic acid of the chromosomes undergoes a sudden increase at prometaphase, when the nuclear membrane breaks down and substances from the cytoplasm have free access to the chromosomes" suggests that the fall in cytoplasmic nucleic acids may coincide closely with the viscosity fall of the cytoplasm. The viscosity rise, which extends through all of telophase, could be accounted for, up to the time of nuclear membrane formation, by a return of the pentosenucleic acids to the cytoplasm during the retrogressive chromosome changes, and, after re-constitution of the nuclear membrane, by the synthesis of new pento- senucleic acid. 120 J. GORDON CARLSON SUMMARY Observations on the rapidity of brownian movement of the mitochondria in different regions of the grasshopper neuroblast during the entire mitotic cycle indicate that : 1 ) The viscosity of all parts of the cytosome is relatively high during interphase and prophase, begins to fall in late prophase, reaches a minimum at anaphase, and rises gradually to its original high level during telophase. 2) The viscosity of the portion of the spindle between the pole and the plane in which the proximal ends of the chromosomes are situated appears to fall slightly from a high prometaphase level through anaphase. 3) The viscosity of the portion of the spindle situated between the separating daughter chromosome groups shows, during anaphase, a slight drop from a me- dium value, and this is followed at the beginning of telophase by an abrupt rise to a very high level, which is maintained through early and middle telophase. Alterations in water content and nucleic acid content of the cytoplasm are suggested as possible explanations of viscosity changes during mitosis. I wrish to express my indebtedness to Dr. L. V. Heilbrunn of the University of Pennsylvania, Dr. Franz Schrader of Columbia University, Dr. Jesse P. Green- stein of the National Cancer Institute, and Dr. Alexander Hollaender of the National Institute of Health for their kindness in reading the manuscript of this paper and in offering valuable suggestions for its improvement. LITERATURE CITED BELAR, K., 1929a. Beitraege zur Kausalanalyse der Mitose. II. Untersuchungen an den Sper- matocyten von Chorthippus (Stenobothrus) lineatus Panz. Rou.v' Arch. Entivick. Organ., 118: 359-484. BELAR, K., 1929b. Beitraege zur Kausalanalyse der Mitose. III. Untersuchungen an den Staubfadenhaarzellen und Blattmeristemzellen von Tradescantia virginica. Zeits. Zellf. u. mik. Anat., 10 : 73-134. BRACKET, J., 1933. Recherches sur la synthese de 1'acide thymonucleique pendant le developpe- ment de 1'oeuf d'Oursin. Arch, dc BioL, 44: 519-576. BRACKET, J., 1937. Remarques sur la formation de 1'acide thymonucleique pendant le developpe- ment des oeufs a synthese partielle. Arch, dc BioL, 48 : 529-548. BRACKET, J., 1940. fitude histochemique des Proteines au cours du developpement embryonnaire des Poissons, des Amphibiens et des Oiseaux. Arch, dc BioL, 51 : 167-202. CARLSON, J. G., 1941. Effects of x-radiation on grasshopper chromosomes. Cold Sprin 4O- 3O- 20- 10- o n o o n a. o 10 15 20 25 30 35 ~r 40 Copper Leaching Rate FIGURE 2. Attachment of Bugnla larvae in relation to copper leaching rate. Symbols as in Figure 1. tion of toxic surfaces plays an important role in the prevention of fouling. Effective antifouling surfaces must function in part by stopping development of attached larvae. This is contrary to the previously mentioned views of Visscher, Neu, and Edmondson and Ingram. The latter authors report a restricted growth of Bugnla and serpulid worms on toxic surfaces exposed in Hawaiian waters, but attribute this to "delayed attachment rather than to a slower rate of growth after settling." 128 MILTON A. MILLER Dates of attachment in their tests are not known, and hence rates of growth on treated and untreated surfaces cannot be calculated and compared from their data. In view of the present studies, it would seem that the alternate interpretation of a growth retarding effect is equally plausible, if not more probable. To illustrate further the growth retarding effect of toxic surfaces, another experi- ment involving growth of Bugula on unseasoned copper paint surfaces may be cited. In this experiment, the first two weeks of growth of Bugula on two hot-plastic, copper-paint surfaces {A and B) was compared to that on a non-toxic, hot-plastic control (C). The two toxic paints have identical copper content (32 per cent 16 1.4 1.2- 1.0- E 6 c o 0.8- 0.6- 0.4- 0.2- — l — 60 120 30 90 150 160 210 240 270 300 330 360 Time (hours) FIGURE 3. Growth of Bugula ancestrulae on unseasoned hot-plastic paints. A is an ex- cellent copper paint with adequate leaching rate, B is an inferior paint with low copper leaching rate, and C is a non-toxic control. Cu2O), but paint A (w.w. rosin-paraffin matrix) has an excellent field record correlated with adequate leaching rates, while paint B (ester gum-paraffin matrix) fouls quickly in the field because of low leaching rate. Hot-plastic paints of the type here used have low initial leaching rates and do not develop characteristic leaching rates until they have been submerged for a period of time. Hence, Bugula larvae attach abundantly to unseasoned panels coated with these paints and begin to grow normally on them. As shown in Figure 3, the growth curves in the present experiment do not begin to diverge significantly until after the second day. Growth on paint surface A was practically stopped at this time at the first polypide stage. Some individuals developed the first polypide, most did not, and no colonies were established. Colon- ies were developed on paint surface B, but growth and differentiation were sig- TOXIC EFFECTS OF COPPER ON BUGULA 129 niikantly less than for controls. The three curves become more and more divergent with time. The foregoing differences between growth inhibiting properties of paints A and B must be attributed to differences in their leaching rates as their copper content and growth of Bnyitla on their respective controls are identical. Repeated leaching rate determinations of these paints have shown both to have low initial leaching rates (which accounts for the initial similarities in the Biigula growth curves for the two surfaces) but paint A eventually develops an adequate leaching rate, while leaching rates of paint B do not attain the adequate level. Thus, again the growth retarding property of copper paint surfaces is demonstrated and associated with the copper leaching rate. TABLE II Effect of copper on growth and development of Bugula ancestrulae Growth and differentiation in copper solutions Av. copper concentration Mean length increments ± S.E. Polypides per colony mg. /liter 1 day 3 days 5 days 3 days S days mm. mm. mm. Controls 0.36±.007 0.61±.016 0.77±.021 1-2 1-2 0-.05 0.30±.015 0.55±.013 0.63±.018 1 1-2 0.05-.10 0.26±.009 0.37±.010 0.46±.012 1 1-2 0.10-.15 0.25±.007 0.36±.008 0.48±.013 1 1 0.1 5 -.20 0.14±.008 0.22 ±.009 0.24±.011 0-1 0-1 0.20-.25 0.08±.007 0.10±.008 0.12±.010 0 0 0.25 -.30 0.02 ±.007 0.03 ±.008 0.02 ±.009 0 0 0.30-.35 0.04 ±.011 0.03±.011 0.03 ±.009 0 0 0.35 -.40 -0.03 ±.008 0.01 ±.009 0±.008 0 0 0.40 -.45 0.03±.007 0.02 ±.011 -0.03±.010 0 0 0.45 -.50 0.02 ±.008 0.03±.010 0.03±.010 0 0 The experiments cited in this section are in close accord with results obtained by Ketchum ct al. (1945) and substantiate the leaching rate theory of action of anti- fouling paints. It is noteworthy that their estimate of minimum adequate copper leaching rate (10 micrograms per sq. cm. per day), which was based on comparisons of numerous leaching rate determinations with field exposure tests, is confirmed by the Bugula attachment and growth tests. Data presented in this section show that copper paints with leaching rates less than this value permit larvae to attach in large numbers and to grow and differentiate. Paints with leaching rates greater than 10 micrograms per sq.cm. per day allow only a small percentage of larvae to attach and completely inhibit their growth. Large percentages of larvae occasion- ally attach to paint surfaces with leaching rates between 10 and 15 micrograms of copper per sq. cm. per day, but these do not develop into colonies. Growth of Bugula in copper solutions To determine more precisely the effect of copper ions on growth of Bugula, non- toxic panels bearing attached larvae (about 25) were immersed in sea-water solu- tions of graded copper concentrations. Cuprous oxide was the salt used in making 130 MILTON A. MILLER the copper solutions, but the ions were presumably cupric since the oxidation of cuprous ion is supposedly rapid. The young Bugulae were measured just before immersion in the experimental solutions and at one, three, and five day intervals thereafter. Polypide development was also observed at these times. Mean length increments after immersion were computed as an index of toxicity of the solutions. c 0> -£ 0 c. mm 08H 07- 0.6- 05- 04- 03- D 0.2- oi- o- 5 days a o O.I 02 I I 0.3 0.4 I 05 Copper Concentration (mg: per liter) FIGURE 4. Growth of Buyula ancestrulae in relation to copper concentration. Curves for one-day (squares) and five-day increments (circles) based on grouped data (class intervals = 0.05 mg. copper per liter). Filled squares indicate one or more polypides per colony, half filled squares show that some stalks did not differentiate polypides, and open data points indicate that no polypides were developed. See also Figure 5. The copper concentrations of each solution were determined before and after the experiment, and the average values were used. As clearly shown in Table II and Figures 4 and 5, the degree of early growth and differentiation of Bngitla is closely related to the copper concentration in the water surrounding the organisms. Up to about 0.3 mg. per liter, increment in TOXIC EFFECTS OF COPPER ON BUGULA 131 length is inversely proportional to copper concentration of the solution. Higher concentrations inhibit growth completely or allow but slight, insignificant increment. The sharp break or inflection in the curves in Figures 4 and 5 at about 0.3 mg. copper per liter apparently marks this as a critical concentration for growth. No stimulating effect of small amounts of copper on growth was noted, since even in the greatest dilutions used growth increments were less (with minor exceptions) than in fresh sea-water controls. 0.6 — • o-H >> o Q ~ 0.4— | | - 0.3—1 0.2 — O.I — o— 3 day s O O O O O O O I O.I 0.2 I 0.3 0.4 05 0.6 Copper Concentration (mg. per liter) FIGURE 5. Three day growth of Bugula ancestrulae in relation to copper concentration. Black circles indicate fully developed polypides, half filled data points represent partly formed polypides, and clear circles indicate no differentiation of polypides. See also Figure 4. Polypide differentiation is inhibited by copper concentrations greater than 0.2 mg. per liter, the minimal value being less than that required to prevent growth completely. Concentrations less than 0.2 mg. per liter only delay polypide de- velopment. This is shown by the fact that only controls attained the second polypide stage at three days (Table II). by the several cases of incomplete or non- functional polypides at three and five days in the range between 0.1 and 0.2 mg. 132 MILTON A. MILLER per liter (Figs. 4 and 5), and by the fact that at five days only the first polypide was developed in concentrations greater than 0.1 mg. per liter. A series of critical copper concentrations may be postulated from these and other studies (Table III). The minimum lethal dose (MLD) for free swimming larvae is about 0.3 mg. per liter (Miller and Cupp, 1942), the same concentration required to stop growth of attached larvae. No data are available on concentrations necessary to prevent larval attachment. Presumably these would be at least as great as the minimum lethal dose, and probably greater, since larvae in solutions containing as much as 0.4 to 0.5 mg. copper per liter do not die immediately but swim about more and more slowly for a half hour to an hour or more. Given a suitable surface, it is conceivable that some might attach in this interval. TABLE III Critical copper concentrations Copper concentration Physiological effect (mg. per liter) >0.3 Kills larvae. Inhibits growth. 0.2-0.3 Retards growth. Inhibits polypide development. <0.2 Retards growth and development of polypides. Recovery of Bugnla from copper poisoning The question arises, are the above-described toxic effects of copper permanent,, or can organisms recover to any significant degree after exposures to sublethal dosages? Preliminary experiments showed that Bugnla ancestrulae can recover nearly normal development after immersion for 6 to 24 hours in sublethal copper sea-water solutions (Miller and Cupp, 1942). In this experiment, the periods of immersion in copper solutions were extended (three to seven days) and alternated with periods of immersion in fresh sea water to determine the effect on recovery of longer exposures and of intermittent as compared to continuous dosages. The initial procedure was similar to that of preceding experiments. A non- toxic test panel with attached larvae was immersed in each of eight copper sea- water solutions (Nos. 1-8, Table IV) for a period of three days (time period I). The subsequent procedure features a staggered schedule of transfers of the animals from copper solutions to fresh sea water, so that, while one group was exposed to copper, the other was immersed in untreated sea water and vice versa. After the first three days exposure, half of the panels (group A, sets 1-4) were transferred to fresh sea water, and the remainder (group B, sets 5-8) were left in the original copper solutions. After four days (period II), group A panels were transferred from sea water back to the original copper solutions, while group B sets (that had now been in the toxic solutions for a week) were transferred to fresh sea water. After another four days (period III), the process was again reversed — those in sea water were transferred back to copper solutions and vice versa. The experi- ment was terminated on the fifteenth day, the final four days constituting period IV. Length measurements were made before the original immersion (a few hours after attachment of the larvae), at the time of transfers (on the 3d, 7th, and llth days), and at the termination of the test on the 15th day. From these data, the rates of growth relative to that of controls for the four time periods in copper solutions and sea water were computed. Concentrations of the copper solutions TOXIC EFFECTS OF COPPER ON BUGULA 133 were estimated from their biological effects (using curves in Figs. 4 and 5) as chemical determinations were not made for this test. Data are given in Table IV, and the salient features are shown in Figure 6. In all but one case, transfers to fresh sea water were followed by marked and significant increases in growth rate and polypide differentiation, while the reciprocal transfer resulted in sharply decreased, it" not completely inhibited, growth (Fig. 6). The exceptional case (set 7) presumably received a lethal dose in the first period as no significant increases in length and certainly no differentiation were observed following transfer to fresh water. All others, however, exhibited an amazingly high degree of plasticity and regulatory ability in recovering from the effects of the TABLE IV Growth and development of Bugula ancestrulae during and after immersions in copper sea-water solutions Symbol No. Cu. in solns. Mean length increments, time periods I-IV * Total no. of polypides per colony at end of periods: mg./liter I II III IV (3 days) (3-7 days) (7-11 days) (11-15 days) I II III IV mm. mm. mm. mm. A-l 0.26 0.05 0.34 0 0.14 0 1 1 2 -2 0.27 0.04 0.22 0.01 0.09 0 1 1 2 -3 0.24 0.09 0.27 0.10 0.14 0 1 1 2 -4 0.23 0.11 0.25 0.09 0.06 0 1 1 2 B-5 0.25 0.08 0.01 0.12 -0.02 0 0 1 ab. 1 -6 0.23 0.10 0.01 0.17 -0.02 0 0 1 ab. 1 -7 0.3 0.03 0.03 0.05 0 0 0 0 0 -8 0.20(?) 0.14 0.16 0.23 0.10 1 1 1 2 Control 0.68 0.64 0.11 0.25 1 2 3 4 (0.68) (0.53) (0.28) (0.19) * Group A immersed in copper solutions during periods I and III, and in fresh sea water during periods II and IV. Group B immersed in copper solutions during periods, I, II, and IV, and in fresh water during period III. poison. Even those immersed an entire week in solutions that allowed but slight growth increment and no visible differentiation (e.g., sets 5 and 6) were able to recover significantly, from growth rates of zero to as much as 60 per cent that of controls, when restored in fresh sea water. They also differentiated polypides though these were slightly abnormal in some instances. Set 8 in group B was apparently exposed to a weaker solution than any of the others since these indi- viduals not only grew considerably but also developed polypides even while im- mersed in the copper solution. Nevertheless, significant increases and decreases in relative growth rates for this set were observed in the various periods. Seem- ingly, one could control rate of growth and differentiation almost at will by appro- priate exposures to copper. The ability of young Bugulae to recover each time after repeated and long immersions in rather toxic copper solutions is truly remarkable. 134 MILTOX A. MILLER In mi case was recovery complete as judged either by growth rates or polypide development. Growth rates after transfers to sea water ranged roughly between a third and four-fifths of the normal (see Fig. 6). and the number of polypides developed was less than that of controls. (0-3doys) m (7-11 days) IZ (H-15 days) Time Periods FIGUKE 6. Growth rates of Bugitla ancestrulae (in per cent of normal) during and after exposure to subletlial copper concentrations. Black data points represent averages for group A panels (Table IV), and clear circles are averages for group B panels (omitting set 7). Solid lines indicate periods of immersion in copper solutions, and dashed lines represent transfer to fresh sea water. Vertical lines through data points show the range between minimum and maximum growth rates for each average. TOXIC EFFECTS OF COPPER ON BUGULA 135 Length of exposure in sublethal concentrations has relatively little apparent effect on recovery. This is shown by the fact that Bugulae exposed to copper for seven days were able to develop after transfer to sea water at a rate comparable to that of specimens which had been exposed for only three days. As previously noted, however, the longer exposures caused abnormal polypide development in some cases. In contrast to this finding, there is evidence that in the case of barnacles and mussels the length of exposure is as important as the concentration in determining the toxic effects of copper solutions. Clarke (1943) found that low concentrations acting for a long time produced effects equivalent to high concentration acting for a short time. In other words, the toxic effect is proportional to the product of duration and intensity. Further experiments of this general type are needed to determine more precisely immediate and long-term effects of various exposures and concentrations and of continuous versus discontinuous exposures. These experiments have practical implications since, under natural conditions, the concentration of toxic ions adjacent to a copper paint surface might vary from time to time as a result of fluctuation in leaching rate, formation of surface films, currents flowing past the surface, and other factors. The effect of such changes on attached Bugitla larvae might be surmised from the foregoing studies showing that growth is an inverse function of copper concentration in the medium and that stunted individuals can resume development when restored to non-toxic situations. Presumably, Bugula ancestrulae attached to a surface with fluctuating toxicity might recover to some degree with each significant decrease, if the organisms were still viable. With each increase in length, the ancestrulae greatly improve their chances to establish colonies for reasons which will become apparent in the next section of this paper. Toxicity gradients of copper paint surfaces The question of zones or gradients of toxicity adjoining toxic paint surfaces is involved in understanding their antifouling action, and is the last problem to be considered in this paper. Although a toxicity gradient has been assumed as a consequence of diffusion of toxic ions emanating from an antifouling paint, a demonstration of this seemed desirable. Furthermore, information on the effective limits and other characteristics of the toxic zone is of particular interest in connec- tion with the establishment of colonial fouling organisms (e.g., erect bryozoans and most hydroids) that grow, plant-like, more or less perpendicularly away from the surface to which they attach, and develop new individuals at the ends of their branches. It would be useful to know how far from the toxic surface growth inhibiting and growth retarding concentrations are maintained, or how much a colony would have to grow before its terminal polypides were out of danger from poisoning. In the following experiment, the problem was to demonstrate, if pos- sible, the existence of the toxic zone of an antifouling paint and to determine its general characteristics. The preceding studies, showing rather precise relationships between growth of Bugula ancestrulae and copper concentration, suggested a method for attacking the problem of toxic gradients. The essential features of the procedure used and of the results obtained are illustrated in Figure 7. A non-toxic panel bearing at- 136 MILTON A. MILLER tached and growing Hut/iilac was placed perpendicularly against a panel coated with a cold-plastic copper-paint (Table I, No. 7). With this arrangement, the develop- ing Biigula stalks maintain practically constant distances between their axes and the toxic surface since they grow parallel to the latter. The non-toxic panel was ruled in millimeter divisions paralleling the toxic surface, and these were used as class intervals of distance in analyzing the data. The ancestrulae in each division were measured at the start of the tests and two days afterwards to determine the effect of diffusing copper ions on their growth at various distances from the paint surface. For control, a non-toxic panel was substituted for the toxic panel. Ex- perimental and control racks were placed in an aquarium containing ten liters of sea water. The water was aerated during the tests by fine streams of air bubbles delivered through pinholes in a piece of rubber tubing stretched along the bottom of the aquarium. This method of aeration caused some circulation of water but no appreciable agitation. TABLE Y Tiixicily 'j/nJit'nts i>f a copper paint demonstrated by growth of nt measured distances from the toxic surface Distance from paint surface Test 1 Mean growth increment ± S.E. (2 days) Test 2 Mean growth increment ± S.E. (2 days) Experimental Control Experimental Control mm. mm. mm. mm. mm. 0-1 0.10±.02 0.48±.02 0.01±.01 0.43±.02 1-2 0.27±.03 0.42 ±.05 0.03±.01 0.46±.02 . 2-3 0.40±.06 — 0.1 6 ±.03 0.44±.03 3-4 0.38±.06 0.50 0.30±.06 0.49±.02 4-5 0.35±.10 0.50 0.40 0.50±.04 5-6 0.45 ±.08 0.45 ±.04 0.37 ±.04 0.50 6-7 0.55 ±.04 0.45 ±.04 0.35 — 7-8 0.45±.08 0.40 0.45 0.40 8-9 0.50±.07 0.50 0.35 ±.04 0.40 9-10 0.50 0.43 ±.03 0.41 ±.02 0.50 Two tests were made : the first, using toxic panels that had been seasoned for ten weeks in the laboratory ; the second, with the same panels that were seasoned another two weeks. In the first tests, the slime film which had accumulated during seasoning was not removed but for the second tests the slime was wiped off. Data are given in Table V and graphically illustrated in Figures 7 and 8. As clearly demonstrated by the retarded growth of the colonies near the painted surface, the toxicity arising from the surface decreases rapidly with distance from the surface. The outer effective limit of the toxic zone is variable. Beyond two millimeters from the surface in the first test and four millimeters in the second, no significant difference in growth between experimental and controls was demon- strated. These values, then, represent the respective outer limits of the toxic gradient in the two tests and indicate the order of magnitude of the width of the toxic zone for a good antit'ouling paint. Within these zones, growth increments are roughly proportional to perpendicular distance from the toxic surface as might be expected from the diffusion gradient of toxic ions. Ancestrulae immediately TOXIC EFFECTS OF COPPER ON BUGULA 137 Fn.i'RK 7. The toxic gradient extending from a copper paint surface as shown by growth of Bngula ancestrulae at measured distances from the toxic surface. Biigula figures are camera lucida drawings made four days after start of test 2 (for two-day growth, see Table V), and are twice enlarged in comparison to the millimeter rulings shown on the non-toxic panel. mm Range of Control Means 34567 Distance from Paint Surface (mm.) FK.I-RE 8. Gradients of toxicity of an antifouling paint: test 1— paint seasoned for ten weeks and coated with slime film, and test 2 — after twelve weeks seasoning and with slime film removed before testing. 138 MILTON A. MILLER adjacent to the toxic paint, however, showed no increase in length indicating that growth inhibiting concentrations of copper were maintained there. This result was anticipated since larvae attached to surfaces coated with this paint do not grow (Table T, No. 7). Growth within the first millimeter interval is contributed by ancestrulae in the outer part of it. The width of the growth inhibiting portion of the toxic gradient (practically, the most important part of it) could be more pre- cisely determined by further tests using smaller intervals of distance. The panels used in the second test were clearly much more toxic than those first tested as shown by the greater width of the toxic zone, by the greater growth inhibition near the surface, and by the fact that the curve of growth increments at distances greater than four millimeters tends to fall below that of controls. With respect to the latter point, the mean growth increment for all individuals between four and ten millimeters in test two is significantly less than the corresponding mean in test one or that for controls. The means for single millimeter intervals in this range are not significantly different probably because of the small numbers of individuals in each. The characteristics of the toxic zone are undoubtedly affected by various factors such as leaching rate and velocity of flow of water across the surface. In the field, currents or movement of the painted surface through the water would probably alter the character of the toxic zone. This and other factors could be simulated in the laboratory and their effects analyzed using the above-illustrated methods with appropriate modifications. The demonstrated vertical gradient would probably differ from a horizontal gradient that extends outward from the edge of a painted surface. The latter type was nicely demonstrated by Pomerat and Weiss (1943) in field tests with panels on which areas of various shapes and sizes were left unpainted. The horizontal gradient observed by these authors was expressed both by graded growth of fouling on large unpainted areas, and by absence of fouling on smaller areas encircled by the paint. -Their effects might be attributed in part to delayed larval attachment as well as retarded growth since the settling of larvae cannot be controlled in the field. For practical purposes, the activity of a vertical gradient is probably of greater importance. The foregoing experiment together with those reported in preceding sections of this paper clearly indicates that the prevention of Bitgula fouling on copper paint surfaces is dependent upon their ability to maintain growth inhibiting concentrations of copper in a narrow zone at the surface. The Fength of newly attached Bitgula ncritina larvae (about 0.2 mm.) presumably represents the minimum adequate widtli of the growth inhibiting zone required to prevent establishment of this or- ganism. If attached forms are permitted to grow, their apical developing parts move away from the toxic surface and hence into regions of lower toxicity with consequent acceleration of development. As new polypides are produced by distal budding, they find themselves in a less toxic environment than their predecessors and eventually the terminal polypides would lie entirely outside of the toxic zone. Since the polypides are functionally independent (except for support), the colony can flourish even though its basal members are dead. If the toxic zone extends outward only a few millimeters, just a few of the basal polypides would be affected since Bngiila colonies at the first polypide stage average about 0.9 mm. in length and each successive polypide adds at least a half millimeter to the length. TOXIC EFFECTS OF COPPER ON BUGULA 139 Larvae that settle on previously attached forms or on any inert particles elevated appreciably above the surface plane of the paint would clearly stand a better chance of survival than those attached directly to the toxic surface. They would occupy less toxic regions of the toxic gradient which might permit their growth, while those attached to the surface itself might be killed or permanently stunted by the higher concentration of the toxic prevailing there. Judging from the steep slopes of the toxicity gradients for an effective paint (Fig. 8), a fraction of a millimeter from the paint surface might make a significant difference in development, especially on surfaces with borderline toxicity. To summarize : the foregoing preliminary tests clearly demonstrate a zone or gradient of toxicity that extends outward a few millimeters from an effective copper paint surface. Further experiments, using the method illustrated, are indicated to determine more precisely the limits and other characteristics of the toxic zone under various conditions. SUMMARY Copper paint surfaces prevent the establishment of Bitgnla ncritina (1) by repelling or killing the larvae and (2) by inhibiting growth and metamorphosis of attached larvae. Copper paints with leaching rates less than 10 micrograms of copper per square centimeter per day permit the larvae to attach in large numbers and to grow and differentiate. Paints with leaching rates greater than 15 micrograms per square centimeter per day allow only a small percentage of larvae to attach and completely inhibit their growth. Large percentages of larvae occasionally attach to paint surfaces with leaching rates between 10 and 15 micrograms per square centimeter per day, but these do not develop colonies. No consistent effect of slime film on larval attachment was noted. Its presence is not prerequisite to attachment. Precise relationships between copper concentration and growth of Bugula ancestrulae are demonstrated. Growth in sea-water solutions of copper is in- versely proportional to the concentration up to 0.3 mg. per liter. Higher concen- trations completely inhibit growth. The critical copper concentrations affecting various stages of the early life cycle of Bugula are as follows : ( 1 ) Concentrations greater than 0.3 mg. per liter kill larvae and completely inhibit growth of attached forms, (2) concentrations between 0.2 and 0.3 mg. per liter retard growth and prevent polypide formation, and (3) concentrations less than 0.2 mg. per liter retard growth and polypide development. No stimulation of growth by copper solutions was observed. There was some evidence that small concentrations of copper stimulated attachment of larvae. Bugula ancestrulae can recover and develop almost normally after seven days exposure to sublethal concentrations of copper. They can recover after repeated immersions in copper solutions that practically prevent growth. Length of ex- posure has relatively little effect on their ability to recover from copper poisoning. A gradient of toxicity extending outward a few millimeters from a copper paint surface is demonstrated. 140 MILTON A. MILLER LITERATURE CITED BRAY, A. W., 1924. Fouling of ships' bottom (4th Kept.), Bur. Construction and Repair, Navy Department. (Unpublished) CARRITT, D. F. AND G. A. RILEY, 1943. Note on the relative toxicity of cuprous and cupric ions. Woods Hole Oceanogr. Ins., Sixth Annual Report to the Bur. of Ships, 2, Biological Investigations Pertaining to the Fouling of Ships' Bottoms, Paper 13. (Unpublished) CLARKE, G. L., 1943. The effectiveness of various toxics and the course of poisoning and re- covery in barnacles and mussels. Woods Hole Oceanogr. hist., Sixth Semi- Annual Rcpt. to Bur. Ships, 2, Biological Investigations Pertaining to the Fouling of Ships' Bottoms, Paper 11. (Unpublished) EDMONDSON, C. H., 1944. Incidence of fouling in Pearl Harbor. Occas. Papers Bishop Mus. 18 (No. 1) : 1-34. EDMONDSON, C. H. AND W. M. INGRAM, 1939. Fouling organisms in Hawaii. Occas. Papers Bishop Mus., 14 (No. 14) : 251-300. GRAVE, B. H., 1930. The natural history of Bugula flabellata at Woods Hole, Massachusetts, including the behavior and attachment of the larvae. Jour. Morph., 49: 355-384. GRAVE, C. AND P. A. NICOLL, 1939. Studies of larval life and metamorphosis in Ascidia nigra and species of Polyandrocarpa. Papers Tortug. Lab., 32: 1-46. JONES, J. R. E., 1941. The effect of ionic copper on the oxygen consumption of Gammarus pulex and Polycelis nigra. Jour. Exp. Biol.. 18: 153-161. KETCHUM, B. H., J. D. FERRY, A. C. REDFIELD AND A. E. BURNS, JR., 1945. Evaluation of antifouling paints by leaching rate determinations. Indus, and Engin. Chemistry, 37 (5) : 456-460. MILLER, M. A. AND E. E. CUPP, 1942. The development of biological accelerated tests. (Prog- ress Report.) Ann. Rcpt. San Diego Naval Bio!. Lab. to Bur. Ships, Navy Depart- ment. (Unpublished) NEU, W., 1932. Untersuchung iiber den Schiffsbewouchs. Intcrnat. Rev. der gcsamtcn Hydro- biologic, 27: 105-119. PARKER, G. H., 1924. The growth of marine animals on submerged metals. Biol. Bull., 47 : 124-142. POMERAT, C. M. AND C. M. WEISS, 1943. The influence' of toxic paints on fouling of adjacent unpainted surfaces. Woods Hole Oceanogr. hist., Sixth Scmi-Annual Kept, to Bur. Ships, 2, Biological Investigations Pertaining to the Fouling of Ships' Bottoms, Paper 10. (Unpublished) PRYTHERCH, H. F., 1934. The role of copper in the setting, metamorphosis, and distribution of the American oyster, Ostrea virginica. Ecol. Monogr., 4 : 47-107. RILEY, G. A., 1943. The toxicity of heavy metals to fouling organisms. Woods Hole Oceano- gr. Inst., Sixth Scmi-Annual Rcpt. to Bur. Ships, 2, Biological Investigations Pertaining to the Fouling of Ships' Bottoms, Paper 12. (Unpublished) VISSCHER, J., 1927. Nature and extent of fouling of ships' bottoms. Bull. Bur. Fisheries, 43 (Pt. II) : 193-252. (Doc. No. 1031) WHEDON, W. F., E. E. CUPP, M. A. MILLER, M. L. DARSIE, J. C. RAPEAN AND M. L. MARSHALL, 1943. Investigations pertaining to the fouling of ships' bottoms. Ann. Rept. San Diego Naval Biol. Lab. to Bur. Ships, Navy Department. (Unpublished) WHEDON, W. F., R. C. NELSON, E. E. CUPP AND M. A. MILLER, 1942. Investigations pertain- ing to the fouling of ships' bottoms. Semi-Ann. Rept. San Diego Naval Biol. Lab. to Bur. Ships, Navy Department. (Unpublished) A NEW GRAPHIC METHOD OF DESCRIBING THE GROWTH OF ANIMALS LIONEL A. WALFORD * Aquatic Biologist, United States Fish and Wildlife Service Growth curves, when conventionally plotted as length on age, are difficult to compare and classify. Moreover, the usual mathematical methods of fitting them, such as, the logistic and the Gompertz are rather laborious and incon- venient for application to large numbers of individuals. Fortunately, for many purposes, it is unnecessary to describe the whole growth curve; for the part below the inflection point is completed early and the part above the inflection point — the "self-inhibiting" phase, covers the period of life when differences in growth are likely to be most striking. That phase of the growth curve can be appropriately represented by a straight line, the charac- teristics of which can be treated statistically, by the following graphic method: Using arithmetic graph paper, with body length represented along both the x axis and along the y axis, plot length at ages 1, 2, 3, 4, 5 • -r on the x axis against length at ages 2, 3, 4, 5, 6- • •« + 1, respectively, on the y axis. For several species on which I have found published length data, these points fall along a straight line. This line can be regarded as a sort of transformation of the usual growth curve, and in the following discussion I .vill call it that. The nine examples given in Figures 1-3 are based on average lengths of large samples. When lengths of individual specimens are plotted by this method, a straight-line relationship is still obvious, though the points deviate more widely from the line than when averages are used. These deviations doubtless result from several causes, among which random variations in environmental experience and errors of observation readily suggest themselves. For a few species the published growth data failed to produce a straight line. In these cases, the course of growth may differ from that in other animals; or the observed anomalies may reflect some artifactual effect in the data. Among those species for which this "transformation" results in a straight line, the growth increments corresponding to equal time intervals, for example, be- tween years of age (k -- /i, /3 -- /2, h •- h,- • -ln •- L-i), have the following inter- relations; where /„ refers to the length at any given age, i.e., at the end of any given time interval:2 /3 — /2 /4 — '3 ' _n n— , 1% ~~ l\ h — ll 1 4 " h l-n-l ~~ ln-1 1 For advice and assistance in the mathematics of this paper, I am indebted to Professors George Polya and Harold Bacon of Stanford University. I am also grateful to my colleagues, Mr. O. Elton Sette and Dr. Frances Felin, for their constant interest and help. 2 In this discussion, l\ is not necessarily the value directly obtained by measurement, but a value calculated on the basis of all measurements (see Fig. 4). 141 142 LIONEL A. WALFORD or /„ - - /„_! = k (ln-l — It is interesting to note that, with 10 = 0, k-li = k (/, -/„)=£ /i h — h — k (/3 — /a) = k k h ' ~ h = k (/4 - - Is) = k k = k'2 /„ -- /„_! = k (ln-l - ln-d = k k"~- /! = k-1 /,. O < z >- i o i a •* s e 7 12 (1) (2) PACIFIC RAZOR CLAM . from Masiet, B.C. - (rent Swickshak, Alai. e e 10 iz 14 ie LENGTH AT AGE n 5 6 7 8 9 10 II It \3 14 AGE, YEARS FIGURE 1. Left-hand series: growth data plotted according to the method lescribed, and fitted empirically with straight lines. Right-hand series: the same data converted to conven- tional size-on-age curves. Data on fresh water mussel from Chamberlain (1931); on Pacific razor clam from Weymouth (1931). The constant k is positive and less than one; that is, the yearly growth increments decrease. These relationships are consistent with there being a growth capacity, which is approached from the inflection point at a constant percentage rate. This is in accord with interpretations made by a number of students of growth, among A NEW METHOD OF DESCRIBING GROWTH 143 400 .380 MO SW 320 300- 160 240 Z20 100 180 160 too 900 I Z34-5S783IO lit 200 MO 400 500 WO 100 800 1100 - INI 800 700 • .' STRIPED BASS I £345 76 600 700 800 300 1000 1100 X i i ' i - 100 PACIFIC HALIBUT — male female I i i I 1 1 1 L BOO 700 600 SOS 4-OC 300 too 1100 1000 300 800 600 45 6 10 II \i >3 14 15 16 17 ii 1 T— 1 1 1 1 1 1 1 1 1 1 1 1~~ — 1 1 1 1 1 1 1 1 1 1 — I_J_ — f=* 111 10 x ^-p*"~r~" IB 16 16 • X' : ; - X : Ik I* 14 x / 12 12 / ID 10 8 X /• SMALL MOUTHED / BLACK BASS 8 c 4- - / ;/ : 4 L • , , s* ',,,,, , , . < I i 468 IOI£I4MI62«U242£ LENGTH AT AGE n I t 3 4 S 6 7 8 9 AGE, YEARS FIGURE 2. Left-hand series: growth data plotted according to the method described, and fitted empirically with straight lines. Right-hand series: the same data converted to conven- tional size-on-age curves. Data on Norwegian herring from Runnstrom (1936); on striped bass from Merriman (1941); on Pacific halibut from Thompson and Bell (1934); on small mcuthed black bass from Bennett (1938). 144 LIONEL A. WALFORD them Minot (1908), Wright (1926), Brody (1927 a, b, c), and \\Vymouth, McMillan, and Rich (1931). The length at infinite age, /x, which can be regarded as the ultimate length or limiting length, can be calculated as follows: The length /„ is attained by adding to /i the successive increments. ' /„ = /I + (/2 -- /,) + (/3 " k) + (1-4 - /3)+- '•+(/„ - Zn-l). Yet these increments were expressed in the formulas following (1), thus: k ~ /I = k /,, I, ~k = & /!, •••/„- /„_! = fe"-1 /„. (3) zn 7 2* ISC - / RAT z.u 100 ISO c B -IM 180 130 WO ZJO Length (mmJ at dftc n days 80 100 120 140 1(0 180 ZOO ZIO 240 IbO LW .3m 310 340 JCO A$e, days o X - 170 - 160 - ISO MAN 130 E no :• cu 100 2: SO 80 e 10 11 14 Aije, years is ie to JERSEY CATTLE no o X ' 60 30 MX) HO liO 130 ISO Htljht at withers (ctn.)aT *Qt n months 4 6 8 10 12 14 « 18 ZO 22 24 ZG A(}e. months FIGURE 3. Left-hand series: growth data plotted according to the method described, and fitted empirically with straight lines. Right-hand series: the same data converted to conven- tional size-on-age curves. Data on rat from Donaldson (1931); on man from Thompson (1942); on Jersey cattle from Ragsdale, Elting, and Brody (1926). A NEW METHOD OF DESCRIBING GROWTH 145 -Z 30 10 . 10 50 60 70 Length at age N FIGURE 4. Transformation of the growth curve of a hypothetical animal, drawn in a heavy line to illustrate the dimensions given in the text. Note that the ultimate length, k,, can be located graphically as the point where the length at age n equals the length at age n + 1 : also where the transformation intersects a line drawn at 45° through the zero point. Therefore, /„ = 1 - kn 1 - k (4) by the well-known formula for the geometric series. Therefore, when n ap- proaches oo , /„ tends to /„, and kn to 0, and so we obtain /i 1 - k (5)' 3 Where this graphical representation gives precisely a straight line, the above calculation shows that /„ is expressed by the formula /„ = Ja>(l — &")• It is a modification of the "Modified Exponential" (cf., Croxton and Cowden, 1940, pp. 441 ff.), but contains one less parameter, and of course applies only to that segment of the growth curve above the inflection point. 146 LIONEL A. WALFORD Thus the limiting length lx may be computed readily from the y intercept, l\ and the slope of the fitted straight line, k; and k, in turn, is readily calculated from any of the ratios preceding equation (1). According to the series of equations (1) to (4) the slope of the transformed growth curve, k, is the constant given not only by the ratio y-^ = k but l>n—l 'n-2 / / i i ^i ' ' '"+1 i also by the ratio -y— —7— = k. 'x ' n In other words, the amount of growth which remains unfulfilled at the be- ginning of any time interval, is a constant percentage of the amount of it which had remained at the beginning of the preceding time interval. Consequently the higher the k value, the more slowly growth approaches the limiting length. This method of plotting growth data permits an objective determination of the limiting length, lx, even before that length is "attained." 4 It provides two growth characteristics, k and /», from which the upper segment of the length-on- time growth curve can be reproduced. These constants are so simply derived, that it is practical to determine them for large numbers of individuals. They can be used for studying growth variation within and between populations, hence for distinguishing between races of animals with differing growth patterns. The method is particularly useful in studies of fishes whose scales bear annual rings, from which the growth history of individual specimens can be estimated. LITERATURE CITED BENNETT, G. W., 1938. Growth of the small-mouthed black bass, Micropterus dolomieu Lacepede, in Wisconsin waters. Copeia: 157-170. BRODY, SAMUEL, 1927a. Growth rates, their evaluation and significance. University of Missouri, Agr. Exp. Station, Research Bull., 97. BRODY, SAMUEL, 1927b. Growth and development with special reference to domestic animals. University of Missouri, Agr. Exp. Station, Research Bull., 96. BRODY, SAMUEL, 1927c. Equivalence of age during the self-inhibiting phase of growth. Uni- versity of Missouri, Agr. Exp. Station, Research Bull., 102. CHAMBERLAIN, THOMAS K., 1931. Annual growth of fresh-water mussels. U. S. Bureau of Fisheries, Bulletin, 46: 713-739. CROXTON, FREDERICK E. AND DUDLEY, J. COWDEN, 1940. Applied general statistics. Prentice- Hall, Inc. 822 pp., 257 figs. DONALDSON, HENRY R., 1924. The rat, data and reference tables for the albino rat Mus norvegicus albinus and the Norway rat Mus norvegicus. Philadelphia, 458 pp., 72 figs. MERRIMAN, DANIEL, 1941. Studies on the striped bass Roccus saxatilis of the Atlantic coast. 1941 U. S. Fish and Wildlife Service, Fishery Bulletin, 35: 1-77. Mi NOT, C. S., 1908. The problem of age, growth and death. John Murray, London. KAGSDALE, A. C., E. C. ELTING, AND SAMUEL BRODY, 1926. Growth and development with spe- cial reference to domestic animals. 1. Quantitative data, growth and development of dairy cattle. 1. Weight growth and linear growth. University of Missouri, Agricultural Experiment Station, Research Bulletin, 96: 1-84. UINNSTROM, SVEN, 1936. A study on the life history and migrations of the Norwegian spring- hen ing based on the analysis of the winter rings, and summer zones of the scale. Fiskeridirektoratets Skrifter, Serie Havundersokelser. Report on Norwegian Fishery and Marine Investigations, V (2): 103 pp., 15 figs., 3 plates. THOMPSON, D'ARCY WENTWORTH, 1942. On growth and form. Cambridge University Press, 1097 pp., 554 figs. 4 Strictly speaking, of course, under the terms of this description the limiting length is only approached; it is never attained. A NEW METHOD OF DESCRIBING GROWTH 147 THOMPSON, WILLIAM F. AND F. HEWARD BELL, 1934. Biological statistics of the Pacific halibut fishery , (2) effect of changes in intensity upon total yield and yield per unit of gear. International Fisheries Commission, Report No. 8, 49 pp., 15 figs. WEYMOUTH, F. W., 1931. The relative growth and mortality of the Pacific razor clam Siliqua patula Dixon, and their bearing on the commercial fishery. U. S. Bureau of Fisheries, Bulletin 1930, 46: 543-567. WEYMOUTH, F. W., H. C. MCMILLAN AND WILLIS H. RICH, 1931. Latitude and relative growth in the razor clam, Siliqua patula. Jour. Exp. Biol., 8 (3): 228-249. WRIGHT, SEWALL, 1926a. A frequency curve adapted to variation in percentage occurrence. .Jour. Amer. Stat. Assoc., 21: 162-178. WRIGHT, SEWALL, 1926b. Reviews. Jour. Amer. Stat. Assoc., 21: 493^197. INVESTIGATION ON THE LOCUS OF ACTION OF DDT IN FLIES (DROSOPHILA) DIETRICH BODENSTEIN Medical Division, Entomology Section, Edyeivood Arsenal, Maryland INTRODUCTION The experiments reported herein were designed to gain information as to where DDT produces its poisonous effect in the insect. DDT poisoning in insects is characterized by symptoms of hyperactivity and dis- coordination of neuromuscular system, followed by convulsions and terminating in death. Isolated legs of DDT treated roaches continue to twitch after complete separation from the body, and isolated legs of normal roaches were induced to twitch by the application of DDT to the cut surface, although they remained quiescent if DDT was not applied (Yeager and Munson, 1945). Likewise the isolated legs of adult blowflies (Phonnia reghia) showed twitching movements when dipped before or after they were cut off from the animal into a one per cent DDT acetone solution, while untreated isolated legs remained motionless (Chadwick, 1945). From somewhat different experiments, Tobias ct al. (1945) working with roaches suggested that the thoracic ganglia were the critical loci for the action of DDT. From these observations it appeared likely that the symptoms of DDT poisoning in insects resulted from an effect on the central nervous system, but that there existed also peripheral components which were not as yet exactly delimited. MATERIAL AND METHODS The experiments were performed on the larvae and adults of the fruit-fly (Drosophila virilis}. The DDT preparation used was in form of an emulsion, (one per cent DDT, one per cent lecithin, ten per cent peanut oil, emulsified in a 0.95 per cent NaCl solution). This emulsion was injected by means of a micro- pipet into the abdominal cavity of the insect. The physiological saline solution used throughout the investigation was a Ringer solution modified for Drosophila (H2O, 1000 cc. ; NaCl, 7.5 gm. ; KC1. 0.35 gin.; CaCl,, 0.21 gm.). The various concentrations of phenobarbital used were also always made up in this Ringer solution. Imaginal discs were transplanted with the usual Drosophila transplanta- tion technique. EXPERIMENTAL Behavior of larvae and adults after poisoning When DDT emulsion was injected into the abdomen of adult flies which had been slightly narcotized with ether, the response to the poison was immediate. Legs and wings at once went into violent, uncoordinated movements. About twenty seconds after the injection the abdomen, previously motionless, began to 148 LOCUS OF ACTION OF DDT IN FLIES 149 convulse ; its movements were a rapid succession of short, uncoordinated spasmodic twitches. At first the convulsions were strong and a great deal of the injected emulsion was thus pressed out through the puncture wound. About five minutes after the injection the legs and wings went into a spasm and took up a characteristic position. The legs were drawn toward the body and crossed over ventrally, while the wrings were folded backward. This position was maintained until the animal died. The contractions of the abdomen continued for about four to seven hours but became gradually weaker. After seven hours the animal was apparently dead. Although during all of this time the legs and wings remained in their spasmodic condition, one occasionally observed slight twitches of the tarsal segments and of the antennae ; the wings, however, showed no movement. It was thus clear that the muscular response to the poison varied in different regions of the body, for the wing and leg muscles soon went into contraction and remained that way, while the muscles of the abdominal wall continued to convulse for a long period. When an emulsion prepared in the same manner, but containing no DDT, was injected into flies, no effect was noted. The animals recovered from narcosis in the usual way, and were still alive the following day without any apparent injurious effects. Thus the symptoms described above were due to DDT and were not caused by the emulsion itself. Larvae narcotized with ether were motionless, except for the pulsating heart tube visible through the transparent skin. When such larvae were injected with DDT emulsion, one observed at first a great acceleration of the heart-beat. Con- vulsions of the body wall began about twenty seconds after the injection. It was difficult to observe the heart-beat while the convulsions were in progress, but it was found in ligatured larvae, which will be described below, that the heart-beat soon became normal again after the initial acceleration. The larval convulsions were very strong and uncoordinated. Contraction wave after contraction wave passed over the creature, somewhat resembling crawling movements, yet the animal was unable to move from its place. The forward and backward movements were much more rapid than the normal crawling movements. Moreover, the animal never extended to its full length but remained partly contracted all the time. Short twitching contractions occurred in various parts of the body, and broke the wave-like contraction into a complex, uncoordinated movement. The larvae moved continuously in this \vay, some of them for twenty or more hours. The majority of such larvae died within ten hours, but some of them lived for twenty- thirty hours after the injection. The symptoms were the same, whether last (3rd) or late 2nd instar larvae were used for the experiments. In control experiments, where emulsion containing no DDT was injected, no such symptoms occurred. ETHER NARCOSIS AND DDT SYMPTOMS Normal flies etherized only until their movements stopped were completely relaxed if removed at that time from the ether. Their wings were in normal resting position and their legs were bent in the way assumed by flies at rest. Such flies recovered about one half hour after their removal from the anaesthesia and then seemingly behaved normally. Yet flies left for a longer time in the etherizer behaved quite differently. Their wings were folded back and their legs were 150 DIETRICH BODENSTEIN stretched out and held stiffly away from the body. Usually flies thus over-etherized did not recover from the narcosis and died in that position. The reaction of these differently etherized flies to DDT poisoning was quite interesting. The slightly etherized individuals showed, after DDT injection, the typical DDT symptoms described in the foregoing section. On the other hand, when over-narcotized flies were injected with DDT, one noticed that their legs and FIGURE 1. Diagram illustrating larval ligature experiments for separating the central nervous system from the rear part of the body. Note location of central nervous system (stippled) in anterior part. wings did not respond to the poison. Since such flies never recovered from the narcosis it might be thought that they were already dead at the time of the DDT injection. This, however, was not the case, for their abdomens showed the typical DDT convulsion. These convulsions continued for about three hours, but were somewhat weaker than those of the slightly etherized animals. About that time the uninjected but over-etherized control animals were still completely immobile and apparently dead. In comparing the uninjected and injected flies, one gained the impression that the DDT treatment in some way partly released the ether block. LOCUS OF ACTION OF DDT IN FLIES 151 THE IMPORTANCE OF THE CENTRAL NERVOUS SYSTEM The central nervous system of Drosophila is concentrated in the anterior part of the body. In the larva the central nervous system is located in the third thoracic and the first abdominal segments; it consists of the two brain hemispheres, to which is attached a large ganglionic mass (Figs. 1 and 3). This large ganglion is a compound structure, for it includes the sub-oesopliageal ganglion, the three thoracic ganglia, and the eight abdominal ganglia. In the adult insect the three thoracic ganglia are separated, but all eight abdominal ganglia unite into one ganglionic mass which extends into the first segment of the abdomen (Fig. 2). FIGURE 2. Diagram illustrating ligature experiments on adult flies, showing the separation of progressively smaller abdominal parts from the anterior region of the animal. Note location of central nervous system in anterior part of fly. The described topography allowed one to separate experimentally large parts of the insect body from its central nervous system. The separation was accom- plished by means of a fine silk ligature tied around the body of the animal. De- pending upon the position of the ligature, smaller or larger parts of larvae and adults were thus isolated. In all experiments of this kind, the part in front of the ligature was cut away, in order to be sure that no connection to the ganglia remained (Figs. 1 and 2). Abdominal parts isolated in this way were completely motionless and stayed alive for several days. The most useful symptom for the experimental approach was the contraction of the abdominal wall muscles under the influence of DDT. This effect was very uniform, clearly observable, and lasted a considerable time. The question then arose : were these movements under the control of the central nervous system ? 152 DIETRICH BODENSTEIN Jn other words, was the muscular activity caused by an effect of DDT on the central nervous system? To test this possibility, isolated portions of larval and adult abdomens containing no ganglia were injected with DDT emulsion. The response to the injection was called forth immediately. Both larval as well as adult abdomens exhibited the same typical movements that were observed in the abdomens of injected intact flies. Again, as in the intact animals, it was found that the isolated adult abdomen responded somewhat faster than the isolated larval abdomen. That is, the larval abdomen responded at about thirty-sixty seconds and the adult abdomen about twenty seconds after the injection. The DDT convulsions of the isolated parts continued for a considerable time, but not as long as in the intact animals. In the isolated adult abdomen the convulsions clearly became weaker two hours after injection, yet weak contractions were observed five hours after the injection. In the isolated larval abdomens the contractions were still strong four hours after the injection, yet these parts never survived for twelve hours as injected whole larvae commonly did. A similar sequence of events was observed when only the distal two or three segments of the adult abdomen were isolated from the rest of the body and then injected with DDT emulsion (Fig. 2) or when only two or three segments from the middle of the larval abdomen were isolated by means of two ligatures and then injected. These findings showed that the central nervous system did not control the abdominal symptoms produced by DDT, since they occurred also in the absence of the central nervous system. However, it must be recognized that this does not imply that the central nervous system was unaffected by DDT. BODY FRAGMENTS AXD DDT SYMPTOMS In order to reduce the structural complexity of the insect body, in an effort to localize more closely the site of action of DDT, the following fragmentation experi- ments were performed. Rectangular pieces of skin were cut from the dorsal, lateral, or ventral wall of adult flies and placed in a drop of saline solution. These pieces, about two segments wide, included some muscles of the body wall, fat tissue, tracheae, and nerves but no ganglia. The pieces of dorsal body wall included also part of the heart tube and some alary muscles. In saline solution these pieces remained motionless, but if a few drops of DDT emulsion were added, the isolated muscles in the piece began to twitch. This response began about thirty seconds after DDT was added and continued for about two or three minutes. In control experiments where emulsion containing no DDT was added, no such response occurred. These experiments confirmed those above by showing that the central nervous system was not necessary for the response of the abdominal muscles to DDT. It would appear that DDT affected either the muscles directly or the peripheral nerves. One way of distinguishing between these two possibilities would be fo test a nerve-free muscle preparation with DDT. Technically, however, it is imp< ssible to obtain such a preparation. Other methods were sought to settle this qu stion and are described below. LOCUS OF ACTION OF DDT IN FLIES 153 TREATMENT WITH PHENOBARBITAL Phenobarbital is a drug which depresses central nervous activity in vertebrates.- The effect of this substance on flies has not been described previously. When a ten per cent phenobarbital saline solution was injected into the abdomen of adult flies the animals were apparently killed instantly. The slight vibrating movements characteristic of lightly etherized flies stopped immediately after the injection. All of the muscles seemed to relax and the legs and wings were held in normal position. Flies which had just come out of ether narcosis but were still sluggish lost their coordination when injected with a one per cent phenobarbital solution. Wings, legs, and abdomen moved uncoordinatedly for several hours until the animal finally died. The movement of the abdomen, also uncoordinated, was very different from the symptoms produced by DDT. Injection of 0.1 per cent phenobarbital solution into adult flies induced narcosis, followed by complete recovery of the animal. One hour after the injection some flies crawled about slowly, while others \vere still unable to hold themselves vip and fell over from time to time. By this time the effect of the ether had worn off and these symptoms were regarded as phenobarbital effects. One-half hour later the coordination of the animals had improved but their movements were still slow. But the next day the animals had recovered and behaved normally. That phenobarbital affects the nerves rather than the muscles in flies was indicated by the following experiments. Fully grown larvae were split open along the mid-dorsal line. The intestine, Malipighian tubes, fat body and the main tracheae were then removed. Care was taken not to disturb the brain and ganglia in the anterior part of the body. This manipulation was carried out in physiologi- cal salt solution. The skin with its muscular layer and the adhering nervous system was placed in a small dish with a wax bottom and covered with fresh physiological salt solution. The preparation was then stretched out and by means of fine pins was fastened to the wax bottom of the culture dish, as shown in Figure 3. In this condition the tissues stayed alive for several hours. From time to time the muscles contracted, showing the typical wave-like contraction pattern of a crawling larva. After these movements had been observed for ten minutes the physiological solution was removed and replaced by a one per cent phenobarbital solution. The movements in the preparation stopped immediately. Again, after about ten minutes during which time no movements occurred the phenobarbital solution was removed and replaced by physiological saline. This solution was in quick succession drawn off and replaced about two or three times. The preparation was thus washed clean of any phenobarbital. It was now observed that normal movements had returned. Ten minutes later the physiological solution was removed and replaced by one per cent phenobarbital solution ; all movements again stopped. When washed and placed in physiological solution again the movements in the preparation reappeared. This procedure was repeated three or four times at intervals of about ten minutes and stil} the tissues continued to contract spontaneously when in physiological solution The;,e experiments indicate that the one per cent phenobarbital solution was appare. tly not injurious to the tissues for the length of exposure tested. More- over, while in phenobarbital, no muscular movement was observed, yet if the 154 DIETRICH BODENSTEIN muscles were stimulated directly by touching them with a fine needle, localized contraction in the region of the stimulus was noted. Following the mechanical stimulus the activated muscles contracted rather rapidly, stayed in the contracted state for some time and relaxed very slowly. This localized response of the muscles to mechanical stimuli when under phenobarbital indicates that the drug had not paralyzed the muscles directly but rather the nerves. front spiracle brain compound ganglion nerves attachment pins rear spiracle FIGURE 3. Diagrammatic representation of the skin muscle preparation of a whole larva, pinned to the wax bottom of the culture dish. COMBINED PHENOBARBITAL AND DDT TREATMENT The knowledge that phenobarbital apparently acted on the nerves but not on the muscles provided a tool for determining whether DDT affects the nervous tissue or the muscles. If DDT could affect the muscles directly DDT symptoms should be provoked in animals paralyzed by phenobarbital. If, on the other hand, DDT affects only the nerves, no DDT symptoms should occur in animals treated with phenobarbital. That the latter is the case was shown by the following experiments. In physiological solution the muscle preparation (Fig. 3) showed spontaneous crawling movements as described above. If a small amount of DDT emulsion was dropped onto such a preparation, the rhythm of the movements was interrupted ; it became uncoordinated rapidly and resembled the movements observed in DDT- treated larval abdomens. If, on the other hand, DDT emulsion was dropped in the same manner onto preparations kept in one per cent phenobarbital solution, no LOCUS OF ACTION OF DDT IN FLIES 155 movement whatsoever took place. Moreover, if the physiological solution con- taining DDT emulsion was washed off from the actively moving preparation, and was replaced by one per cent phenobarbital solution, the movements ceased immedi- ately. However, it was impossible to bring the preparation once treated with DDT back to its normal way of movement by washing the DDT solution off and replacing it with pure physiological solution. Such preparations still showed DDT symp- toms. Apparently it was impossible to wash all the DDT out by the methods used. The DDT symptoms were of course stopped by placing the preparation again into phenobarbital. Similar results were obtained when DDT-treated larval or adult abdominal parts w-ere injected with one per cent phenobarbital solution. The DDT symp- toms of such parts ceased immediately after phenobarbital was administered by injection. The reciprocal experiment, where phenobarbital was injected first, yielded the same results, for when DDT was injected into such treated abdomens no DDT symptoms occurred. The symptoms of whole larvae or adults which had been treated with DDT were also stopped immediately by injecting a one per cent phenobarbital solution. Certainly the effects of phenobarbital on isolated, DDT-treated abdomens indi- cate that the absence of the central nervous system did not limit the action of the drug, showing that in insects phenobarbital may act on the peripheral nerves. At this point the discussion of the phenobarbital effects on DDT poisoning must be augmented by an experiment showing how phenobarbital in weaker concentra- tions caused at least a partial recovery from DDT poisoning. It has been stated before that the legs and wings of adult flies injected with DDT were completely paralyzed five minutes after the injection. Now, when 0.1 per cent phenobarbital solution was injected into such animals, the movements of their legs were restored. These movements were uncoordinated and were similar to those observed in the beginning stages of DDT poisoning before the organism went into spasm. Also, the wings were able to move somewhat and were not folded backward. Even the convulsions of the abdomen were much less pronounced. Leg movements continued for about two hours, which of course was a much longer time than ever noted in animals injected with DDT only. These findings clearly show the antagonistic effect of phenobarbital on DDT. CAPACITY OF DIFFERENT TISSUES FOR GROWTH AND DIFFERENTIATION AFTER DDT POISONING If it is true, as the experiments seemed to indicate, that DDT affects only the nervous system, one might expect other tissues to be largely unharmed by the DDT treatment. This expectation can be tested experimentally. It is known (Boden- stein, 1943) that larval tissues will grow and differentiate normally when trans- planted into the abdomen of larval or adult flies. In the larvae the transplanted tissues will develop in synchrony with their hosts to imaginal completion, and in adult flies, the transplant will undergo a considerable amount of growth. The capacity of the tissue for growth and differentiation, it is believed, offers a good criterion for testing the condition the tissue is in after being exposed to the poison. Tissues affected by DDT should certainly not develop normally. The following experiments were designed to clarify this issue. 156 DIETRICH BODENSTKIN * Fifteen last-instar larvae were injected with DDT and placed on moist filter paper. Twenty-two hours later four larvae were still alive and showed typical DDT symptoms. Three of these larvae were opened and their eye discs, leg discs, and antennal discs dissected out. These discs are the primordia for the future eye, leg, and antenna of the adult fly. They were transplanted into the ahdomen of the mature larvae, one disc to each host. The several hosts carrying the transplants from one donor were kept separated from the hosts carrying the transplants of the other donors. If the transplanted organs were ahlc to develop normally, they should have been found as fully differentiated organs in the abdominal cavity of their respective hosts after metamorphosis. From the fifteen original hosts comprising the cases of all three series, eleven survived the operation. These animals completed their metamorphosis seven days after the operation and emerged. They were then dis- sected. Three completely differentiated legs, one eye, and one antenna supplied by the first donor were recovered. Hosts which received transplants from the second donor yielded three legs, two antennae, and two eyes, and from the third donor, three legs. All transplants were fully differentiated. As far as the detailed morphological differentiation of the tested organs was concerned, they were found to be completely normal, and there was no reason to believe that the histological differentiation likewise was not normal. For another experimental series, five leg discs, dissected from the fourth living larva of the original set of DDT-treated animals, were transplanted into the ab- domen of five adult flies. Each host in addition to the leg disc also received two ring glands. This structure is necessary for the continued growth of the trans- plant, for it furnishes a growth-promoting hormone (Bodenstein, 1943). Three days after the operation one host was killed and the leg disc dissected. It had grown but little. Three other hosts were killed six days after the operation and the transplants dissected. These leg discs had clearly become larger. Finally the last host was killed 24 days after the operation and the disc dissected. In this case the transplant had grown considerably and had reached an advanced state of differentiation. The results of this series of experiments were very similar to those obtained in transplanting normal discs in the same manner (Bodenstein, 1939 and 1943). In conclusion, these two experimental series show that exposure to DDT for twenty hours in no way affected the capacity of the imaginal tissues for growth and differentiation. Hence these findings are further evidence that the nervous system alone is affected by DDT. Sl'M MARY 1. The larvae and adults of Drosophilla virilis were fatally poisoned by injecting a one per cent DDT emulsion into the abdominal cavity. The poison produced a typical pattern of symptoms. 2. The neuromuscular system of the wings and legs was apparently very sensi- tive to the poison, for they went into spasm long before the muscles of the ab- dominal wall. There was also a difference in sensitivity to the poison between the larva and adult, the larva being more resistant to the DDT emulsion. LOCUS OF ACTION OF DDT IN FLIES 157 3. Phenobarbital was found to affect the nervous system. Paralysis by pheno- barbital was also produced in the absence of the central ganglia. This shows that the drug also affected the peripheral nerves. Muscles of larvae treated with pheno- barbital responded to mechanical stimulation. 4. Since DDT produced no symptoms in animals treated with phenobarbital and since animals treated with DDT lost their DDT symptoms when injected with phenobarbital, it was shown that DDT acted on the nervous system. Moreover, body parts which had been isolated from the central nervous system and then treated with DDT stopped convulsing after phenobarbital administration. This shows that DDT affected the peripheral nerves. 5. The methods used do not allow one to determine what part of the peripheral nervous system might be affected. There are three possibilities. The poison might affect (1) the motor nerves leading to the periphery; (2) the myoneural junctions; (3) the peripheral nerve net. It is however still questionable whether such a nerve net exists in Drosophila. 6. The antagonistic effect of phenobarbital on DDT was clearly indicated by the fact that the spasm of the legs and wings in DDT-treated flies was partly relieved by treatment with phenobarbital. 7. The conception that only the nervous system is affected by DDT has been strengthened by the fact that larval organs (imaginal discs) which had been ex- posed to DDT for twenty hours grew and differentiated normally when trans- planted into untreated larvae. LITERATURE CITED BODENSTEIN, D., 1939. Investigations on the problem of metamorphosis IV* ; Developmental relations of interspecific organ transplants in Drosophila. Jour. Exp. Zoo/., 82 : 1-30. BODENSTEIN, D., 1943. Hormones and tissue competence in the development of Drosophila. Biol. Bull., 84 : 34-58. CHADWICK, L. E., 1945. Personal communication. TOBIAS, J. M., 1945. Personal communication. YEAGER, J. F. AND MUNSON, S. O., 1945. Physiological evidence of the site of action of DDT. Science, 102 : No. 2647, 305-307. THE CONTRACTILE VACUOLE AND THE ADJUSTMENT TO CHANGING CONCENTRATION IN FRESH WATER AMOEBAE DWIGHT L. HOPKINS * Mundelcin College, Chicago, Illinois INTRODUCTION In considering the relationship existing between an organism and its environ- ment, it is necessary to distinguish clearly between the environment of the organism and the environment of the individual cells. The effect of the environment on an organism will depend specifically on the extent to which the cellular environment is controlled by the organism as a whole. As pointed out by Krogh (1939), the osmotic environment of the tissue cells of the vertebrates — the body fluid — is main- tained at a very constant level. This is true to a lesser degree among the lower organisms. In the free-living protozoa, the cellular environment is the outside environment of the organism and is controlled little, if at all, by the organism. We find consequently that the individual cells of higher organisms have lost their re- sistance to osmotic changes while the cells of many of the protozoa, of necessity, have retained the ability to withstand considerable variation in the osmotic environ- ment. It is well known that certain bacteria and molds can withstand enormous osmotic changes — which also have been shown to be true of a considerable number of protozoa (Hopkins, 1938; Butts. 1935; Mast and Hopkins. 1941 ; and Kitching, 1938). The nature of the adjustments made by cells capable of withstanding great osmotic changes is not clearly understood. In the literature, the assumption is rather generally made that the osmotic concentration of the cellular contents must be maintained at a fairly constant value in order to be consistent with the metabolic processes. This assumption is made, regardless of the fact that evidence for the constancy of the intracellular osmotic pressure is conspicuously absent, especially for the lower organisms. We find that the prevailing opinion as to the function of the contractile vacuole of the protozoa is that it is an organelle which operates to maintain the intracellular osmotic pressure at a relatively constant level which is higher than that of the environment. This theory maintains that excess water, diffusing through the outer membrane into the hypertonic cell, is extracted, isolated into the vacuole, and thus is eliminated when the vacuole is discharged (Kitching, 1938). However, Mast and Hopkins (1941) have shown that the marine Amoeba mira (Flabellula mira Schaeffer), which can withstand enormous changes in the 1 1 am greatly indebted to the American Philosophical Society for grants which have made this investigation possible. I am also much indebted to the following students of Mundelein College for valuable technical assistance : Collette Bergeron, Winifred Greenspahn, Dorothy Homan, Marie Kioebge, and Catherine Miller. To Kathleen Warner, Instructor in Biology at Mundelcin College, is due v.nttrful acknowledgment for help in the preparation of the manuscript. 158 FUNCTION OF CONTRACTILE VACUOLES 159 concentrations of the salts of sea water, does not maintain a constant intracellular osmotic pressure. The intracellular pressure varies with that of the environment. They have shown that, if the amoeba is living in, and is adjusted to a medium, the intracellular pressure is always only slightly higher than that of the environment. Thus, a change in the osmotic pressure of the environment results in a correspond- ing change in the intracellular pressure. In view of the fact that Amoeba mira is a marine organism and does not form contractile vacuoles even in dilute media, it was considered important to extend the investigations to a fresh water amoeba, which forms and eliminates contractile vacuoles, in order to ascertain whether these forms actually maintain a constant osmotic pressure by means of the action of the contractile vacuoles. The amoeba selected was a fresh water form collected from a swamp near Chicago. A single amoeba was isolated and allowed to multiply on an agar culture, using the methods of Rice (1935). All experiments were performed on the progeny of this amoeba. METHODS AND RESULTS Description of the amoeba As in the case of most amoebae, it is difficult to decide upon its correct name. Figure 1 is a composite diagram illustrating its structure as observed in fresh water. Dujardin's Amoeba lacerata (1841) fits it fairly well. We shall then adopt the name of Amoeba lacerata Dujardin. Schaeffer (1926) has placed amoebae of this type in the genus Mayorella Schaeffer. Thus, using the system of Schaeffer, the name is Mayorella lacerata Dujardin. Amoeba lacerata is a small amoeba, rounds up readily when disturbed, and in that condition has an average diameter of fifteen micra or a volume of about 1500 cubic micra. It varies from a limacine to a flattened fan shape and has a clear hyaline border (Fig. la) which is thick anteriorly and very thin posteriorly. A few, fine, pointed, indeterminate pseudopodia may project from the surface — an- teriorly, superiorly, and posteriorly. Cytoplasmic granules are numerous — (1) an abundance of small, barely visible granules and (2) larger granules (or vacuoles) which are less numerous and more variable in number and size (Fig. 16). The living nucleus (Fig. Id) is spherical with a central endosome. Food is engulfed into the under surface as the amoeba moves over the food particles. Except in rare cases, food is engulfed with little or no water. Ingested food particles coalesce with clear vacuoles (Fig. Ir) of the cytoplasm, and the resulting small food vacuoles coalesce with each other as they do in Amoeba mira (Hopkins, 1938) to form large food vacuoles (Fig. If). These vacuoles are readily stained with Nile blue sulfate and neutral red. Egestion of food vacuoles (Fig. \g) occurs at the posterior tip. No direct connection or relationship between the food vacuoles and the con- tractile vacuoles has been observed except that occasionally a food vacuole may be ruptured into the contractile vacuole instead of to the outside. At egestion, the food vacuoles are rarely large and watery. Just before egestion they may be ob- served to decrease slowly in size until little more than food residues remain. This would indicate a protoplasmic absorption of water from the food vacuoles. It is also interesting to note that at the time when this apparent absorption takes place, the food vacuoles are located posteriorly and in close proximity to the region of 160 DWIGHT L. HOPKINS contractile vacuole formation which MacLennan (1941) and others have called the nephridioplasm. It is entirely possible that the water leaving the food vacuoles is absorbed into the contractile vacuoles. The contractile vacuoles In an amoeba moving forward rapidly, the contractile vacuoles form at the posterior end, as shown in Figure If. When the individuals are flattened out and feeding actively, it is difficult to distinguish the posterior end unless the location o • •" • ... *SKi SSSSI FIGURE 1. Diagram showing the characteristic structure of Amoeba lacerata Dujardin. a, hyaline border; b, large cytoplasmic granule (vacuole) ; c, contractile vacuole; d, nucleus; e, small clear vacuole adhering to a recently engulfed bacterium ; /, food vacuole ; g, food vacuole in position for egestion ; /;, nephridioplasm. of the contractile vacuoles is accepted as a distinguishing feature of the posterior part of the cell. Observations on the origin of contractile vacuoles in active amoebae have revealed the following interesting facts: FUNCTION OF CONTRACTILF. VACUOLES 161 (1) They develop in the posterior third of the cell, or, if they begin to develop in an apparently anterior part of the cell, that part will soon become the posterior part — as evidenced by change in the direction of protoplasmic streaming. If the posterior region of the amoeba becomes separated into two parts by a large food vacuole, contractile vacuoles will arise and be expelled independently on both sides. Pressure on the amoeba by a coverslip will cause contractile vacuoles to arise and be expelled independently in several parts of the cell. This phenomenon was also observed by Rowland (1924b) in Amoeba vcrrucosa. A freely moving amoeba, as a rule, has only one formative region for contractile vacuoles. f,~ a —a _a -~ c FIGURE 2. Diagrams showing contractile vacuole cycles, 1,2,3 and 4. Vacuoles a, are expelled to outside leaving residues, b, which by coalescence with other vacuoles, small and large, and swelling, form vacuoles, c, which are expelled. (2) The nephridioplasm (Fig. l/i) is in the region in which plasmogel is rapidly becoming plasmosol. This region is often under considerable pressure (Mast, 1926). This is evidenced in Amoeba lacerata by the observations that the contractile vacuoles are often squeezed into oblong shapes, that food vacuoles and contractile vacuoles which normally never coalesce are sometimes squeezed so tightly together that the food vacuoles rupture violently into the contractile vacuole. The contraction of the contractile vacuole is, however, not always violent. Some- times the contraction is very slow, while at other times, it may be only partial. (3) While the contractile vacuoles increase in size by absorption, as will be shown later, the most conspicuous methods of growth are the coalescence of small droplets or vacuoles with one another, of large vacuoles with each other, and of small vacuoles with large vacuoles (Fig. 2). Small vacuoles less than 0.5 micron in diameter — can be observed coalescing with large vacuoles whose diameters are many times greater. Careful focusing on the outer boundary of a large contractile vacuole reveals this phenomenon. The coalescence of two contractile vacuoles is seldom violent. It occurs at about the same speed as the coalescence of two drops of olive oil. (4) When a vacuole contracts, its contents are expelled to the outside. This occurs usually when the vacuole has a diameter of about 4 micra, but often smaller vacuoles contract, and larger ones may be formed before contraction occurs. Rup- 162 DWIGHT L. HOPKINS ture usually takes place through the postero-superior surface of the cell. Some- times the contracting vacuole leaves no residue, but usually a small residue (Fig. 2h] remains, which coalesces with other small vacuoles forming new, large vacuoles. The residues are usually spherical, but often may assume oblong, stellate, or other shapes. "Whether a residue remains or not, the region from which the previous vacuole was expelled contains small vacuoles which coalesce to form a large vacuole (Fig. 2r). These small droplets, or vacuoles, are continually appearing in the nephridioplasm. The region is generally devoid of food vacuoles. The diagrams in Figure 2 illustrate four sequences which have been observed to occur in the nephridioplasm. (5) All efforts to associate the contractile vacuoles with cytoplasmic granules of any sort have been futile. - Tolerance of Amoeba laccrata for cliangcs in the concentration of the medium The swamp water from which the amoeba was collected was very dilute. The amoebae were first cultured in water from Lake Michigan to which grains of wheat were added. Attempts to culture it in various dilutions of sea water proved quite successful. It can be transferred, without deleterious effects, directly from lake water cultures to 5 per cent sea water to which grains of wheat are added. Excel- lent cultures of healthy amoebae develop within five days. Direct transference of the amoebae from 5 per cent sea water cultures to similar cultures made up in various dilutions of sea water results in excellent growth, even when the percentage is as high as 50 per cent. Direct transference to cultures having percentages higher than 50 per cent results for the most part in complete failure or very slow growth. If, however, the amoebae are first cultured in 50 per cent sea water and then trans- ferred to 75 per cent or 100 per cent, survival and growth result. Slow growth may even be obtained in 125 per cent by adjusting the amoebae to 100 per cent and then transferring them to 125 per cent sea water. The volume of the amoebae and the concentration of the culture medium A study of the relation between the volume of the amoebae and the concentration of the culture medium in which they were grown has been made. The amoebae for this study were all from the same 5 per cent sea water culture and were inoculated into cultures of varying concentrations from one per cent to 75 per cent sea water. To each culture five sterile grains of wheat were added. Five cultures were set up at each concentration in exactly the same way and at the same time. When abun- dant growth and reproduction had occurred in all the cultures the amoebae were caused to round up by pipetting the culture fluid back and forth. Their diameters were then measured under the microscope with an eyepiece micrometer. At least 100 amoebae were measured for each concentration. The age of the cultures was the same at the time of measurement. A total of over 1,750 amoebae were meas- ured. The results are presented in Figure 3, in which the average volume is plotted against the concentration of the culture. The increase in volume with an increase in concentration appears significant statistically. (The standard errors are graph- - The very small vacuok-s which form the larger contractile vacuoles undoubtedly have hern mistaken for granules by various authors. FUNCTION OF CONTRACTILE VACUOLES 163 ically indicated in the curve by the length of the vertical lines drawn through the points.) The total range of variation in average volume is, however, only a little over 750 /*3. This increase in volume with concentration, if significant, is probably due to a lower rate of fission in respect to growth in higher concentrations than in lower concentrations. It should be mentioned here, however, that much greater 1 1800 rt 1 1400 3 o .s '« 1 000 600 * ^- .^ H- z • 10 TO 20 30 4O 50 60 Concentration in percentage of that of sea water FIGURE 3. Graph showing relation between volume of amoebae and the concentration in which they are grown. variations occur in average volume of amoebae from cultures when the nature of food is allowed to vary more than was the case in this series of cultures. (See average volumes of cultures of Figure 4.) Relation between the protoplasmic osmotic pressure and osmotic pressure of the external medium In amoebae which have become adjusted to 50 per cent sea water, is the cyto- plasmic concentration the same, higher, or lower than that of amoebae adjusted to 5 per cent or 20 per cent sea water ? What relation does the cytoplasmic concen- tration bear to the concentration of the culture medium? Since these amoebae round up into spheres when agitated, the osmotic concen- tration of the protoplasm can be readily approximated. When the amoebae are placed in sea water of a given strength with a resulting decrease in volume, we can safely conclude that the osmotic pressure is higher externally than internally. If, on the other hand, the volume of the amoebae increases, the osmotic pressure of the protoplasm is greater than that of the medium. If, however, the plasmalemma is permeable to the solutes of the medium the permanence of the change will depend upon the degree and speed of solute penetration. 164 DWKiHT L. HOPKINS The results of experiments designed to answer the above questions were as follows : Amoebae raised in 5 per cent sea water, and consequently adjusted to it, are placed in one per cent sea water. They immediately swell to a maximum volume and then gradually decrease in volume until approximately their original volume is regained. The smaller the change in concentration, the more rapid is the return tc original volume. The same is true when amoebae cultured in any strength in which they will live are placed in higher and lower concentrations. 4000 • 3000 tJ c '« 2000 J3 0 1000 6 0 10 15 20 25 30 40 50 60 Concentration in percentage of salts of sea water FIGURE 4. Graph showing volume change occurring immediately when amoebae grown in various percentages of sea water are transferred to lower and higher percentages respectively. A, amoebae grown in 5 per cent transferred to one per cent and 10 per cent; B, grown in 20 per cent and transferred to 15 per cent and 30 per cent; C, grown in 50 per cent and transferred to 40 per cent and 60 per cent. The results of a statistical study of the initial changes in volume when the concentration of the medium is increased or decreased from that of the culture mi-dium are given in Figure 4. In this experiment the amoebae were cultured in 5 per cent, 20 per cent, and 50 per cent sea water (Fig. 4A, B, and C respectively). The amoebae were agitated in the culture ; the diameters of from fifty to one hun- dred individuals measured ; the volumes calculated and averaged. The concentra- tion was increased on some and decreased on others; and then the volumes of fifty amoebae whose medium was concentrated, and fifty amoebae whose medium was diluted were measured and averaged. Measurements of volume were all made between 5 and 30 minutes after the concentration was changed. In the figure, the FUNCTION OF CONTRACTILE VACUOLES 165 circles give the average volume of the amoebae in culture medium. The other points represent the average volume of amoebae which have been transferred to the concentration indicated. Volume is indicated on the ordinates. The concentration of the culture is indicated on the abscissas. The length of the vertical lines through the points represents the extent of the standard error. These results indicate very definitely that the osmotic pressure of the proto- plasm of Amoeba laccrata varies in the same direction as that of the external medium as was found to be the case in Amoeba mira (Mast and Hopkins, 1941). The osmotic pressure of protoplasm of amoebae cultured in 5 per cent is not over that of 10 per cent, nor under that of one per cent sea water. The osmotic pres- sure of those cultured in 20 per cent sea water is not over that of 30 per cent nor under that of 15 per cent. Finally the osmotic pressure of those cultured in 50 per cent sea water is not over that of 60 per cent nor under that of 40 per cent. Due to the rapid return of the amoebae to their original volume when small concentra- tion changes were made, we were unable to make more accurate measurements of protoplasmic concentration. Therefore, we are unable to say definitely whether the adjusted amoebae were isotonic, slightly hypotonic, or hypertonic to the culture medium. Since the amoeba can live as well in 50 per cent sea water as it can in 5 per cent, and since the protoplasm of amoebae living in 50 per cent assumes a concentration greater than that of 40 per cent sea water, a wide range of proto- plasmic concentrations appears to be compatible with the metabolic processes. Factors affecting the activity of tJie contractile vacuole system Contractile vacuoles form and are expelled in all concentrations of sea water in which the amoebae will live and reproduce. The rate of formation and expulsion is, of course, slower in concentrated than in dilute culture media. On encystment, growth and expulsion of contractile vacuoles cease and are resumed when excyst- ment occurs. Kitching (1938) has shown that the oxygen tension of the medium is an important factor in the activities of the vacuoles of some protozoa, including various Peritrichs, Amoeba proteits, and Amoeba mira. During the present ex- periments, it has been observed that if the amoebae are sealed under a coverslip and left for some time, the rate of vacuolar expulsion is invariably retarded. On the other hand, if fresh aerated culture fluid is continuously circulated under the cover- slip, the vacuoles remain active. This would suggest that oxygen is important in the activity of the contractile vacuoles. More precise experiments on the effect of oxygen are necessary. A study of the effect of the hydrogen ion concentration of the medium was made. Sudden increases and decreases cause an increase in the rate of vacuolar activity. On the other hand, if amoebae are adjusted to and living at a given pH, there is no consistent relation between the pH of the medium and the rate of fluid elimina- tion. The effect oj the concentration of the culture medium on the activity of the con- tractile vaatoles An extensive study was made of the relation between the concentration of the culture medium and the rate of fluid elimination by means of the contractile vacu- oles. In view of the fact that uncontrollable factors act to affect the activity of the 166 DW1GHT L. HOPKINS vacuoles, the study had to be statistical. For these experiments amoebae were inoculated into cultures made by adding six grains of wheat to various percentages of sea water in Petri dishes. The rate of elimination was measured only after adjustment evidenced by growth and reproduction was complete. Measurements were made of the rate in amoebae living in 5, 10, 15, 20, 25, and 50 per cent sea water. As we have seen, this amoeba flourishes equally well in all of these concentrations. The rate of fluid elimination was measured as follows : A coverslip was cleansed and dropped onto the bottom of a culture dish; the culture fluid was agitated with a pipette and the amoebae were allowed to settle and attach them- selves onto the coverslip. When attachment had occurred, the coverslip was then inverted and placed on a slide between two parallel streaks of vaseline. The streaks of vaseline served to stabilize the coverslip, prevent the amoebae from being crushed between slide and cover, and allow free circulation of water underneath. During an experiment, fresh aerated sea water of the same strength as the culture medium of the particular amoebae being observed was continuously added at one end of the cover and drawn out at the other by means of a strip of filter paper. Observations were made under an oil immersion lens. The diameters of vacuoles were measured with a micrometer eyepiece and the volumes calculated. The vacuole was always measured just before expulsion. The fluid elimination of an amoeba for a ten minute interval was obtained by adding the volumes of the vacu- oles formed during that time. The amoeba was then mechanically agitated, causing it to round up, and its volume obtained by measuring its diameter while thus rounded. The rate was calculated in terms of the cubic micra eliminated per minute by 100 cubic micra of protoplasm. The results of this study of the rate of elimination of amoebae living and re- producing in various concentrations of sea water are presented in Figure 5. The rates of over one hundred amoebae were measured to obtain the curves in this figure. The rate of each amoeba measured is represented by a point. The average rate of all the amoebae measured in each percentage of sea water is represented by a square, while the highest rate obtained for each concentration of medium is repre- sented by a circle. The average rates have been connected by a curve, as have the maximum rates. A third curve has been drawn whose shape is determined by the formula : RC = K, where R is the rate of elimination, C is the concentration of the medium to which amoebae are adjusted, and K is a constant. If we take into consideration cysts whose rates are zero and which are not in- cluded in the figure, we find that the rates of amoebae grown in any concentration may vary from zero to a maximum. The maximum rates vary inversely with the concentration. Neglecting those amoebae, or cysts, whose rates are zero, the average rates also vary inversely with the concentration. Neither the curve for the maximum rates, nor that for the average is exactly parallel with the curve expressed by the equation RC = K, but the parallelism in both cases is sufficient to justify the conclusion that under ideal conditions in which concentration is the only variable, the equation would express accurately the relationship existing between the concentration of the medium and the rate of elimination. Kitching (1938) has pointed out that if a high unalterable protoplasmic con- centration was maintained bv the action of the contractile vacuoles, the rate of FUNCTION OF CONTRACTILE VACUOLES 167 0 10 ZO 50 HO Concentration in percentage of that of sea water FIGURE 5. Graph showing the relation of the rate of fluid elimination to the concentration of salts in sea water. The amoebae were tested in the concentration in which they were grown. Squares, average rates; circles, maximum rates; dots, individual rates. Solid line is curve for equation; R (rate) X C (concentration) =K (constant). elimination should be inversely proportional to the external concentration, and that the curve should be a straight line whose slope is constant instead of a curve whose slope increases as the external concentration decreases. The curves obtained by Kitching (1938) for Peritrichs, those obtained by Mast and Hopkins (1941) for Amoeba mira, and those obtained here, never approach a straight line. The actual rates are either always too low in the intermediate concentrations or too high in dilute and concentrated solutions, depending of course upon the rate which is assumed to be correct. Therefore, in view of these facts, it is not possible that the correct curve is a straight line. 168 DWIGHT L. HOPKINS The effect oj changing the external concentration on the volume of individual vacuoles (A) Vacuoles remaining in the amoeba: When an amoeba with a large con- tractile vacuole is adjusted to 5 per cent sea water and is placed suddenly in 100 per cent sea water, the amoeba shrinks. By careful observation it can also be observed that the contractile vacuole shrinks. The membrane of a vacuole for a while shows sufficient elasticity to maintain a spherical shape, but the shape is not maintained. As more and more water is drawn out of the vacuole, the mem- brane collapses (see Fig. 6). B% 6EA WATER RtPUACtO BY 100 X SEA WATCR RtPUACtD BY S X SCA WATfR FIGURE 6. Diagram illustrating the collapse of a contractile vacuole when subjected to hypertonic sea water, and its recovery when placed in sea water to which it originally was isotonic. If the amoeba, now in 100 per cent sea water with the vacuole in the condition shown in Figure 6, is returned to 5 per cent sea water, the amoeba and the collapsed vacuole both absorb water and swell. The vacuole soon regains its original volume and may even become larger. It would appear from this experi- ment that the vacuole is acting as a simple osmometer. When the hypertonic sea water was added, the water was extracted from the vacuole but not the dissolved substances. Consequently, when dilute sea water was again added this substance acting osmotically drew water back into the vacuole. (B) Vacuoles removed from the amoeba: In the preceding experiment, it could still be argued that the swelling of the vacuole was due to a secretory process requiring oxygen (Kitching, 1938), since the swelling of the vacuoles took place in the amoeba and in the presence of the cellular activities. If, however, the vacuole was removed from the cell and suspended in the medium free of all contact with protoplasm the influence of other cell parts would be eliminated. By crushing amoebae under a coverslip while watching the vacuoles closely it was observed that the vacuoles do not always disintegrate ; some of them remain intact after they are liberated, free of protoplasm, into the medium. When the vacuoles are thus exposed to the outside medium they swell noticeably, indicating that the elastic forces of the cell were inhibiting swelling to some extent. When such isolated contractile vacuoles from amoebae adjusted to 5 per cent sea water are changed to 100 per cent, they shrink. Then, on being returned to 5 per cent sea water, they swell just as they do while still in the amoebae. It is obvious then that the\ act as simple osnioineters; that their contents contain osmotically active sub- FUNCTION OF CONTRACTILE VACUOLES 169 stances in solution, and that the swelling of the vacuolcs docs not depend upon other protoplasmic processes. The adjustment of Amoeba laccrala to changing concentration We have described the immediate effects upon the amoeba and contractile vacuole of changing the concentration of the medium. The story is not completed, however, until we have described the complete series of changes in volume and in the rate of elimination which occur during the adjustment process. Observations have been made upon the volume and vacuolar activities during the adjustment process, both when amoebae cultured in 5 per cent sea water were changed to 50 per cent and when amoebae cultured in 50 per cent sea water were changed to 5 per cent. (I) Adjustment of amoebae cultured in 5 per cent sea water to 50 per cent sea water : When an actively moving amoeba, raised in 5 per cent sea water, is changed to 50 per cent sea water and allowed to remain, it rounds up and its contractile vacuole shrinks greatly. The amoeba remains for some time in the inactive spherical condition. The contractile vacuole increases in volume very slowly. Re- peated measurements of the diameter of the amoeba itself reveals the fact that it increases slowly in volume. This increase continues until it has regained approxi- mately the value it had in 5 per cent sea water. As soon or soon after it has re- gained this original volume, protoplasmic streaming becomes more definite, attach- ment to the substratum is made, and normal locomotion begins. The rate of elimination of fluid by the vacuole remains at a low level. In Figure 7 A and B are plotted curves showing the observed changes in protoplasmic volume and the rate of elimination of two amoebae when they were suddenly changed from 5 per cent sea water in which they had been cultured to 50 per cent sea water. In these experiments the rate of fluid elimination by the contractile vacuoles was measured, 50 per cent sea water added, and the volume and rate measured again. Since the rate of elimination was so extremely slow it was possible to make only one measure- ment before adjustment was completed, but the diameter of the amoeba was measured at frequent intervals until adjustment was complete and locomotion resumed. (II) Adjustment of amoebae cultured in 50 per cent sea water to 5 per cent sea water : In Figure 7C and D, we have plotted curves showing observed changes in vol- ume and the rate of elimination when amoebae were suddenly changed from their 50 per cent sea water culture to 5 per cent sea water. In these curves it will be observed that (1) the amoebae swell and then shrink until their original cul- tural volume is regained, (2) the rate of elimination increases rapidly to a high value which may either remain high, or after an initial increase may fluctuate, and (3) that the time required for the amoebae to regain their original volume may be long or short. In this experiment it appears quite possible that the activity of the contractile vacuole may have been helpful in bringing the volume of the amoeba back to normal. But we already know from experiments that when adjustment is complete the amoebae will now shrink if placed in 10 per cent sea water. Therefore, while the activity of the contractile vacuole may have helped the amoebae to regain their 170 DWIGHT L. HOPKINS u 1000 4000 3000 2000 1000 2000 1000 4000 3000 2000 1000 3 CULTURE SOLN. 57. < — » c* i L SERIMEN1 50% AL SOLN A +4-+-^ \^^^ -^ 1 ••___> _v 1 /- ^ . Y B i — ' ^^ EXPERU ENTAL s can aRE50UK. n _^ _^__^ 2 ______ c ( >•—-'-». 13 1 1 I I X i -^* D —*- — r >- 5210 1 23 Hours FIGURE 7. FUNCTION OF CONTRACTILE VACUOLES 171 normal volume, the protoplasmic concentration at the same time has dropped from a value approximating that of 50 per cent sea water to a value approximating that of 5 per cent sea water. DISCUSSION Krogh (1939) after reviewing the works of various authors on the functions of the contractile vacuoles of the Protozoa, concluded that the evidence was pre- dominantly in favor of the "osmoregulatory theory," which holds that these vacuoles serve to maintain the internal protoplasmic concentration at a constant level which is higher than that of the outside medium. In view, however, of the work of Buchthal and Peterfi (1937) who found only slight potential differences existing between the protoplasm of Amoeba sphaeronucleus and the outer medium, Krogh concludes his discussion with a "warning against too schematic conceptions." He says further, "It seems possible that the contractile vacuole may come much nearer in its function to the kidney of higher animals than is indicated by direct studies of osmotic regulation." The contents of the contractile vacuoles Weatherby (1927) was unable by means of very delicate tests to show the presence of urea or ammonia in the contractile vacuoles of paramecia. Kitching (1938) points out that the maintenance of a constant concentration higher than the outer medium would require the elimination of fluid more dilute than the proto- plasm. He believes from his experience that it is entirely possible that these vacu- oles contain only distilled water. Ludwig (1928) argues that carbon dioxide is eliminated by way of the contractile vacuoles, but without evidence. The possibility of the elimination of carbon dioxide, carbonic acid, and carbonates by the contractile vacuoles, however, has not been disproven. Various authors have maintained that granular anlagen give rise to contractile vacuoles (Lloyd and Scarth, 1926; Taylor, 1923; Rowland 1924a; MacLennan, 1933). Hopkins (1938) and Mast and Hop- kins (1941) maintain that the food vacuoles of the marine amoeba, Amoeba mira, in addition to being the seat of digestion, act as do contractile vacuoles and that the anlagen of these vacuoles may contain granules. They also show that the sub- stances that make up these granules may never aggregate in the form of granules but remain in solution in the vacuoles or their anlagen. The present observations on Amoeba lacerata do not show definite connections between granules of the cytoplasm and the contractile vacuoles or that their anlagen contain granules, but they do show that they contain osmotically active substances. The concentration of the protoplasm The actual concentration of the protoplasm has never been. ascertained. When we speak of the osmotic strength or activity of a cell we can of necessity refer only to the attraction of the cell for the solvent and compare the strength of this attrac- FIGURE 7. Graphs showing changes in volume and rate of fluid elimination of individual amoebae when they were transferred from the percentage of sea water in which they were grown to lower or higher percentages. Solid curves, volume changes ; broken curves, changes in rate of fluid elimination. Arrows show the direction of the concentration changes. A and B from 5 per cent to 50 per cent sea water ; C and D from 50 per cent to 5 per cent sea water. DWIGHT L. HOPKINS tion to the attraction exerted by known solutions. A cell is said to be isotonic to a solution of known strength when the affinity of the cell and the solution for solvent are equal ; i. e., a further increase in the strength of the solution would shrink the cell. No deductions can be made as to the total molecular concentration of the cell components. The osmotic concentration, the intermolecular attractions and repulsions of the protoplasmic molecules and tendencies of the protoplasm to be- come hydrated are all involved. The tensile or elastic strength of the protoplasm, membranes, and cell wall all work in opposition to the osmotic concentration and hydration. The shrinkage or swelling of a cell depends upon a summation of these inside forces. Most cells are in equilibrium with their environment ; i. e., the summation of forces inside equals those outside. The cells forming contractile vacuoles are said to be exceptions according to the osmosis-secretion theory of Kitching (1938). Cells forming contractile vacuoles live in media much more dilute than solutions necessary to plasmolyze them. Such cells must be absorbing water continuously from their more dilute outer medium and continuously bailing out fluid, more dilute than the protoplasm, by the action of the contractile vacuole. Even though the concentration of the protoplasm is higher than the outside medium, this does not follow. The elastic or tensile strength of the membrane and ecto- plasm or plasmagel, and for that matter intermolecular attractions of the proto- plasmic molecules must exert considerable opposition to the entrance of water. The results here presented and those of Mast and Hopkins (1941) show that the osmotic activity involved in collecting water eliminated by the contractile and food vacuoles is resident in the vacuoles themselves not in the surrounding cytoplasm. The surrounding cytoplasm serves merely as a membrane through which the water is drawn. It is also clearly seen from these results that we cannot speak accurately of the osmotic activity of the protoplasm in general. Only of specified parts which have no internal differentiation can we hope to accurately be specific as to osmotic activity. The concentration of the protoplasm of Amoeba lacerata is only slightly differ- ent from the surrounding medium. The present experiments demonstrate this fact, (see Figures 4 and 7). The results presented in Figure 5 are consistent only if we assume that the concentration of the protoplasm outside of the vacuoles varies in the same direction as the medium. If we also assume that the protoplasm by oxidative or other metabolic processes gives rise to a constant supply of metabolic by-products for isolation into the vacuoles which swell and coalesce in a constant manner (which is true in the average amoeba), then the curve of Figure 5 is ex- plained perfectly. It is unfortunate that the experiments on the rate of elimination could not be continued below 5 per cent sea water. The reason for this inability was the extreme variability of rates obtained. In other words, in extremely dilute media, the rate of elimination showed little or no dependence upon the osmotic pres- sure of the external medium. This leads to the conclusion that in extremely dilute media, the osmotic pressure of the external medium is of little consequence in de- termining the rate of fluid elimination. The junction of the contractile vacuoles If water is absorbed mainly by the cell clue to the osmotic activity of the contents of the contractile vacuoles, the primary function of the vacuoles is not the elimina- tion of water. The ivatcr eliminated by the vacuoles is only wa^r absorbed as a FUNCTION OF CONTRACTILE VACUOLES 173 consequence of their own contents. The osmotically active contents of the vacu- oles are discharged to the outside and are wasted. These wastes are in all proba- bility metabolic by-products. They could be incompletely oxidized food, carbonic acid, carbonates, ammonia, urea, uric acid, or other waste products. Kitching (1938) has shown that activities of contractile vacuoles and volumes of various protozoa are controlled by the oxygen tension. He takes this fact to mean that oxidative energy is utilized in extracting water from the protoplasm and the secretion of this water into the vacuoles. He argues that energy is needed since this is accomplished against an osmotic gradient. He also finds that cyanide inhibits the action of the contractile vacuoles in these forms and that coincident with cyanide inhibition, the body volume increases. When cyanide is removed, the con- tractile vacuoles resume their activity and the cell volume returns to normal. His interpretation of these facts is that the removal of oxygen or the addition of cyanide inhibits oxidation which in turn deprives the vacuoles of energy and they cease extracting water from the protoplasm ; but meanwhile the cytoplasm continues to absorb water from the medium thus swelling the cell. He omits any consideration of the effect of cyanide on protoplasm except the inhibition of oxidation. His in- terpretation does not take into consideration the following possibilities : ( 1 ) While the relation of oxygen tension to vacuolar activity is demonstrated, this relation could be as well explained by the formation of osmotically active substances in vacuoles as a result of oxidation. (2) Cessation of the oxidative process may initiate many changes in the protoplasm, for example, its degenerative breakdown which would cause swelling. Kitching's own results, the results of others, and the present results are more consistent with the conclusion that the contractile vacuoles function to eliminate metabolic wastes, and that the elimination of water is merely a consequence of the osmotic activity of these wastes. When one stops to consider the possibility of the formation of vacuoles more dilute than the protoplasm, it is seen to be an impossibility and contrary to known physico-chemical laws for the following reasons. For it to be possible, water would have to be isolated and separated against solution and chemical forces. This water would have to be isolated into regions of equal hydrostatic pressure by some force which cannot exist since water has been attracted into the protoplasm by exactly opposite forces. There are two ways by which water can be separated from proto- plasm, i.e., setting up forces which attract or repulse water more strongly than protoplasm. (1) By chemical changes occurring in regions of the protoplasm, for instance, oxidation which results in localized increases in chemical and osmotic forces (vacuoles). (2) By changes in protoplasm in general. Energy resulting from the oxidation of food can be used in polymerization whereby the forces leading to hydration are lost. This releases water from combination with the protoplasm. The water would tend to collect in localized regions, but during this collection, waste products become dissolved and consequently, the vacuoles so formed would contain waste and salts, not distilled water. Oxidation in the protoplasm enhances the water-combining powers of some substances, but not all. If the vacuoles contain a solution whose osmotic pressure is less than that of the surrounding cytoplasm, then water will be drawn back into the cytoplasm and not expelled to the outside. This statement, however, would not be true if the vacuole membrane were impermeable to water. 174 DWIGHT L. HOPKINS In the experiments concerned with isolated vacuoles, it has been clearly shown that the internal contents of the vacuoles do have an osmotic pressure and that the vacuole membrane is permeable to water. The osmotic pressure of the vacuoles is either equal to or greater tlian that of the cytoplasm. The relation bet-ice en the osmotic activity of the protoplasm as a zvlwle and that of the external medium in u'liich the amoebae live We have demonstrated that the osmotic activity of Amoeba laccrata varies di- rectly with that of the environment. This also is true of Amoeba mira (Mast and Hopkins, 1941) and of the peritrichs investigated by Mast and Bowen (1945). In addition, it appears that Amoeba proteus (Mast and Fowler, 1935), peritrichs (Mast and Bowen, 1945), plant cells (many authors) and many other cells main- tain a higher osmotic pressure than their natural medium. This is the cause of turgidity. In the protozoa with contractile vacuoles as well as other animal and plant cells, the extent of this hypertonicity of a cell over its environment must depend upon the strength of the membranes, and other ectoplasmic structures. The external medium remaining constant, some of the factors that would cause swelling of cells are: (1) weakening of cell boundaries, (2) peptization of proto- plasmic proteins, (3) increase in cellular anabolism without increase in cellular excretion, or (4) merely failure of the excretory processes. The extent of the difference in osmotic pressure existing between the cell con- tents and the outside medium has not been demonstrated either for Amoeba mira or Amoeba lacerata, but if we include consideration of the osmotic pressure of the vacuoles, then the difference is appreciable but variable, being higher inside than outside. Factors determining the rate of vacnolar output In a given protozoan, the rate of fluid output depends on a number of factors. These factors and how they operate are not clearly defined but the following un- doubtedly are involved : ( 1 ) Respiration. The work of Kitching has demonstrated this. If the basic- osmotic pressure of a cell is in equilibrium with that of the environment, the oxi- dative (aerobic or anaerobic) breakdown of foodstuffs will result in an increase in the osmotic activity of the cell as a whole even though these by-products a im- probably isolated into the vacuoles immediately by the formation of a precipitation membrane. It is readily seen from the facts presented here and by Mast and Hop- kins (1941) that the rate of fluid elimination would depend upon the respiratory rate. Condensation of the by-products in the form of condensation granules would decrease their osmotic activity. While the rate of oxidation does have this con- trolling influence upon the rate of fluid elimination, the shape of the curve of Figure 5 is not determined by variations in oxidative rate. The oxygen supply, of course, was not absolutely constant but reasonably so. Since the curve represents an average, the variations are accounted for statistically. (2) Food. In protozoa, food taken into food vacuoles and digested there adds to the over all osmotic activity of the cell, but if digested food is rendered insoluble by condensation into crystals or other food reserve bodies, this osmotic increase is nullified. This storage may not take place before appreciable increase in intake of FUNCTION OF CONTRACTILE VACUOLES 175 water has been affected. Also, the water engulfed with food would naturally have an influence on the rate of fluid uptake, but not upon the output as suggested by Kitching (1939). (3) Agents causing polymerization or breakdown of substances in the proto- plasm. (4) Changes in osmotic activity of outer medium will cause increases or de- creases in water intake and consequently the output. The permanence of these effects will depend upon the permeability of the membrane to the substances in the external medium. Two vacuole systems as opposed to one Amoeba mira (Hopkins, 1938) forms only one system of vacuoles. By this statement it is meant that all vacuoles which arise in this amoeba coalesce to form large cloacal vacuoles, which are eliminated in order of formation. The functions of excertion and digestion are accomplished by this system. On the other hand, Amoeba laccrata forms two entirely different sets of vacuoles: one which goes to make up the contractile vacuoles, and is excretory in function while the other system goes to combine with the engulfed food and is digestive in function. All protozoa probably can be separated into two groups : one primitive group, having only one vacuolar system ; and a specialized group, having two systems. SUMMARY 1. Amoeba laccrata, a fresh water amoeba, is able to adjust and live in any concentration of the salts of sea water up to 125 per cent. 2. Contractile vacuoles are formed in all concentrations and are separate from food vacuoles. 3. Food is engulfed with little or no water and small protoplasmic vacuoles coalesce with it. When digestion is completed, the food residues are eliminated with little or no fluid. 4. The major part of fluid elimination occurs by way of the contractile vacuoles. 5. Contractile vacuoles grow in size by coalescence and osmotic swelling. 6. The rate of fluid elimination is affected by a number of factors, but under constant optimal conditions it varies inversely with the concentration of the medium. It is suggested that the rate is proportional to the rate of catabolism. 7. The osmotic pressure of the protoplasm varies in the same direction as that of the outside medium. The protoplasm of completely adjusted amoebae is very nearly equal to that of the medium, the osmotic difference being less than that of 5 per cent sea water (0.13 atmosphere). 8. Contractile vacuoles have been shown to contain osmotically active sub- stances ; they shrink in hypertonic and swell in hypotonic solutions regardless of whether they remain in the cell or have been removed from the cell. 9. The volume of the amoeba is only temporarily dependent on the concentra- tion of the medium. After adjustment, the volume shows no dependence upon concentration of the medium. If the amoeba is placed in a hypertonic solution the cell shrinks and then swells to its original volume within three hours time. Conversely, if placed in a hypotonic solution, it swells and then shrinks to its original volume. 176 DWIGHT L. HOPKINS 10. The cell membrane is either permeable to substances in the medium especi- ally when adjusting to a change in the concentration of the medium, or is capable of modifying its internal osmotic activity in some other way. 11. The contractile vacuoles are obviously excretory and also osmoregulatory organelles since they remove waste substances which would otherwise cause in- creases in the osmotic pressure in the cell. LITERATURE CITED I'.UCHTHAL, F. AND T. PETEKFi, 1937. Messungcii von Potentialdifferenze an Amoeben. Proto- plasma, 27 : 473-483. BUTTS, HELEN E., 1935. The effect of certain salts of sea water upon reproduction in the marine amoeba Elabellula niira Schaeffer. Physiol. Zool., 8 : 273-289. DUJARDIN, F., 1841. Histoire Naturelles des Infusoires. Pp. 226-239. Paris. HEILBRUNN, L. V., 1943. An Outline of General Physiology. 2nd ed. Philadelphia. HOPKINS, D. L., 1938. The vacuoles and vacuolar activity in the marine amoeba, Flabellula mira Schaeffer and the nature of the neutral red system in protozoa. Biodynamica, No. 34. ROWLAND, RUTH B., 1924a. On excretioa of nitrogenous waste as a function of the contractile vacuole. Jour. Exp. Zool., 40: 231-250. HOWLAND, RUTH B., 1924b. Experiments on the contractile vacuole of Amoeba verrucosa and Paramecium caudatum. ] our. Exp. Zool., 40 : 251-262. KITCHIXG, J. A., 1938. Contractile vacuoles. Biol. Reinews, 13: 403-444. KITCHING, J. A., 1939. The physiology of contractile vacuoles. IV. A note on the sources of water evacuated, and on the function of contractile vacuoles in marine protozoa. Jour. Exp. Biol, 16: 34-37. KROGH, AUGUST, 1939. Osmotic regulation in aquatic animals. 241 pp. Cambridge. LLOYD, F. E. AND G. W. SCARTH, 1926. The origin of vacuoles. Science, 63 : 459-460. LUDWIG, \V., 1928. Der Betreibsstoffwechsel von Paramaecium caudatum Ehrb. Zugleich ein Beitrage zur Frage nach der Function der Kontractilen Vacuolen. Arch. f. Protist., 62: 12-40. MA< LENNOX, R., 1933. The pulsation cycle of the contractile vacuoles in the Ophryosolecidae of cattle. Univ. Cat. Publ. in Zool., 39 : 205-249. MACLENNON, R., 1941. Cytoplasmic inclusions. In "Protozoa in Biological Research." Chap. III. Lancaster. MAST, S. O., 1926. Structure, movement, locomotion, and stimulation in Amoeba. Jour. Morph. and Physiol, 41 : 347-425. MAST, S. O. AND D. L. HOPKINS, 1941. Regulation of the water content of Amoeba mira and adaptation to changes in the osmotic concentration of the surrounding medium. Jour. Cell. Comp. Physiol., 17: 31-48. MAST, S. O. AND W. J. BOWEN, 1945. The food-vacuole in the peritricha, with special refer- ence to the hydrogen-ion concentration of its content and of the cytoplasm. Biol. Bull, 87: 188-222. MAST, S. O. AND C. FOWLER, 1935. Permeability of Amoeba protons to water. Jour. Cell. Comf. f'ltvsiol.. 6: 151-167. RICE, N. E., 1935. The nutrition of Flabellula mira. Arch. I. Protist., 85: 350-368. SCHAEFFER, A. A., 1926. Taxonomy of amoebae. Carnegie Inst. Wash., Publication 345. TAYLOR, C. V., 1923. The contractile vacuole in Euplotes ; an example of sol-gel reversibility of cytoplasm. Jour. Exp. Zool, 37 : 259-289. WEATHERBY, J. H., 1927. The function of the contractile vacuole in Paramecium caudatum; with special reference to the excretion of nitrogenous compounds. Biol. Bull, 52 : 208-218. Vol. 90, No. 3 June, 1946 THE BIOLOGICAL BULLETIN PUBLISHED BY THE MARINE BIOLOGICAL LABORATORY DILUTION MEDIUM AND SURVIVAL OF THE SPERMATOZOA OF ARBACIA PUNCTULATA. II. EFFECT OF THE MEDIUM ON RESPIRATION THRU HAYASHI Department of Zoology, University of Missouri, Columbia, Missouri INTRODUCTION The first section (Hayashi. 1945) of this investigation presented evidence for the existence of a factor in the seminal fluid of Arbacia which influences the duration of fertilizing capacity of the sperm of the same species. The experimental results indicated that the factor was adsorbed on the sperm surface, and subsequently was lost into the surrounding medium. Since Gray (1928a) has shown a relation be- tween "mechanical crowding" of sperm and the length of metabolic life of sperm, the question at once arose as to whether the seminal fluid factor influenced the ferti- lizing power of sperm through an effect on the metabolism of sperm. Conse- quently, a series of experiments was done investigating the effects of seminal fluid on the respiratory activity of sperm. MATERIALS AND METHODS The microrespirometer used for these experiments was a compensating type, a modification of Krogh's respirometer.1 It consisted of two similar vessels of con- ventional design connected through a common U-shaped manometer. To remove the carbon dioxide, filter paper wet with 20 per cent KOH was placed in the manom- eter vessels out of contact with the respiring cells. The manometer vessels were im- mersed in a constant-temperature bath maintained steadily at 25° C. The shaker, upon which the microrespirometers were mounted, moved to and fro in a straight line a distance of 15 centimeters. A steady rate of 40 cycles per minute thoroughly agitated the respiring sperm suspensions. The sensitivity of the respirometer was such that a respiratory rate of one mm." of oxygen consumed per hour could be de- tected in ten-minute readings. Two points of difficulty were encountered in the course of these experiments. First, capillary action prevented the easy transfer of any liquids from the side-arm into the respirometer vessel. Second, the effect of dilution on the respiration of sperm was found to be very rapid, and therefore difficult to measure. To overcome these difficulties, a measured amount of packed sperm was placed on the glass wall 1 Kindly lent by Dr. Titus C. Evans, formerly of Iowa State University. 177 178 THRU HAYASHI of the- manometer vessel above tbe point readied by a given amount of liquid in the vessel, when this liquid was agitated by the movements of the shaker. The packed sperm remained in place by their own adhesive action. When the time came to di- lute the sperm in the medium to be studied, the adhered sperm on the side of the vessel were washed down into the liquid by a slightly greater agitation of the manom- eter vessel. This technique overcame the first of the above-mentioned difficulties satisfactorily, but only partly the second. To wash down and to disperse completely the sperm cells in the medium required time in the order of minutes. Since, during this period, changing numbers of cells were being affected by the medium, neither the total res- piration nor the total respiratory rate could be accurately measured. This effect was minimized by manipulation of the vessels and by extrapolation of the respiration curve. a. z o ^- o. r o -0- SEMINAL FLUID 20 60 30 JOO TlME IN MINUTES 120 140 160 FIGURE 1. Effect of seminal fluid, sea water on respiratory rate of sperm. The pH of the medium in the experiments here presented was not checked after each run, although it was known before the run, and has been published elsewhere (Hayashi, 1945). In later, similar experiments (to be published) the pH was ad- justed to equality and checked before and after each run. In the time of the run, no change in pH occurred, and the results did not differ from those presented. It was noted that the sperm in all the respiration experiments, irrespective of the medium used, and in the concentrations of these experiments, were dead at the end of five hours. A sharp rise in the rate of oxygen consumption occurs at about this time, due, probably, to disintegration of the sperm cells caused by the constant shak- ing of the manometers. This same effect was noted by Gray ( 19281), p. 350). This dtect afforded no trouble in the interpretation of the results, however, for it occurred uniformly in all sperm suspensions, and the pertinent data were obtained before the end of the five-hour period. DILUTION MEDIUM AND SPERM RESPIRATION ENPERIMKNTS AND 179 For the first experiments, the respiration of sperm in seminal fluid was com- pared with the respiration of an equal amount of sperm in sea water. Each sus- pension contained 0.00155 cc. of packed sperm per cc. of medium. The contrast- ing changes in the respiratory rates of the two suspensions are shown in Figure 1. Using the same data, the total oxygen consumed was plotted as a function of time, to produce the curves of Figure 2. The experiment showed that the seminal fluid did not affect the sharp in- crease in respiratory activity immediately following the dilution of the sperm. The main effect of the seminal fluid was to regulate, or delay, the sharp drop in respiratory rate shown by the sperm in sea water. Figure 1 also showed that the total amount of energy expended by equal amounts of sperm in the same length of time was greater when the sperm were suspended in seminal fluid. 30 I 40 60 «O 100 TIME IN MINUTES MO FIGURE 2. Total oxygen consumption as a function of time for sperm in seminal fluid and sea water. As a further check on the effect of dilution on the respiration of sperm, one of Gray's (1928a) experiments was repeated, with both seminal fluid and sea water. A measured quantity of packed sperm (0.00155 cc.) was placed on the wall of the respiration chamber, and 1.0 cc. of medium was placed on the bottom of the vessel. In addition 0.5 cc. of medium was placed in the side arm, and this amount of medium was "dumped" over into the respiration chamber at an appro- priate time. The results of the experiment are shown in Figure 3. The results confirmed Gray, but compared with his results, the degree of rise in the respiratory rate at the second dilution was less. When the sperm concen- trations used by Gray were compared with those used here, it was seen that the present sperm suspensions were far more dilute. Therefore, the results were taken to mean that the concentration of sperm in this experiment was near the 180 TERU HAYASHI upper limit to obtain tin- maximum burst of energy at first dilution from the amount of sperm used. In other words, greater original dilution of the packed sperm would not result in very much more initial activity. This interpretation was partly confirmed by the following experiment. Lillie (1913) had found that egg water stimulated sperm to greater activity, and Gray (1928c) concluded from his experiments that egg water stimulated sperm to greater respiratory activity. However, Gray had used egg water as the first and only diluent for the sperm. In the following experiment, by contrast, the egg water was used as a diluent after the initial burst of activitv induced by the original dilution - j fj of the medium. The same procedure as the preceding experiment was employed. Instead of the respective media in the side arms, however, egg water was used, and this was ** z o p Q. r u. o UJ I- U -©- -O 3EMINAU o -o -©- -©- *0 BO ]00 TlME IN MINUTES 1*0 160 FIGURE 3. Effect of dilution by respective media on sperm in sea water and seminal fluid. "dumped" in at the appropriate time. The results are given in Figure 4. Study of Figure 4 showed that, after the original burst of activity, egg water, far from stimulating the sperm, seemed to inhibit them. The effect was probably due to the agglutination of the sperm, and seemed to overcome the slight stimulation due to dilution shown in Figure 3. DISCUSSION .-Inalysis oj the effect o\ the seminal fluid on sperm respiration The respiratory rate of sperm in seminal fluid is compared with that of sperm in sea water in Figure 1. The seminal fluid does not seem to influence the initial steep rise in activity due to dilution. There was some variation shown in this first burst of respiration, but a check of all the runs made showed that the variations were not significant. In some cases the sperm in seminal fluid showed more intense DILUTION MEDIUM AND SPERM RESPIRATION 181 activity at dilution; in others, the sperm in sea water were more active at dilution. The variations were caused probably by differences in the rate at which the sperm were washed from the sides of the respiratory vessels. The effect of the seminal fluid is on the "senescent period" following the initial respiratory rise. The seminal fluid seems to prevent the rapid decline of the meta- bolic rate evident in the sea- water control. The effect of the seminal fluid on the respiration of sperm thus parallels the effect of seminal fluid on the fertilizing power of sperm. In both cases, the seminal fluid maintains the sperm at a high functional level for a longer time than does sea water. A comparison of the areas under the curves shows that sperm in seminal fluid expend more energy than an equal amount of sperm in sea water during the course of their active lives. There are two possible explanations for these results. It is possible that the sperm in seminal fluid utilized 10 ec 1. z o a 6 z 3 UJ z O u z . Ul ^ o X o a. o Ul 2 -®- -O— 40 60 80 ico TIME IN MINUTES 1*0 FIGURE 4. Effect of dilution by egg water on sperm in sea water and seminal fluid. their internal store of fuel more completely than the sperm in sea water. The second possibility is that the seminal fluid is serving as a source of fuel for the sperm. A clue to these possibilities was sought by utilization of some of Gray's analytical methods. Gray (T928b) originally advanced the view that the sperm cell was a tiny com- bustion engine with a limited supply of fuel. From his original assumption, Gray derived the equation dx Activity = -77 = k(a - x) which, integrated, gave k - - In. a a — x (Equation 1) (Equation la) 182 TERU HAYASHI lii this equation, .r represented the amount of fuel used up in time. /; a, the initial, total amount of fuel in the sperm cell; and k, the activity produced by each unit of fuel consumed. Gray found that this "law of exponential decay of a homogeneous population" did not fit the experimental results. He, therefore, expanded the equa- tion with a consideration of a heterogeneous population, and derived, as one form of the equation. / / \ (Equation 2) in which O, represented the total oxygen consumed during active life; a, another constant ; and .r and /, possessing the same meanings as before. Gray found that this equation could be made to fit the experimental results, and concluded that the sperm population was heterogeneous in a special way, and that its activity decreased because of the depletion of an internal supply of energy. The notion of heterogeneity was used as a basis for a later, more detailed mathematical analysis (Gray, 1930). The above methods were applied to the results of a number of experiments of this investigation. At two different values of time t in a single run. experimental values of the total oxygen consumption were arbitrarily selected, and the values of the constants k, a, (X, and a in equations la and 2 were calculated. The values of the constants thus derived were then substituted back in the equations and the the- oretical values of .r for the entire run calculated and compared to the experimental values. The results showed that, with the proper selection of points, both equations la and 2 could be made to fit the experimental data rather closely, but other points could be selected to demonstrate that neither of the equations fit the experimental data. The latter case, in a typical experiment, is shown in Table I. TABLE I Comparison of the experimental and theoretical values of the total oxygen consumption of sperm in sea water and seminal fluid. Theoretical values calculated from equation la and equation 2 Total oxygen consumed in mm.3 Time (minutes) Sea water Seminal fluid Exp. £=.01275 a =12.24 02 = 15.25 a=71.15 Exp. £=.00876 a =16.31 02=22.54 a = 135.7 0 0.00 0.00 0.00 0.00 0.00 0.00 10 1.05 1.45 1.87 0.95 1.36 1.55 20 2.07 2.75 3.32 1.99 2.62 2.90 30 3.09 3.89 4.50 3.08 3.77 4.08 10 4.06 4.89 5.46 4.06 4.83 5.13 50 4.96 5.77 6.26 5.06 5.78 6.07 60 5.84 6.55 6.94 6.04 6.66 6.91 70 6.68 7.23 7.53 6.97 7.47 7.67 90 8.08 8.36 8.49 8.60 8.90 8.99 100 8.70 8.86 8.88 9.37 9.52 9.56 110 9.23 (9.23) (9.23) 10.09 (10.09) (10.09) 210 11.41 1 1 .40 11.37 13.80 13.72 13.69 220 11.50 (11.50) (11.50) 1 3.94 (13.94) (13.94) 255 11.76 1 1.76 1 1 .90 14.31 14.56 14.71 DILUTION MEDIUM AND SPERM RESPIRATION 183 It may be noted that the points selected are those covering the end of the run, and the data of this portion tit the experimental values rather closely, whereas wide variation is evident in the early portions of the run. Points selected near the begin- ning of the run, in contrast, would show a fit in that portion of the run, and variation at the end. Intermediate points, of course, would show an intermediate fit. From such analyses, these conclusions may he drawn. ( 1) There is no need to assume heterogeneity of the sperm population whether in sea water or seminal fluid. (2) The simple "law of exponential decay" is not adequate to explain sperm ac- tivity ; the decrease in activity of the sperm cells with time is not due to the exhaus- tion of an internal supply of fuel alone, whether the sperm cells are in sea water or seminal fluid. Since sperm cells in sea water show a decline in activity whose controlling factor is not the exhaustion of their internal source of fuel, it is possible that even at the end of active life in this medium, the sperm cells contain unused, potential energy. The utilization of this unused energy by the sperm in seminal fluid would explain the increased oxygen consumption of the sperm cells in seminal fluid. There is, however, the possibility that the sperm in sea water reach the end of active life be- cause of starving plus other factors. The question of nutrition by seminal fluid, therefore, is unsettled by the above analyses, although strong possibilities are afforded. Gray (1928b) analyzed the senescence of spermatozoa from still other considera- tions. He assumed that the cause of the death of the spermatozoon was the accumu- lation of products of metabolic activity inside the sperm cell. This would render inactive part of the active system originally present in the cell. Active system + product of activity ^ inactive system. Based on this concept Gray derived an equation which he called the "theory of autointoxication.'' \2Ka r .v = \/~~T — *t' (Equation 3) Here x, a, and t designated the same quantities as before, and K and b were con- stants. Equation 3, based primarily on the assumption of a homogeneous popula- tion, showed that the amount of fuel, .r, consumed in time t was directly proportional to the square root of t. Therefore, the total oxygen consumed should show a straight-line relationship to the square root of the age of the active suspension. Gray found that this relationship did not hold for sperm in sea water, for toward the end of active life the points fell below the expected values. He concluded that part of the active system was not only being inactivated, but also being irreversibly destroyed. Equation 3 was applied to the experimental figures of the present investigation. When the total oxygen consumption of the sperm in sea water is plotted against the square root of the time, the results confirm Gray's findings, for, toward the end of the run, the points fall below the expected values. (The variation from the straight- line relationship at the beginning of the run is due to the time requird for the cells to be completely dispersed in the medium.) For the sperm in sea water, therefore, an irreversible destruction of part of the active metabolic system is taking place to- 184 THRU HAYASHI ward the end of the run (Fig. 5). In contrast, the straight-line relationship is fol- lowed for a much longer time in seminal fluid in an equivalent run (Fig. 6). In some of the experiments, a departure from the straight-line relationship occurred in the seminal fluid, but in all cases, this departure occurred very much later from the beginning of active life. In the seminal fluid, therefore, the irreversible destruction of the active system is much delayed, as compared to the case of the sperm in sea water. The decay of activity of the sperm in seminal fluid is apparently due, at least for a great part, to a process of autointoxication. The difference between the metabolism of sperm in sea water and in seminal fluid apparently lies in the fact that some sort of metabolic system is kept intact for a longer time when the sperm cells are suspended in seminal fluid. The sperm and a factor in the seminal fluid, therefore, constitute a closed system. In sea water, r r o 111 O 10 111 o s _l © © b 10 VT t IN MINUTES FIGURE 5. Total oxygen consumption as a function of the square root of the time in minutes. Sperm in sea water. part of this closed system is destroyed quite early during the active life of the sperm. The above findings, supported by the knowledge that seminal fluid contains no reducing sugars (Hayashi, 1945). lends further support to the idea that the effect of seminal fluid on sperm respiration is not due to the replenishment of a store of en- cr^y, but rather to the maintenance of a metabolic system. ( )ther experiments (un- published data) measuring the R/J. of sperm cells both in sea water and seminal fluid show a characteristic carbohydrate metabolism for these cells, further support- ing the contention that the seminal fluid does not provide added nutrients for the sperm cells. It may be concluded that, although there is the possibility that sperm can be nourished experimentally, it is most important for future experimentation that the system, of which the sperm cells are only a part, be kept intact. Workers who have diluted sperm cells in sea water have diluted not only cells, but, also, system. DILUTION MEDIUM AND SPERM RESPIRATION 185 Buth the fertilizing power (Hayashi, 1945) and the respiratory activity of sperm are maintained by a common factor, the seminal fluid. As a first possibility, the role of the seminal fluid factor may be considered fundamental to the respiratory activity only, the fall in fertilizing power being a manifestation of the loss of activity of the sperm cells. In other words, the fall in fertilizing power is only a secondary effect in relation to the seminal fluid factor. The second possibility is that the seminal fluid plays a part directly in both fertilization and respiration. Lack of data does not permit the choice of these possibilities. However, the fact that a seminal fluid factor influences sperm respiration is implicit in both possibilities. A mechanism, tentatively proposed earlier as the wearing off of a protein sub- stance from the surface of the sperm cell, was adequate to explain the loss of ferti- lizing power of sperm in sea water. The same mechanism is also applicable to ihe £ Z O in 2 o * O J o 0 2 4 6 8 10 U. M »* Vt t IN MINUTES FIGURE 6. Total oxygen consumption as a function of the square root of the time in minutes. Sperm in seminal fluid. results of respiration studies, if the assumption be made, that the removal of each molecule of protein substance from the surface of the sperm cell releases to the sperm cell a finite amount of the total internal energy. The rate of sperm respiration thus would be controlled by the rate of removal of substance from the sperm surface, which is in turn controlled by other factors. The principal factor affecting the rate of removal of a sperm-surface-substance is the "mechanical crowding" factor studied by Gray (1928a), who found that sperm cells exhibited a burst of activity when diluted. This was confirmed in the present study (Figs. 1, 3, 4). According to the mechanism proposed, the rate of loss of sperm-surface-substance is inhibited when the cells are in a crowded condition, but when relatively far apart, the removal of surface-substance is enhanced. The removal of sperm substance from the cell surface presumably constitutes the irreversible destruction of part of the active system noted in Fig. 5. In seminal fluid, 186 TKKT HAYASHI however, the replacement of the surface-substance would de-lay tin- onset of this ir- reversible destruction (Fig. 6), the senescence of the sperm cells being conditioned by internal autointoxication, and perhaps also by depletion of the internal energy. It would seem that for the sperm cells in sea water, the destruction of part of the system, in addition to autointoxication and probable depletion of fuel, controls the fall in activity. It is understood that the mechanism as outlined is tentative at best, but it serves as an explanation of the facts so far known. The facts fit well with the earlier re- sults of the effect of seminal fluid on the fertilizing capacity of sperm. It may be concluded that possibly a common mechanism underlies the effect of seminal fluid on the fertilizing function and the respiration of sperm. The relationship between the surface-substance and the respiratory mechanism of the sperm cell was not in- vestigated. The action of the surface-substance while attached to the sperm cell is possibly enzymatic. Effect of egg water on sperm respiration Egg water does not stimulate the respiration of sperm cells but seems to condi- tion a sharp drop in the respiratory rate (Fig. 4). When egg water was added to sperm suspended in sea water and, also, to sperm suspended in seminal fluid, the effect was a sharp decrease in the respiratory rate. This decrease of metabolic ac- tivity upon dilution with egg water is in contrast to the effect of dilution with sea water and seminal fluid shown in Figure 3. The increase in respiratory rate brought about by dilution with sea water and seminal fluid shows that even at the dilution used, "mechanical crowding" was still apparently a factor inhibiting the respiration of spermatozoa. Dilution with egg water also relieved the "mechanical crowding," but the agglutinating effect of egg water apparently overcame the effect of dilution so that the respiratory activity decreased. The stimulation of metabolism by egg water noted by Gray (1928c) and Carter (1930) is not confirmed in these experiments. However, it may be recalled that Carter (1931) had found no stimulation of ripe sperm by egg water. In addition, the egg water of the present investigation was added, not to undiluted sperm, but to sperm that had already been activated by an initial dilution. Therefore, it is possible either that egg water did not affect the respiration of sperm at all, or that the differ- ence in the time of addition of egg water was the cause of the disparity of the results of Gray's experiments and the results of the present investigation. SUMMARY 1. Seminal fluid has the property of delaying the fall of respiratory activity of the- sperm after the original burst of activity upon dilution. 2. Gray's "theory of exponential decay" is not adequate to explain sperm meta- bolic activity, whereas the "theory of autointoxication" fits the activity of sperm cells suspended in seminal fluid. 3. The seminal fluid delays appreciably the onset of an irreversible destruction M|" part of the metabolic system. 4. Fertilization studies have led to the formulation of a tentative mechanism based on the adsorption of a protein substance and its removal from the surface of the sperm cell. DILUTION MEDIUM AND SPERM RESPIRATION 187 5. The proposed mechanism also explains adequately the results of the respiration experiments. Therefore, it is concluded that a seminal fluid factor, hy its action while on the surface of the sperm, influences both the fertilizing capacity and the respiratory rate of spermatozoa. 6. Egg water added to a sperm suspension after the original dilution, causes a sharp decrease in the respiratory rate. LITERATURE CITED CARTER, G. S., 1930. Thyroxine and the oxygen consumption of the spermatozoa of Echinus miliaris. Jour. Exp. BioL, 7: 41-48. CARTER, G. S., 1931. Iodine compounds and fertilization. II. The oxygen consumption of sus- pensions of sperm of Echinus esculentes and Echinus miliaris. Jour. E.vp. BioL, 8: 177-192. GRAY, J., 1928a. The effect of dilution on the activity of spermatozoa. Brit. Jour. E.vp. Biol., 5 : 337-344. GRAY, J., 1928b. The senescence of spermatozoa. Brit. Jour. E.rp. BioL, 5: 345-361. GRAY, I., 1928c. The effect of egg secretions on the activity of spermatozoa. Brit. Jour. E.rp. BioL, 5 : 362-365. GRAY, J., 1930. The senescence of spermatozoa, II. Brit. Jour. Exp. BioL, 8: 202-210. HAYASHI, TERU, 1945. Dilution medium and survival of the spermatozoa of Arbacia punctulata. I. Effect of the medium on fertilizing power. Biol. Bull., 89 : 162-179. LII.LIE, F. R., 1913. The behavior of the spermatozoa of Nereis and Arbacia with special refer- ence to egg extractives. Jour. E.rp. ZooL, 14: 515-574. THE EFFECTS OF LITHIUM CHLORIDE ON THE FERTILIZED EGGS OF NEREIS LIMBATA CATHERINE HENLEY Marine Biological Laboratory. U'oods Hole. Mass., and Department of Biology, The Johns Hopkins University INTRODUCTION The action of the lithium ion on the eggs of certain animals has long been known. Using a number of lithium salts, Herbst (1892) obtained abnormal embryos of two types. In the first group, the endoderm was found not inside the body, but outside; these forms are designated "exogastrulae." Larvae of the second group showed an apparent conversion of ectoderm into endoderm so that in extreme cases, the whole blastula wall was endodermized (Herbst, 1892, 1893, 1943). Herbst concluded that this was a specific and typical affect of lithium, a view shared by Gurwitsch (1895), Morgan (1902), and other early investigators. Studying the effects of 0.2 per cent to 1.0 per cent solutions of lithium chloride on frog and toad eggs, Gurwitsch found that in the resulting embryos, gastrulation was abnormal, but no exogastrulae were produced. Morgan (1902), using similar concentrations of lithium, obtained embryos in which the black hemisphere had sunk into the interior of the egg, but here again there was no definite indication of exogas- trulation. Experiments by Holtfreter (1931) showed that amphibian exogastrulae could be obtained if the blastulae were treated with a modified Ringer solution. Further experiments to ascertain the effects of lithium on the amphibian egg were carried out by Bellamy (1919) who observed a number of cases of fusion of the lateral sense organs of the larvae, as well as abnormalities in gastrulation. Tondury (1938) treated urodele embryos in early stages of gastrulation with LiCl and de- scribed a high percentage of head and foregut abnormalities in the larvae. Still an- other effect of lithium on urodele eggs was observed by Cohen (1938) who treated early gastrulae or late blastulae and observed cases in which the myotomes formed a continuous sheet across the midline. separating the nerve cord and notochord. He also obtained embryos in which exogastrulation had occurred, seemingly as a result of the treatment with lithium. The effects of LiCl on the eggs of the pond snail, Liinnaca stuuiuilis. were studied by Raven (1942) who was able to produce forms which he designates as exogas- trulae. He also describes a number of larvae in which varying abnormalities of the eyes were apparent. These experiments are of interest in connection with the present series because the egg of Limnaea, like that of Nereis, is an example of the so-called "mosaic" type of development. A wide variety of chemical and physical agents has since been found to produce echinoderm exogastrulae,' thus invalidating the theory of an ion specificity. These agents include liypotonic sea water, isotonic solutions of magnesium chloride and lithium chloride, and combinations of isotonic solutions of magnesium chloride, so- 188 EFFECTS OF LITHIUM ON NEREIS EGGS 189 clium chloride, potassium chloride, and calcium chloride as studied by Waterman (1932) ; solutions of nickel chloride and copper chloride (Waterman, 1937) ; lithium in low concentrations augmented by carbon monoxide (Runnstrom, 1935) ; and 97 per cent carbon monoxide with 3 per cent oxygen in the presence of light (Runn- strom, 1928a). MacArthur (1924) treated sea urchin eggs with calcium chloride, stale or diluted sea water, copper sulfate, mercuric chloride and potassium cyanide, and obtained exogastrulae. Using "auxin," glycogen and potassium chlorate, Moto- mura (1934) likewise obtained echinoderm exogastrulae. Most of the eggs previously tested with lithium have been those of echinoderms and amphibians, both of which are characterized by an invaginative type of gastrula- tion. It seemed, therefore, that an egg showing a strictly epibolic form of gastrula- tion should be tested, in an effort to ascertain if the same effect could be produced. The following experiments were performed to study the effects of lithium on the fertilized eggs of the annelid, Nereis limbata.1- METHODS Gametes of the heteronereis form of Nereis limbata were obtained at Woods Hole from June to September, during appropriate phases of the moon. Usually the eggs from one female were sufficient for an average experimental series ; they were insemi- nated according to the methods outlined by Just (1939) and were then washed with freshly drawn, filtered sea water. The fertilized eggs were allowed to remain undisturbed for a period of 75 minutes after insemination; at the end of this time (shortly before the first cleavage), they were transferred with as little sea water as possible to a series of 25 cc. stender dishes, to which the various solutions of LiCl were then added. Appropriate con- trols were kept in all series, these being cultured in filtered aerated sea water which was changed at daily intervals. A stock solution of 0.54 M LiCl in distilled water was used for all experimental mixtures ; such a solution is approximately isotonic in all dilutions with sea water. Experimental mixtures were designated according to the percentage of this stock solution used — e.g., a "15 per cent solution" indicates that 85 cc. of filtered aerated sea water were added to 15 cc. of the 0.54 M LiCl stock solution. About 15 cc. of the mixture were added to each stender dish; at intervals ranging from 15 minutes to 36 hours, the solutions were decanted carefully and the eggs washed in three or four changes of sea water. LiCl-sea water mixtures ranging from 2 per cent to 100 per cent were tested for varying periods of time. No attempt was made to control the room temperature, which varied from 20° C. to 28° C. However, all cultures were kept in moist chambers which stood in running sea water ; the temperature of this sea water averaged about 20° C., and did not vary more than one or two degrees at any time during the entire series of experiments. Less extensive test series were also conducted for purposes of comparison, using the eggs of Nereis from which the vitelline membrane had been removed, according to the method described by Costello (1945a). 1 The author wishes to express sincere appreciation to Prof. D. P. Costello for advice and encouragement throughout the course of the investigation, to Prof. Viktor Hamburger and Dr. Charles Packard for use of laboratory space at the Marine Biological Laboratory during the sum- mers of 1944 and 1945, to Prof. B. H. Willier for his kindly interest, and to Mr. John S. Spurbeck for valuable assistance with the drawings. 190 CATHERINE HENLEY RESULTS The most striking characteristic of the experimental larvae is the complete ab- sence of exogastrulation, or of any definite evidence of true vegetalization. Ab- normal larvae occurred to some extent in almost all the cultures ; these abnormalities included the absence of an apical tuft, lengthening of the prototrochal cilia (Fig. 2A), absence of these cilia in varying degrees (Fig. 2). absence or abnormality of the eye spots (Fig. 3), deficiencies in the number of prototrochal cells (Fig. 3), atypical seta sacs with derangements in the number and position of these organs in older larvae (Fig. 4), and abnormalities in the number and position of the oil drop- lets (Figs. 2C, 3D, 4B). Most of the experimental larvae had abnormalities of the anal and prototrochal pigments; these abnormalities included both the absence of pigment and an abnormal concentration of pigment granules in various regions of the larvae (Figs. 2, 3, 4, 5). The specific occurrence of these anomalies was roughly proportional to the severity of treatment, being far more marked in cultures which PR. C. PR. P. A.P. FIGURE 1. A normal Nereis trochophore at about 30 hours; polar view. Pr. C. = Proto- trochal cilia ; Pr. P. = Prototrochal pigment ; Pr. = Prototrochal cell ; O. D. = Oil droplet ; A. P. = Anal pigment ; E. = Eye spot ; 3A, 3B, 3C, 4D = Entomeres. This and all subsequent text-figures are semi-diagrammatic, based on camera lucida drawings of the living larvae. were exposed to higher concentrations for sublethal periods. No evidence of twin- ning was observed. Complete data describing the results of treatment with the varying concentrations for varying periods are contained in Table I. It is evident that concentrations of 2 per cent to 4 per cent acting for periods up to 12 hours produce larvae which are almost completely normal ; similar trochophores result when 5 per cent to 10 per cent solutions are applied for one to 5 hours. However, when allowed to act for periods of 5 to 24 hours, the same concentrations affect the eggs so that the resulting larvae show marked deficiencies in the number of prototrochal cells, and are usually com- pletely devoid of the normal anal and prototrochal pigment. Abnormalities such as these occur quite commonly in all series ; some extreme examples are shown in Figure 2. For purposes of comparison, a normal trochophore is shown in Figure 1. Accurate quantitative observations of the exact deficiency in the number of proto- trochal cells were impossible in most cases. Twelve per cent solutions acting for periods of I1/} to 5 hours result in trocho- phorcs which are essentially normal, but if allowed to remain on the eggs for 5 to 24 hours, bring about pigment and prototrochal anomalies of the type described above. EFFECTS OF LITHIUM ON NEREIS EGGS 191 TABLE 1 Effects of lithium treatment of Nereis eggs Duration of Percentage2 treatment Appearance of 26-30 hour larvae in hours 2% \Y^ Normal 5 Normal 8 Mostly normal; some with abnormalities in number and position of oil droplets 23 Mostly normal; some with abnormalities in number and position of oil droplets 3 8 10 24 Many with incomplete prototroch and with abnormalities in ber and position of oil droplets Many with incomplete prototroch and with abnormalities in ber and position of oil droplets Abnormal distribution of oil droplets, otherwise normal num- nu rn- 41 IX * / 2 5 8 12 Normal Mostly normal Mostly normal Normal in form; somewhat sluggish in movements 5 14 16 18 20 22 26 28 30 32 36 Normal Mostly normal Mostly normal Some prototrochal and pigment abnormalities Pigment and prototrochal abnormalities Considerable variation in amount of pigment present Variations in amount of pigment present More marked pigment abnormalities than after 28-hour ment Pigment and prototrochal abnormalities Dead treat- 6 1M 5 8 12 23 Normal Normal, but sluggish in movements Mostly normal Normal, but sluggish in movements Quite abnormal; prototrochal and pigment abnormalities 7 1 2 5 24 Normal A very few cases with oil droplet deficiencies Some cases with oil droplet deficiencies Prototrochal cell deficiencies; sluggish in movements; pigment often absent 8 1 Yi Normal 5 Some cases with prototrochal and pigment abnormalities 8 More marked pigment abnormalities 12 Severe prototrochal deficiencies; no pigment 23 Severe prototrochal deficiencies; no pigment. Oil droplets ab- normal in number and position. Many larvae without apical tuft 2 Percentages in the table refer to percentages of 0.54 M stock solution of lithium chloride with sea water : 2 per cent = 2 cc. 0.54 M LiCl + 98 cc. sea water. 1')2 CATHERINE HENLEY TABLE I — Continued Duration of Percentage2 treatment in hours Appearance of 26-30 hour larvae 10 1 4 8 12 14 16 18 20 24 26 28 36 Some localized deposits of pigment; some pigment present in most cases Some pigmented, some unpigmented. A few cases of abnormality in amount and position of anal pigment Pigment abnormalities; prototrochal cells seem fairly complete Pigment abnormalities; prototroch fairly normal No pigment; marked defects in number of prototrochal cells No pigment; marked defects in number of prototrochal cells No pigment, no cilia, no apical tuft No pigment, no apical tuft. Abnormalities in number and posi- tion of oil droplets No pigment, no cilia, no apical tuft. Abnormalities in number and position of oil droplets No pigment; marked prototrochal deficiencies. No apical tuft; oil droplets very abnormal in number and position No pigment; prototrochal defects marked. Oil droplets ab- normal in number and position Dead 12 U-2 5 8 12 23 Fairly normal; some abnormalities in number of prototrochal cells and oil droplets Many deficiencies of pigment and prototroch; oil droplets ab- normal in number and position Pigment deficiencies; prototroch and oil droplets abnormal in number and position Some abnormalities in distribution of oil droplets; no pigment. Prototrochal deficiencies; no apical tuft Dead 14 \Y2 5 8 23 Mostly normal; some pigment deficiencies No pigment; marked prototrochal deficiencies No pigment; prototrochal deficiencies; no apical tuft Dead 15 1 Mostly unpigmented; prototrochal deficiencies. Anal pigment present in many cases. Oil droplets abnormal in number and position 2 Unpigmented; prototrochal deficiencies; anal pigment present in some cases. Oil droplet abnormalities 4 More marked pigment defects; severe prototrochal deficiencies. Some few cases of endodermal extrusion 6 Highest incidence of endodermal extrusion: 2%-7% 8 Somewhat fewer cases of extrusion than in 6-hour culture 10 Some few cases of endoderm extrusion 1 1 A very few cases of endodermal extrusion 14 No pigment; marked prototrochal defects. No apical tut t . 16 No pigment; prototrochal defects; no apical tuft 20 No cilia, no pigment, no apical tuft. -Severe prototrochal de- ficiencies 24 No pigment, no cilia, no apical tuft EFFECTS OF LITHIUM ON NEREIS EGGS 193 TABLE I — Continued Duration of Percentage* treatment in hours Appearance of 26-30 hour larvae 17 1 2 3 6 8 10 14 Anal pigment present; prototrochal pigment absent in many cases. Considerable number of oil droplet abnormalities No pigment; oil droplet abnormalities; prototrochal deficiencies; no apical tuft No pigment; oil droplet abnormalities in number and position. Prototrochal deficiencies, no apical tuft Marked oil droplet deficiencies and abnormalities in position; no pigment. No apical tuft; prototrochal cell deficiencies Many amorphous forms; no cilia; abnormalities in number and position of oil droplets No cilia, no pigment, no apical tuft Dead 20 1 2 4 8 12 14 Normal Pigment abnormalities; some localization of pigment No pigment; prototrochal cell deficiencies; apical tuft apparently missing; oil droplets abnormal in number and position Marked oil droplet deficiencies; severe prototrochal defects Oil droplet abnormalities; marked prototrochal deficiencies Dead 25 1 2 4 8 12 14 Prototrochal cell deficiencies; some pigment defects No pigment; prototrochal cell deficiencies; abnormalities in num- ber and position of oil droplets Severe prototrochal deficiencies; no apical tuft. No anal pigment Prototrochal deficiencies; pigment abnormal. Oil droplet ab- normalities in number and position Marked prototrochal cell deficiencies, no pigment, no apical tuft Dead 30 2 4 No pigment; marked prototrochal deficiencies. No apical tuft in most cases Dead 35 2 No pigment; marked prototrochal deficiencies. No apical tuft 40-90 1 Dead 50 K No pigment; prototrochal deficiencies. Apical tuft usually absent 100 y* Prototrochal and pigment deficiencies Yz Dead Treatment of the eggs with 15 per cent solutions for one to 4 hours produces pig- ment and prototrochal abnormalities, but after 6, 8, and 10 hours of treatment, the resulting larvae are marked by peculiar endodermal derangements which were at first thought to be the result of exogastrulation. However, closer examination of these cases revealed that the extruded cytoplasm was non-cellular insofar as could be de- termined, although it had the characteristic "glassy" appearance of endodermal cyto- plasm; usually these larvae were marked by the complete absence of cilia (Fig. 5). 194 CATHERINE HENLEY Frequently cases were observed in which the endoderm extended up under the sur- mounting cap of ectoderm. The cleavage pattern of the experimental eggs of this series showed no abnormality through the fourth cleavage. Repeated experiments with this treatment and with exposure of the eggs to concentrations of 12 per cent and 17 per cent for varying periods of time definitely established the fact that the abnormalities occur only after the treatment described above. It does not seem likely that such an extremely narrow range of dosage would be necessary for the production of true exogastrulae, inasmuch as these forms are produced in sea urchin eggs after a rather wide variety of treatments. B FIGURE 2. Typical 26-30 hour abnormal trochophores, showing anomalies in ciliary hand, pigmentation and eye spots. Anal pigment and prototrochal pigment art.- absent in all cases ; Figure 2A shows lengthening of the prototrochal cilia. Exposure of the eggs to a 17 per cent solution resulted in generally abnormal trochophores after a treatment of one hour, and in more pronounced pigment and prototrochal abnormalities if treated for longer periods, up to the lethal point at 14 hours. No larvae with the endodermal derangements described above were noted in any of the cultures in this series. The effects of a 20 per cent solution for com- parable periods of time approximate those described for a 17 per cent treatment; a concentration of 25 per cent produces pigment and prototrochal abnormalities alter treatment for 1, 2, 4 and 8 hours. These defects also result from treatment for 2 hours with a 30 per cent solution. Concentrations of 40 per cent to (HJ per cent EFFECTS OF LITHIUM ON NEREIS EGGS 195 are lethal after treatment for one hour, and a 100 per cent solution has the same effect after y2 hour; the 100 per cent concentration produces pigment and proto- trochal defects if applied for 15 minutes, so that its range of dosage is apparently very narrow. The reduction of the prototroch seems to increase quantitatively with an increase in the concentration of lithium and the duration of application. Denuded Nereis eggs proved to be extremely sensitive to the effects of lithium, even in low concentrations for short periods of time. In all series, the controls showed considerable abnormality in shape and in the distribution of pigment, so that B C D FIGURE 3. Abnormalities involving pigment, eye spots, cilia, and oil droplets after various treatments with lithium. Figure 3A is almost normal except for the heavy concentrations of anal and prototrochal pigment ; Figure 3B indicates an abnormality in the number and position of the eye spots and in the size of these structures. it is difficult to come to any definite decision as to the specific effects brought about by the lithium. However, in the cases in which membrane removal was not com- plete, the larvae survived fairly well, and in none of these cases were any indications of exogastrulation observed. After treatment for 2 hours with 15 per cent LiCl, some of the completely denuded larvae seemed to show signs of exogastrulation, al- though no conclusions can be drawn from this because of the small number of cases. Treatment of the denuded eggs with 5 per cent solutions for periods longer than 4 hours killed the eggs in late stages of cleavage, and no definite effects could be noted in larvae surviving in cultures which had been exposed for periods of 1 and 2 hours. 196 CATHERINE HENLEY DISCUSSION From the foregoing results, it is apparent that the main effect of the lithium is to produce a reduction in the number of prototrochal cells, and in the quantity of anal and prototrochal pigments, together with effects on the eye spots and cilia. The absence of the apical tuft in most cases of prolonged or concentrated treatment indi- cates that the la-Id cells may be affected directly or indirectly, since it is known from the observations of Wilson (1892) that this quartet is associated with the formation B SET. FIGURE 4. Marked abnormalities in setae of older larvae (ca. 53 hours). an atypical number of oil droplets. Set. = Setae. Figure 4B also shows of the apical tuft. It is possible that the abnormalities observed in the seta sacs of older larvae (Fig. 4) may be due to the early action of the ion on the material des- tined to be incorporated into the 2d cell. The absence of any true cases of exogastrulation in these experiments is quite striking. As was noted above, most of the forms in which this abnormality has been observed are marked by a type of gastrulation in which invagination plays a main EFFECTS OF LITHIUM ON NEREIS EGGS 197 role. In the amphibian, formation of the endoderm is accomplished by a complex series of cell movements, as a result of which a directed and organized migration of cells brings about imagination of most of the vegetal hemisphere at the region of the blastopore. Meanwhile, the animal cells grow down to cover the vegetal region, thus forming the ectoderm of the embryo. The yolk material itself is invaginated in these forms and comes to be enclosed by the endoderm. Gastrulation in the echino- derms is likewise thought by Herbst (1893) to be brought about by an invaginative process. The strictly epibolic form of gastrulation exhibited by the egg of Nereis, on the other hand, is accomplished by a downgrowth of the ectomeres, so that even- B ECT.-U-— ..I — KND. FIGURE 5. Abnormal trochophores resulting from treatment of eggs with 15 per cent LiCl for 6, 8 and 10 hours. Figure 5A shows no particular endodermal derangements, but has only one small tuft of cilia and is quite atypical in shape. Figures 5B and 5C show examples of "endodermal extrusions." Cell boundaries were not visible except as indicated. Ect. := Ecto- derm ; End. = Endoderm. tually they cover the four entomeres. Such a movement seems to be associated with the presence of a "cellular affinity" as postulated by Costello (1940, 1945a) in connection with his experiments on the development of isolated blastomeres of this egg. Since no exogastrulation occurred, it may be assumed that this dynamic asso- ciation of ectomeres and entomeres is not radically disturbed. However, the abnor- malities in the number and position of the oil droplets in the four entomeres indicate that the lithium may exert some effect on the process of cytoplasmic segregation (Costello, 1945b) preceding the formation of the 3A, 3B, 3C and 4D cells. CATHERINE HENLEY No distinct line of demarcation can be drawn between tbe direct and tbe indirect effects of lithium in these- experiments, since the ion is brought into contact with the egg before cleavage has occurred, and, in the case of the longer exposures, may re- main until considerably after the completion of gastrulation. Runnstrom (1928b), observing lithium-treated sea urchin eggs under dark-field illumination, presented evidence that the element actually penetrates the cells. Spek (1918) maintained that the action of the lithium ion was brought about through its production of a pre- cipitation and swelling effect on the surface of the vegetative cells. Thus, the exact mode of action remains obscure, but it appears fairly clear in the case of the egg of Nereis that exogastrulation is not produced, at least not in cases where the vitelline membrane is present. The effects produced on the pigment of the trochophores seem to be at random, since the anal pigment may occur without the presence of the prototrochal pigment, and vice versa — or both may be present in varying degrees. This action possibly is related to the orientation of the eggs with respect to the bottom of the dish, or to each other; although no particular effort was made to keep the eggs suspended in the solution, the jelly serves to support them during the early stages of development, thus allowing relatively free access of the lithium to all surfaces. SUMMARY 1. The fertilized eggs of Nereis liinbata were treated with mixtures of sea water and a 0.54 M stock solution of LiCl, ranging from 2 per cent to 100 per cent for periods of 15 minutes to 36 hours. Treatment was begun 75 minutes after insemina- tion of the eggs, shortly before the appearance of the first cleavage. 2. No exogastrulae were observed. There were a few cases of marked abnor- malities in the endodermal components of the trochophores within a very narrow range of treatment (15 per cent for 6, 8 and 10 hours). 3. The main abnormalities observed in the experimental larvae were : Absence of the apical tuft, lengthening of the prototrochal cilia, absence of these cilia in varying degrees, abnormalities in the anal and prototrochal pigments, absence or abnormality of the eye spots, deficiencies in the number of prototrochal cells, atypical seta sacs, abnormalities in the number and position of oil droplets. The degree of abnormality in these cases seemed to be roughly proportional to the severity of treatment. LITERATURE CITED BELLAMY, A. W., 1919. Differential susceptibility as a basis for modification and control of early development in the frog. Biol. Bull, 37: 312-361. ( OHKN, A., 1938. Myotome fusion in the embryo of Amblystoma punctatum after treatment with lithium and other agents. Jour. /:>/>. Zoul., 79 : 461-472. CCISTI VI.LO, D. P., 1940. Gastrulation of isolated blastomeres of Nereis eggs. Anal. Rcc. (Suppl.), 78: 132-133. ( "OSTU.I.O, D. P., 1945a. Experimental studies of germinal localization in Nereis. I. The devel- opment of isolated blastomeres. Jour. /f.r/>. Zoul., 100: 19-66. CUSTKLLO, D. P., 1945b. Segregation of ooplasmic constituents. Jour. I'.lislict Mitchell Scientific Soc., 61 : 277-289. GUKWITSCH, A., 1895. Ueber die Einwirkung des Lithionchlorids auf die Entwicklung der I'Yusrh- und Krnteneier (R. fusca und Bufo vulg.). Anat. Anz., 11B: 65-70. llikHST, C., 1892. Experimentelle Untersuchung iiber den Kinrluss der veranderten chemischen Zusammensetzung des umgebenden Mediums auf die Entwicklung der Thiere. 1 Theil. \ risuche an Seeigeleiern. /.cil. ll'iss. /.on!., 55: 446-518. EFFECTS OF LITHIUM ON NEREIS EGGS 199 HERBST, C., 1893. Experimentelle Untersuchung iiber den Einfluss der veranderten chemischen Zusammensetzung des umgebenden Mediums auf die Entwicklung der Thiere. 2 Theil. Weiteres iiber die morphologische Wirkung der Lithiumsalze und ihre theoretische Bedeutung. Mitt. Zool. Stat. Naples, 11: 136-220. HERBST, C., 1943. Die Bedeutung der Salzversuche fiir die Frage nach der Wirkungsart der Gene. Zusammenschau einer Hydrationsbzw. Mikrolaboratorientheorie der Genwirkung. Arch. f. Entw. der Or,/., 142: 319-378. HOLTKRETER, J. 1931. liber die Aufzucht isolierter Teile des Amphibienkeimes. II. Ziichtung von Keimen und Keimteilen in Salzlosung. Arch. f. Entw. der Org., 124: 404-466. JUST, E. E., 1939. Basic methods for experiments on egg of marine animals. Philadelphia: P. Blakiston's Sons and Co., Inc. 89 pp. MACARTHUR, J. W., 1924. An experimental study and a physiological interpretation of exogas- trulation and related modifications in echinoderm embryos. Biol. Bull.. 46: 60-87. MORGAN, T. H., 1902. The relation between normal and abnormal development of the embryo of the frog, as determined by tine effect of lithium chloride in solution. Arch. f. En tit'. der Org.. 16: 691-712. MOTOMURA, I., 1934. Uber die Exogastrulabildung beim Seeigelkeim durch Auxin, Glycogen, und KC1O3. Sci. Rep. Tohoku Imp. Unir. Scries 4, Biology, 9: 123-130. RAVEN, C. P., 1942. The influence of lithium upon the development of the pond snail, Limnaea stagnalis L. Proc. Nederl. Akad. van Wctcns., 45: 856-860. RUNNSTROM, J., 1928a. Zur experimentellen Analyse der Wirkung des Lithiums auf den Seeigel- keim. Acta Zool., 9 (3) : 365-424. RUNNSTROM, J., 1928b. Plasmabau und Determination bei dem Ei von Paracentrotus lividus Lk. Arch. f. Entw. der Org., 113: 556-581. RUNNSTROM, J., 1935. An analysis of the action of lithium on sea urchin development. Biol. Bull.. 68: 378-384. SPEK, J., 1918. Differenzen im Quellungszustand der Plasmakolloide als eine Ursache der Gas- trulainvagination, sowie der Einstulpungen und Faltungen von Zellplatten iiberhaupt. Kolloidchcm. Beih., 9 : 259-399. TONDURY, G., 1938. Weitere Beitrage zur Frage der Kopfentwicklung bei Urodelen. II. Erzeugung von Mikrokephalie durch Einwirkung von Lithiumchlorid auf die Gastrula von Triton alpestris. Arch. f. Entiv. der Org., 137 : 510-539. WATERMAN, A. J., 1932. Effects of isotonic salt solutions upon the development of the blastula of the sea urchin, Paracentrotus lividus. Arch, de Biol., 43: 471-519. WATERMAN, A. J.. 1937. Effects of salts of heavy metals on development of the sea urchin, Arbacia punctulata. Biol. Bull.. 73: 401-420. WILSON, E. B., 1892. The cell lineage of Nereis. A contribution to the cytogeny of the annelid body. Jour. Morph., 6 : 361-480. STUDIES <)X CILIATES OF THE FAMILY ANCISTROCOMIDAE CIIATTON AND LWOFF (ORDER HOLOTRICHA, SUBORDER THIGMOTRICHA). II. HYPOCOMIDES MYTILI CHATTON AND LWOFF, HYPOCOMIDES BOTULAE SP. NOV., HYPOCOMIDES PARVA SP. NOV., HYPOCOMIDES KELLIAE SP. NOV., AND INSIGNICOMA VENUSTA GEN. NOV., SP. NOV. EUGENE N. KOZLOFF Lewis and Clark College, Portland. Orajtni INTRODUCTION The genus Hypocomides was proposed by Chatton and Lwoff ( 1922a) to in- clude two species: Hypocomides invtili, from My til its cditlis L., and H. modiolariac, from Modiolaria inaniwnita (Forbes). These authors gave no formal diagnosis of the genus Hypocomides, however, and their incomplete descriptions of the two spe- cies are not supplemented by illustrations. They did not designate either species as the genotype. In 1926 Chatton and Lwoff published a preliminary diagnosis, again without illustrations, of a third species of Hypocomides, H. zyrpliacae. from Zirjaea ens pat a (L.). It was largely on the basis of their description of H. zyrphacac that Raabe (1938) was led to suppose that one of the ciliates which he found parasitizing the gills of Mytilns editlis was H. inytili. He could not identify on this ciliate, how- ever, a structure to which Chatton and Lwoff referred as the "vestige de frange adorale" and which according to them is in H. mytili "constitue seulement par une dizaine de grands cils." This "vestige de frange adorale" is supposed to be better developed in H. modiolariae than in H. inytili. In H. zyrphaeae it is represented, according to Chatton and Lwoff, by "une touffe de cils." The position of the "ves- tige de frange adorale" in relation to other ciliary structures in these three species of Hypocomides is entirely unclear. On a ciliate which I have studied from Mytilus cditlis from San Francisco Bay and which conforms in most respects to the description given by Raabe of the form considered by him to be H. inytili, I have been unable to detect such a "vestige de frange adorale." The brief notes on the morphology of H. mytili to be found in the papers of Chatton and Lwoff in which mention is made of this species (1922a,l(^22b, l'<24) are not entirely consistent, and it is altogether possible that, as Raabe has pointed out, these authors used the term "vestige de frange adorale" to indicate only a short segment of the distal portion of one of the longer ciliary rows on the right side of the body, which Raabe suggested may be homologous with the two rows bordering the peristotnal groove of ciliates of the family Ancistrumidae. It is quite evident from the brief description of //. zyrphaeae, however, that the "vestige de frange adorale" of this species is entirely separate from the two long 'rows on the right. Perhaps //. inytili and H. modiolariae are not actually congeneric with H. syrphaeae, but this remains to be seen. At any rate, unless it can be established 200 CILIATES OF THE FAMILY ANCISTROCOMIDAE. II 201 with certainty that the ciliate described by Chatton and Lwoff as Hypocomides my- tili is not identical with the species thought by Rnabe to be H. inytili, it seems best to consider the form studied by Raabe to be H. inytili and to refer related forms to the same genus. In the present paper I will give a description of the ciliate from Mytilus edulis which I consider to be H. mytili, and will add three new species to the genus Hypo- courides: H. botulae sp. nov. and H. parua sp. nov., from the gills and palps of the rock-boring pelecypod, Botula californiensis (Philippi), and H. kelliae sp. nov., from Kcllia laperonsii Deshayes, a small nestling clam which is frequently encoun- tered in the excavations made in rocks by other molluscs. Another very interesting ancistrocomid ciliate from Botula californiensis will be described herein as Insiyni- coina vainsta gen. nov., sp. nov. HYPOCOMIDES MYTILI CHATTON AND LWOFF (Fig. 1; Plate I, Fig. 1) The body is elongated and somewhat flattened dorso-ventrally. The ciliary sys- tem, to be described presently, is disposed for the most part on the shallow concavity occupying the anterior three-fifths of the ventral surface ; the dorsal surface and that part of the ventral surface posterior to the ciliary area are convex. The anterior portion of the left margin is usually not quite so rounded as the right margin, and appears typically to be weakly indented. The body is widest near the middle and rounded posteriorly. Thirty living individuals taken at random ranged in length from 34 /A to 48^., in width from 16 /u, to 22 /A, and in thickness from 13 /A to 18^, averaging about 40 /A by 18 p by 14.5 /u. The anterior end of the body is provided with a short contractile tentacle which enables the ciliate to attach itself to the epithelial cells of the gills and palps of the host and to suck out their contents. This tentacle is continuous with an internal tubular canal which can usually be traced in fixed specimens stained with iron hema- toxylin for about one-half the length of the body. The canal nearly always appears to be widest in its anterior portion and to be directed obliquely toward the right side of the body as it extends posteriorly. The cilia of H. mytili are about 9 /A in length and are markedly thigmotactic, espe- cially nedr the anterior end of the body. The ciliary system consists of three sepa- rate complexes. The central complex, occupying the middle and right portions of the ventral anterior concavity, consists of seven rows, the one nearest the right being the shortest (one-third to two-fifths the length of the body), the other rows becoming progressively longer toward the left. The sixth and seventh rows are usually ap- proximately one-half the length of the body; in some specimens the seventh row is appreciably shorter than the sixth. To the right of the central complex are two long rows, each about one-half the length of the body. Both these rows originate on the dorsal surface close to the left margin a short distance behind the level of origin of the rows of the central complex and curve vent rally as they extend poste- riorly. To the left of the central complex is a series of eight rows which usually are more closely-set than those making up the central complex. The innermost row, which originates on the left margin of the body near the base of the suctorial tentacle, is the shortest, and terminates at a point about one-third the distance from the anterior end of the body to the posterior end. The remaining rows become pro- 202 K I 'GENE N. KOZLOFF gressively longer and originate progressively more dorsally and posteriorly, curving ventrally as they extend l)ack\vard. The outer row of this complex is usually the longest and terminates at a point nearly opposite the point of termination of the outer of the two rows constituting the complex on the right. According to Raahe. the three ciliary complexes of H. uiytili are much more dis- tinctly separated than I have observed them to be. Raahe also stated that variations in the number of ciliary rows in the central and left complexes are correlated with two well-differentiated size races: form minor (17 ^ to 26 p. in length) and form major (26 p. to 36 /A in length). I have noted no tendency for the ciliates I consider to belong to this species to be segregated into distinct size races and have seen no examples of H. mytili which were in life as small as those assigned to the form minor by Raabe. Although I have observed few deviations from the typical num- ber of ciliary rows, it may be of interest to record here the fact that frequently some FIGURE 1. Hypocomides mytili Chatton and Lwoff. Distribution of ciliary rows, somewhat diagrammatic.1 A, dorsal aspect ; B, ventral aspect. of the rows of the left complex, particularly the first and second rows, do not stain as well with hematoxylin or impregnate as well with activated silver albumose (pro- targol) as do the rows of the central and right complexes, and hence may easily escape detection. The cytoplasm is colorless and contains numerous lipoid droplets in addition to food inclusions. Several larger food vacuoles are sometimes observed in the poste- rior part of the body. The contractile vacuole is centrally located - and opens to the exterior on the ventral surface. I have observed no permanent opening in the pellicle. The macronucleus is usually ovoid, sometimes sausage-shaped ; more rarely it is spherical, although Raabe considered the macronucleus of H. mytili to be typically All text figures illustrating this paper are based on camera lucida drawings of individuals fixed in Schaudinn's fluid and impregnated with activated protein silver (protargol). - Raabe stated that the contractile vacuole of H . mytili is to be found in a vacuolated area in the posterior part of the body behind the macronucleus. Perhaps his observation was based on specimens which were undergoing degenerative vacuolization and in which the contractile vacuoK- was not conspicuous. In all normal individuals which 1 have examined its position was central. CILIATES OF THE FAMILY ANCISTROCOMIDAE. II 203 spherical. It is situated dorsally in the posterior half of the body with its longi- tudinal axis placed obliquely to the longitudinal axis of the body. In fixed and stained preparations the chromatin appears to be aggregated into a dense reticulum enclosing vacuole-like clear spaces of varying size. In fifteen individuals fixed in Schaudinn's fluid and stained by the Feulgen nuclear reaction the macronucleus ranged in length from 9 /u to 13.2 p. and in width from 4.4 /JL to 6.9 /z. The spherical micronucleus is ordinarily situated near the middle of the body anterior to the macronucleus, although sometimes it is seen to lie to one side of the macronucleus. The chromatin appears in most fixed and stained specimens to be homogeneous, although in some it is aggregated into vague peripheral granules or strands. In fifteen individuals fixed in Schaudinn's fluid and stained by the Feulgen reaction the diameter of the micronucleus ranged from 2.7 //. to 4 p. I found Hypocomides my till to be present in large numbers on the gills and palps of about 80 per cent of the specimens of Mytilus cdulis which I examined from vari- ous localities in San Francisco Bay. It is sometimes the only ciliate infesting the mussels, but more commonly it is associated with Crebricoma carinata (Raabe) and Ancistntina mytili (Ouennerstedt). Hypocomides mytili Chatton and Lwoff Diagnosis: Length 34/x-48/x. (according to Raabe 17 /x to 36 /*), average about 40 fi ; width 16 p.-22 p., average about 18 ^ ; thickness 13 /x-18 /JL, average about 14.5 /x. The central ciliary complex is composed of seven rows (according to Raabe seven or eight rows) which are one-third to one-half the length of the body, becoming pro- gressively longer toward the left side ; the right complex consists of two rows, each about one-half the length of the body ; the left complex consists of eight closely-set rows (according to Raabe five [ ?] or six rows) one-third to one-half the length of the body. The contractile vacuole is central and opens to the exterior on the ventral surface. The macronucleus is typically ovoid or sausage-shaped, rarely spherical. The micronucleus is spherical. Parasitic on the epithelium of the gills and palps of Mytilus cdulis L. (Roscofif [Chatton and Lwoff] ; Hel [Raabe] ; San Francisco Bay, California). HYPOCOMIDES BOTULAE SP. NOV. (Fig. 2; Plate 1, Fig. 2) The body is elongated, narrowed anteriorly, and somewhat flattened dorso- ventrally. The ciliary system, to be described presently, is disposed for the most part on the shallow concavity occupying the anterior one-half of the ventral surface ; the dorsal surface and that part of the ventral surface posterior to the ciliary area are convex. The anterior portion of the left margin is usually less rounded than the right margin and appears typically to be weakly indented. The body is widest near the middle and rounded posteriorly. Twenty living individuals taken at ran- dom ranged in length from 31 ^ to 39 /*, in width from 14 /JL to 17/x, and in thickness from 12 p. to 14 /^, averaging about 33 /*. by 15 /* by 13 /JL. The anterior end is provided with a contractile suctorial tentacle continuous with an internal tubular canal which can usually be traced in fixed specimens stained with iron hematoxylin for a distance equal to about three-fifths the length of the body. 204 EUGENE N. KOZLOH In most individuals the canal appears to be directed obliquely toward the right side of the body. The cilia are about 8 /* to 9 ^ in length and are markedly thigmotactic, particu- larly near the base of the suctorial tentacle. The ciliary system consists of three M'parate complexes. The central complex, occupying the larger part of the ventral anterior concavity, consists of eleven rows about one-half the length of the body which become progressively longer toward the left side. To the right of the cen- tral complex are two longer row's which originate on the dorsal surface close to the right margin a little behind the level of origin of the rows of the central complex and curve downward as they extend posteriorly. Each of the two rows is about three-fifths the length of the body. To the left of the central complex is a series of eleven rather closely-set rows which originate on the left lateral margin and the dorsal surface on the left side and curve ventrally as they extend posteriorly. These I-H.URE 2. Hypocoinidcs hotulac sp. nov. Distribution of ciliary rows, somewhat diagrammatic. A, dorsal aspect ; B, ventral aspect. rows are about one-half the length of the body and originate and terminate progres- sively more posteriorly. The cytoplasm is colorless and contains numerous lipoid droplets in addition to food inclusions. I have not seen any large food vacuoles in this species. The con- tractile vacuole, situated near the middle of the body, opens to the exterior on the ventral surface. The macronucleus is ovoid to sausage-shaped and is situated in the posterior half of the body. Its longitudinal axis is usually placed obliquely to the longitudinal axis of the body. In fixed and stained preparations the chromatin is aggregated into a dense reticulum. Vacuole-like clear spaces of varying size are sometimes apparent near the periphery. In ten individuals fixed in Schaudinn's fluid and stained by the Feulgen reaction the macronucleus ranged in length from 9/x to 13 /x and in width from 4.3 //, to 7 /x. The micronucleus is spherical and is situated dorsally near the middle of the body. The chromatin appears to be homogeneous in fixed specimens. In ten individuals CILIATES OF THE FAMILY ANCISTROCOMIDAE. II 205 fixed in Schaudinn's fluid and stained by the Feulgen reaction the diameter of the micronucleus ranged from 2.4 p. to 3.2 p.. Hypocomidcs botulac was present in small numbers on the gills and palps of twelve of the thirty-four specimens of Botula californiensis which I examined from localities near Moss Beach, California. It is sometimes found in association with Hypocomidcs parva and Insignicoma vcnusta. No specimens of Botula falcata (Gould) which I examined from the same localities were parasitized by these or any other ciliates. Hypocomidcs botulac sp. nov. Diagnosis : Length 31 /t-39 p, average about 33 p. ; width 14 /x-17 //.. average about 15 p.; thickness 12/x-H/u,, average about 13 p.. The central ciliary complex consists of eleven rows each about one-half the length of the body; the right complex is com- posed of two rows about three-fifths the length of the body; the left complex con- sists of eleven closely-set rows about one-half the length of the body. The macro- nucleus is ovoid to sausage-shaped. The micronucleus is spherical. Parasitic on the gills and palps of Botula californiensis (Philippi) (Moss Beach, California). Syntypes are in the collection of the author. HYPOCOMIDES PARVA SP. NOV. (Fig. 3; Plate I, Fig. 3) The body is elongated, narrowed anteriorly, and somewhat flattened dorso- ventrally. The ciliary system, to be described presently, is disposed for the most part on the shallow concavity occupying the anterior two-fifths of the ventral surface ; the dorsal surface and that part of the ventral surface posterior to the ciliary area are convex. The anterior half of the left margin is usually not so rounded as the right margin, and appears typically to be nearly straight or weakly indented. The body is widest near the middle and rounded posteriorly. Twenty-five living indi- viduals taken at random ranged in length from 21 //, to 29 p, in width from 10 p, to 13 p., and in thickness from 8 ju, to 11 /*, averaging about 26 p. by 12 p. by 10 p. The anterior end of the body is provided with a short contractile suctorial ten- tacle continuous with an internal tubular canal. The canal is usually directed ob- liquely toward the right side and can be traced in most fixed specimens stained with iron hematoxylin for about one-half the length of the body. The cilia of Hypocomidcs parva are about 6 p to 7 p. long and are strongly thignm- tactic, particularly near the base of the suctorial tentacle. The ciliary system con- sists of three separate complexes. The central complex, which occupies the larger part of the concave depression on the anterior part of the ventral surface, consists of eight rows which are about two-fifths the length of the body. To the right of this complex are two longer rows which originate on the dorsal surface close to the right margin behind the level of origin of the rows that constitute the central complex. Each of these rows is about three-fifths the length of the body and curves ventrally as it extends posteriorly. The outer row originates and terminates the more poste- riorly. To the left of the central complex is a series of eight rows which originate on the left lateral margin and on the left side of the dorsal surface and curve ven- trally as they extend backward. These rows are about two-fifths the length of the body and originate and terminate progressively more posteriorly. The outermost 206 EUGENE N. KOZLOFF row, however, is somewhat shorter than the other rows and is usually seen to termi- nate a little anterior to the point of termination of the seventh row. The cytoplasm is colorless and contains numerous small lipoid droplets and food inclusions. I have distinguished no large food vacuoles in this species. The con- tractile vacuole is situated near the middle of the body and opens to the exterior on the ventral surface. The macronucleus is typically ovoid, sometimes spherical, rarely sausage-shaped. It is situated in the posterior half of the body with its longitudinal axis usually placed obliquely to the longitudinal axis of the body. In fixed and stained preparations the chromatin appears to be aggregated into a dense reticulum enclosing a few vacuole-like clear spaces of varying size. These are most prominent near the periphery. In ten individuals fixed in Schaudinn's fluid and stained by the Feulgen reaction the macronucleus ranged in length from 4.2 //, to 8.2 ^ and in width from 4.2 /A to 5.3 i>.. 3. Hypocomides parra sp. nov. Distribution of ciliary rows, somewhat diagrammatic. A, dorsal aspect ; B, ventral aspect. The micronucleus is usually spherical and is situated dorsally a short distance anterior to or to one side of the macronucleus. Sometimes it is located as far ante- riorly as the middle of the body. In fixed and stained preparations the chromatin of the micronucleus appears to be homogeneous. In ten individuals fixed in Schau- dinn's fluid and stained by the Feulgen reaction the diameter of the micronucleus ranged from 1.9 /A to 2.3 />.. Hypocomides parra was present on the gills and palps of nineteen of the thirty- four individuals of Botula calijorniensis which I examined from localities near Moss Beach, California. It is sometimes associated with H. botulae and Insignicoma rc- nnstu. In my experience it is the most common species of ciliate parasitizing this mollusc. Hypocomides parra sp. nor. Diagnosis: Length 21/x-29/x, average about 26/<,; width 10^i-13/x, average about 12 /j. ; thickness S/i.-11/u., average about 10/<. The central ciliary complex comprises eight approximately equal rows about two-fifths the length of the body; CILIATES OF THE FAMILY ANCISTROCOMIDAE. IT 207 the right complex consists of two rows about three-fifths the length of the body; the left complex consists of eight rows, each about two-fifths the length of the body, the eighth row being usually somewhat shorter than the others. The macronucleus is typically ovoid. The micronucleus is spherical. Parasitic on the gills and palps of Bo tula californiensis (Philippi) (Moss Beach, California). Syntypes are in the collection of the author. HYPOCOMIDES KELLIAE SP. NOV. (Fig. 4; Plate I, Fig. 4) The body is elongated, narrowed anteriorly, and somewhat flattened dorso- ventrally. The ciliary system, to be described presently, is disposed for the most part on the shallow concavity occupying the anterior one-third of the ventral sur- FIGURE 4. Hypocomides kclliac sp. nov. Distribution of ciliary rows, somewhat diagrammatic. A, dorsal aspect ; B, ventral aspect. face ; the dorsal surface and that part of the ventral surface posterior to the ciliary area are convex. The anterior half of the left margin is not as rounded as the right margin, and typically is nearly straight or weakly indented. The body is widest near the middle and rounded posteriorly. Twenty living individuals taken at random ranged in length from 31 /u. to 37 /x, in width from 13 ^ to 15 /x, and in thickness from 1 1 /j. to 13 fji., averaging about 33 /JL by 14 ^ by 12 /A. The anterior end is provided with a contractile suctorial tentacle continuous with an internal tubular canal. The canal is usually directed toward the right side of the body as it extends posteriorly. It can be traced in most fixed individuals stained with iron hematoxylin for about three-fifths the length of the body. The thigmotactic cilia of H. kclliac are about 8 /* or 9 p in length. The ciliary system is composed of three separate complexes. The central complex, occupying the larger part of the anterior ventral depression, consists of five equal rows about one-third the length of the body. To the right of this system is a single long row about two-thirds the length of the body. This row originates on the dorsal surface 208 EUGENE N. KOZLOFF close to the right margin a little behind the level of origin of the central rows and curves ventrally as it extends posteriorly. To the left of the central complex is a .scries of live rows, one-third the. length of the body, which originate on the left lateral margin or dorsal surface on the left side and curve ventrally as they extend poste- riorly. These rows originate and terminate progressively more posteriorly. The cytoplasm is colorless and contains numerous small lipoid droplets and food inclusions. I have observed no large food vacuoles in this species. The contractile vacuole is situated near the middle of the body and opens to the exterior on the ven- tral surface. The macronucleus is ovoid or sausage-shaped, usually about two times as long as wide. It is situated in the posterior half of the body with its longitudinal axis placed obliquely to the longitudinal axis of the body. In fixed and stained prepara- tions the chromatin appears to be organized into a very dense reticulum which some- times is seen to enclose vacuole-like clear spaces of varying size. These are more evident near the periphery. In ten individuals fixed in Schaudinn's fluid and stained by the Feulgen reaction the macronucleus ranged in length from 7.8 ^ to 14 /x and in width from 3.9 ^ to 7 p. The micronucleus varies in shape from spherical to ovoid ; typically it is ovoid. It is usually situated dorsally a short distance anterior to the middle of the body or to one side of the anterior part of the macronucleus. In most fixed and stained individuals of H ' . kclliac the chromatin of the micronucleus is aggregated into deeply- staining peripheral strands. In ten specimens fixed in Schaudinn's fluid and stained by the Feulgen reaction the micronucleus ranged in size from 1.9^ by 1.5 /A to 2.3 /JL by 1.9/*. Hypocomides kclliac was present in nine of the twenty-eight individuals of K cilia laperousii which I examined from localities near Moss Beach, California. Also asso- ciated with this mollusc is a small ancistrumid ciliate which it may be possible for me to describe in a subsequent paper. Hypocomides kclliac sp. no?'. Diagnosis : Length 31 ^u-37 p, average about 33 ^ ; width 13 /x-15 p, average about 14 ju.; thickness 11 /x-13/t, average about 12 //.. The central ciliary complex consists of five rows about one-third the length of the body; the right complex consists of a single row about two-thirds the length of the body ; the left complex consists of five rows about one-third the length of the body. The macronucleus is ovoid or sausage- shaped. The micronucleus is typically ovoid. Parasitic on the epithelium of the gills and palps of Kcllia laperousii Deshayes (Moss Reach, California). Syntypes are in the collection of the author. INSIGNTCOMA VENUSTA GEN. NOV., SP. NOV. (Fig. 5; Plate I, Fig. 5) The body is elongated, narrowed anteriorly, and somewhat flattened dorso- ventrally. The anterior one-half of the ventral surface, on which the major part of the ciliary system is disposed, is weakly concave; the dorsal surface and that part of the ventral surface posterior to the ciliary area are convex. The anterior half of the left margin is usually not so rounded as the right margin and typically is nearly straight or slightly indented. The body is widest near the middle and rounded poste- C1LIATES OF THE FAMILY ANCISTROCOMIDAE. II 209 riorly. Twenty-five living individuals taken at random ranged in length from 42 /x to 52/i, in width from 18^ to 21 /x, and in thickness from 15 ^ to 18 //,, averaging about 48 ft by 20 /x by 17 /x. The anterior end of the body is provided with a contractile suctorial tentacle continuous with an internal tubular canal. In most fixed specimens stained with iron hematoxylin the canal can be traced for about three-fifths the length of the body. It usually appears to be directed obliquely toward the right side. The ciliary system consists of four separate complexes. Two long, widely-spaced rows on the right side of the body originate on the dorsal surface close to the right margin at the anterior end and are about two-thirds the length of the body. They curve ventrally as they extend posteriorly. A central complex of fourteen or fifteen FIGURE 5. Insiyiiiconia rcniisla gen. nov., sp. nov. Distribution of ciliary rows, somewhat diagrammatic. A, dorsal aspect; B, ventral aspect. rows occupies the larger part of the ventral anterior depression. These rows are on the average about one-half the length of the body and originate progressively more posteriorly toward the left side. The outer two or three rows on the left, however, do not usually terminate quite so far posteriorly as the twelfth row. The rows of this complex are usually a little more closely-set on the right side than on the left. To the left of the central complex is a series of sixteen or seventeen rows about one- half the length of the body which with the exception of the outer three or four rows are very closely-set. The innermost row originates on the left lateral margin near the base of the suctorial tentacle ; the remaining rows originate progressively more dorsally and posteriorly. The distal portions of several of the inner rows of this complex are usually visible on the left side of the ventral surface. Posterior to the middle of the body on the left side is a nearly V-shaped series of cilia which originates 210 EUGENE N. KOZLOI-T on the ventral surface in the posterior third of the body, extends anteriorly and to the left to a point a short distance behind the distal portion of the gap separating the central and left ciliary complexes, then bends abruptly backward and dorsally. The cilia of this fourth complex are about 12 ^ to 14 // in length. The cilia of the other rows are about 8 /u, or 9 //. in length and strongly thigmotactic, especially near the base of the suctorial tentacle. The cytoplasm is colorless and contains numerous small lipoid droplets in addi- tion to food inclusions. A few larger food vacuoles are sometimes observed near the posterior end. The contractile vacuole is situated near the middle of the body and opens to the exterior on the ventral surface. The macronucleus is ovoid or sausage-shaped and is situated in the posterior half of the body with its longitudinal axis usually placed obliquely to the longitudinal axis of the body. In fixed and stained preparations the chromatin appears to be aggre- gated into a dense reticulum enclosing vacuole-like clear spaces of varying sizes. These are most prominent near the periphery. In ten individuals fixed in Schau- dinn's fluid and stained by the Feulgen reaction the macronucleus ranged in length from 12 fj. to 17 /j. and in width from 4.4 /z to 9 /j.. The spherical micronucleus is commonly situated dorsally a short distance ante- rior to or to one side of the macronucleus. In fixed and stained preparations the chromatin appears to be homogeneous. In ten individuals fixed in Schaudinn's fluid and stained by the Feulgen reaction the diameter of the micronucleus ranged from 2.4 p. to 4 /A. Insif/nicoina t'cnusta was found to parasitize the gills and palps of nine of the thirty-four specimens of Botnla californiensis which I collected at localities near Moss Beach, California. Insignicoipia (/en. nov. Diagnosis: The body is elongated and somewhat flattened dorso-ventrally. The anterior end of the body is narrowed and provided with a contractile suctorial tenta- cle continuous with an internal tubular canal. The ciliary system consists of four separate complexes. The central complex, occupying the major portion of the shal- low concavity on the anterior one-half of the ventral surface, is bounded on the right by a small number of widely-spaced rows which curve ventrally as they extend poste- riorly ; to the left of the central complex is a series of closely-set rows about one- half the length of the body which originate progressively more posteriorly on the left lateral margin and dorsal surface on the left side and curve ventrally as they extend backward; the fourth complex consists of a V-shaped series of long cilia which lies immediately behind the distal portion of the gap separating the central and left ciliary complexes. The contractile vacuole is central and opens to the exterior PLATE I All figures have been drawn with the aid of a camera lucida from specimens fixed in Schau- dinn's fluid and stained with iron hematoxylin. X 1870. IMCUUK 1. Hypocomides niytili C'hatton and Lwoff. Ventral aspect. Ki<;n. no?'. Diagnosis : Length 42 /j-52 /x, average ah< »ut 48 /x ; width 18 /x-21 /x, average al>< >ut 20 /x; thickness 15it-18/i. average about 17^. Tlie central ciliary complex consists of fifteen (rarely fourteen) rows about one-half the length of the body which origi- nate progressively more posteriorly toward the left side; the right complex consists ot two widely-spaced rows about two-thirds the length of the body which originate on the dorsal surface close to the left margin and curve ventrally as they extend posteriorly; the left complex consists of sixteen or seventeen closely-set rows about one-half the length of the body which originate progressively more posteriorly on the left margin and dor.sal surface on the left side and curve ventrally as they extend posteriorly; the V-shaped series of cilia constituting the fourth complex originates on the ventral surface in the posterior third of the body, extends anteriorly and to the left to a point a short distance behind the distal portion of the gap separating vhe central and left ciliary complexes, then bends abruptly backward and dorsally. The cilia of the fourth complex are approximately 12^-14^ in length; those of the other three complexes are approximately 8 /x-9 /x in length. The macronucleus is ovoid or sausage-shaped; the micronucleus is spherical. Parasitic on the gills and palps of Botula calijorniensis ( I'hilippi) (Moss Beach, California). Syntypes are in the collection of the author. LITERATURE CITED CHATTOX, E., A\I> A. LWOKK, 1922a. Sur revolution des infusoires cks lamellibranches. Rela- tions des hypocomides avec les ancistrides. Le genre Hypocomides, n. .nen. (.'. A'. Acad. Sci. Paris. 175: 787. CHATTOX. E., AND A. LWOKK, 1922b. Sur revolution des infusoires des lamellibranches. Rela- tions des sphenophryides avec les hypocomides. C. K. .-I cud. Sci. Paris. 175: 1444. CHATTOX. E., AMI A. LWOKK. 1^24. Sur 1'evolution des infusoires des lamellibranches; morpho- logic comparee des hypocomides. Les nouveaux genres Hypocomina et Hypocomella. C. R. .lead. Sci. Paris. 178: 1928. CHATTOX, E.. AND A. LWOKK, 1926. Diagnoses de eilies thigmotriches nouveaux. Bull. Sac. Zatil. l-rancc. 51 : 345. RAABK, Z., 1938. Weitere L'ntersuchungen an parasitischen Ciliaten aus dem polnischen Teil der Ostee. II. Ciliata Thigmotricha aus den l'"amilien: Hypocomidae Biitschli und Sphaenophryidae Cli. & Lw. .-Inn. Mus. ::.m>l. palmi.. 13: 41. NATURAL HETEROAGGLUTININS IN THE HODY-FLUIDS AND SEMINAL FLUIDS OF VARIOUS INVERTEBRATES * ALBERT TYLER William G. Kerckoff Laboratories of the Biological Sciences, California Institute of Technology. Pasadena INTRODUCTION Since the early work of Lanclois (1875) numerous investigators have noted in the normal serum of various species of animals, particularly among the vertebrates, the occurrence of agglutinins that act on the cells of certain other species (cf. Wiener, 1943; Landsteiner, 1945; Thomsen, 1932). Such natural heteroagglutinins have also been frequently reported to occur in the serum or body-fluids of various inverte- brates (see literature in Huff, 1940; Tyler and Metz, 1945). Heteroagglutination reactions are frequently encountered in fertilizin studies and have been reported by several investigators (Lillie, 1913, 1919; Glaser, 1914; Just, 1919, 1930; Sampson, 1922; Godlewski, 1934; Hartmann et al., 1940; Runnstrom ct al., 1944) as occur- ring between spermatozoa and foreign egg-water preparations, body-fluids and sper- matozoa or their extracts. Further information concerning the range and nature of the heteroagglutination reactions is of importance, then, in analysis of problems of fertilization particularly in regard to the specificity and role of the interacting substances that are obtained from eggs and sperm. A study of heteroagglutination reactions with lobster serum (Tyler and Metz, 1945 ; Tyler and Scheer, 1945), which normally acts on the sperm or the blood-cells of a wide variety of species throughout the animal kingdom, showed that at least ten distinct relatively class-specific agglutinins are present in the serum of this spe- cies. This was determined by means of absorption tests. In the controls for those tests the supernatant fluids of the sperm suspensions used for absorption were also examined for agglutinating activity and it was found, particularly when the sperm had not been previously washed, that the fluids from sperm of some species were active on cells of certain other species. Tests were then made with the body-fluids and these, too, were found to be active. The present paper reports the results of this examination of the body-fluids and the sperm-supernatant s of various species of invertebrates for the occurrence of heteroagglutinins. The species examined were for the most part those that have been used or are potentially useful in fertilizin studies. In addition a few absorption tests were made with starfish body-fluid to determine whether or not its agglutinat- ing activity is attributable to the presence of several heteroagglutinins, each with broad specificity such as was found in lobster serum. MATERIALS AND METHODS The body-fluids of twelve species of animals among the annelids, echinoderms, mollusks and tunicates were examined for possible agglutinating action on sperm 1 This work has been aided by a grant from the Rockefeller Foundation. I am indebted to Miss Margaret L. Campbell for technical assistance. 213 214 ALBERT TYLER suspensions of various species of invertebrates and, in a few cases, on erythrocyte suspensions of various vertebrates. The body-fluids were obtained by incision into the body cavity or by insertion of a hypodermic syringe into the body cavity. In the case of Ciona the fluid was obtained directly from the heart. The fluids were clari- fied by centrifugation and tested in the manner previously described (Tyler and Metz, 1945). The supernatant fluids of sperm suspensions of eleven of the same species were also tested. These were obtained by centrifugation of approximately 10 to 20 per cent suspensions of "dry" sperm in sea water. They may be termed diluted seminal fluids. EXPERIMENTAL PART In all of the species that were examined the body-fluids were found to possess agglutinating activity for the cells of certain other species. The results are presented in Table I. In some species (e.g., 1, 6, 18, 19, 21) the fluids exhibited agglutinating action on the cells of most of the species that were tested. However, the fluids of TABLE I Heteroagglulinating action of body-fluid (b) and seminal fluid (s) of various invertebrates Spermatozoa (species 1 to 25) or erythrocytes (species 26-34) of: Fluids of: 1 b s 4 b s 6 b 12 b s 13 b s 14 b s 16 b s 18 b s 19 b s 21 b s 22 b a 23 b s POLYCHAETS 1. Chaetopterus variopedatus 2. Halosydna johnsoni 3. Sabellaria calif arnica. . . . 0 0 0 + + 0 0 0 0 0 0 + + 0 0 0 0 0 0 + + + + 0 0 + + + + 0 0 + + 0 + + + + + + + + + + 0 0 0 0 0 0 0 ECHIUROIDS 4. Urechis caupo + + + + 0 0 0 0 0 0 0 0 + + 0 0 0 0 0 0 0 0 + + + + + + + + + + + + + + + + + + + + + + + + + + 5. Thalassema sp AMPHINEURANS 6. Mopalia muscosa. + + + + 0 0 + o + + + + + 0 0 0 0 0 0 0 + + + + 0 0 + + 7. Ischnochiton magdalensis GASTROPODS 8. Acmea digitalis 9. Lottia gigantea + + + + + + + + + + + + o + + o 0 0 0 0 0 0 0 0 0 0 + + + 0 0 0 0 0 0 0 + + o + + 0 0 0 0 0 0 0 0 0 + o + + + + + + + + + 0 0 0 0 0 + + + + + + + + + + + + 10. Tegula galena . 11. Astraea undosa . 12. Megathura crenulata . . . . PELECYPOD 13. Mylilus calif ornianus . . . + + + 0 0 0 0 + + + + + + + + + ECHINOTDS 14. Strongylocentrotus purpuratiis 0 0 0 0 0 + + + + 0 0 0 0 0 0 + + + + + + 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 + o + 0 0 0 0 0 0 15. S. franciscanus 16. Lytechinus pictus 17. Dendraster exu-nlrii us. . . HETEROAGGLUTININS IN INVERTEBRATES 215 TABLE I — Continued Spermatozoa (species 1 to 25) or erythrocytes (species 26—34) of: Fluids of: 1 b s 4 b s 6 b 12 b s J3 b s 14 b s 16 b s 18 b s 19 b s 21 b s 22 b s 23 b s ASTEROIDS 18. Patiria miniata + + + + + 0 0 0 0 0 0 0 0 + + + + 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 + + 0 0 0 0 0 0 0 19. Pisaster ochraceus .... 20. Astropecten armatus HOLOTHURIOID 21. Stichopus calif ornicus . . . 0 0 0 0 0 0 0 0 0 0 0 0 0 0 ASCIDIANS 22. Ciona intestinalis + + + + + 0 0 0 0 0 0 + + 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 + + + + + + + + + + + + + + 0 0 0 0 0 0 23. Styela barnharti 24. Ascidia ceratodes FISH 25. Leuresthes tennis 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 + + + + + 26. Girella nigricans AMPHIBIA 27. Rana catesbiana + + + + o 0 0 0 0 0 0 0 0 + + + + 28. Bufo halophilus REPTILE 29. Sceloporns occidentalis . . . 0 0 0 0 BIRD 30. Chicken + 0 0 + + + + + + + 0 + 0 0 + 0 0 0 MAMMALS 31. Guinea pig . . . . + 0 0 0 0 0 0 0 + + + 0 32. Rabbit . 33. Sheep . . . .... 34. Man none of the twelve species examined were found to possess agglutinating action on all of the species that were tested. With one exception no heteroagglutination reactions occurred with body-fluids and cells of animals belonging to the same taxonomic class. The exception to this consists in the agglutination of Sabellaria'sperm by Chaetopterus fluid. It is of in- terest in this connection that these two genera are placed ( Pearse, 1936) in separate subclasses (cryptocephala and phanerocephala respectively) of the polychaets. Sabellaria also differs from the other two polychaets that were tested in that its cells fail to agglutinate in the fluid of the mussel (13), sea-urchins (14, 16) and sea cucumber (21) as well as that of the lobster previously (Tyler and Metz, 1945) reported. Another feature of the results is that closely related species behave alike with respect to the ability or inability of their cells to be agglutinated by the various body- fluids. This was evident in the previously reported experiments with lobster serum. 216 ALBERT TYLER In the present data differences arc observed between species belonging to different orders, as in the case of the two species (8 and 9) belonging to the docoglossid gas- tropods which differ from the three rhipidoglossid species (10, 11, and 12) in their reactions to two ol the llnids (6 and 21). With the other twenty-nine species tested, similarity in behavior is exhihited hy memhers of the same class or suh-class. The tests with the diluted seminal fluids gave results that in most cases paralleled those obtained with the body-fluids. Eight exceptions (1 on 29; 4 on 9 and 12; 12 on 6 ; 13 on 1 1 ; 18 on 8 and 30 ; 21 on 14) may be noted in Table I out of a total of 186 cross-combinations in which both seminal fluid and body-fluid were examined. These exceptions are all in the same direction ; namely, a failure of the dilute seminal fluid to cause agglutination while the corresponding body-fluid is active. It can, then, be stated that in those cases in which the diluted seminal fluid is found to pos- sess heteroagglutinating activity, the corresponding body fluid is likewise found to be active. Thus, similarly acting heteroagglutinins are found in both body-fluid and seminal fluid. The above-mentioned few exceptions may be attributed to failure to obtain, in certain seminal fluid preparations, sufficient concentration of a particular heteroagglutinin to produce a visible reaction with the cells of some species. How- ever, the tests necessary to determine the validity of this explanation have not, as yet, been made. From the similarity in action of seminal fluid and body-fluid, it might be in- ferred that the activity of the former is due to contamination with the latter. How- ever, it may be noted that in most of the species employed for preparation of seminal fluid (e.g., 4, 12. 14, 16, 18. 19, 21, 22) the sperm is readily obtained without any appreciable admixture of body-fluid. Another inter] >retation is that seminal fluid is normally similar to body-fluid in composition. Against this may be cited the fact that readily recognizable constituents of body-fluid, such as hemocyanin in the mol- lusks, are not observed in the seminal fluids. A third possibility is that identical heteroagglutinins are present in both fluids. However, serological similarity does not imply entirely identical molecular constitution. Reaction with a specific antigen implies similarity only on the part of the specific combining groups of the antibodies from diverse sources. In the present case it has not been shown that the hetero- agglutinin in seminal fluid and that in body-fluid both react with the same antigenic group or structure on the sperm that they agglutinate. However, the generally parallel behavior of the two fluids, when tested with spermatozoa of different species, favors that view. Absorption tests, to determine whether or not more than one heteroagglutinin is involved in the action of a particular fluid, were carried out with Patiria body-fluid. These were done in the manner previously described (Tyler and Metz, 1945). Be- fore being used for absorption the sperm were washed repeatedly in order to free them of agglutinins contributed by the seminal fluid, and this was checked in each test by examination of the supernatant of an aliquot part of the sperm for aggluti- nating activity. Samples of Patiria body-fluid were absorbed with sperm of six species of animals and tested for agglutinating activity on sperm of nine species. The results are given in Table II. This limited set of tests reveals the presence of at least four distinct heteroagglutinins in Patiria body-fluid. These evidently comprise : — one for the two polychaets, one for the two echiuroids and Mytilus, one for the two gastropods, and one for the two ascidians. It seems likely, then, that the situation in Patiria body- HETEROAGGLUTININS IN INVERTEBRATES 217 fluid is similar to that previously reported for the lobster; namely, the presence of a number of distinct agglutinins. each with broad group specificity. DISCUSSION Heteroagglutinins are evidently normal constituents of the body-fluids of animals. They have generally been considered to be non-specific agents. However, from the fact that the fluids of various species act on the cells of different assemblages of other species, the heteroagglutinins must be regarded as having some degree of speci- ficity. The previously reported (Tyler and Metz, 1945) absorption tests with lobster-serum gave evidence of the presence of ten distinct heteroagglutinins which are, for the most part, each specific for a taxonomic class of animals. The present results with Patiria body-fluids are indicative of similarly broad specificity on the part of the four heteroagglutinins found therein. It is clear, however, that hetero- agglutinins of different species and also different heteroagglutinins of the same ani- mal may differ considerably in the range of species on which they act. The rule is TABLE II Agglutinative activity of Patiria body-fluid after absorption with spermatozoa of various species Spermatozoa of: Body-fluid absorbed with washed sperm of: Chaetopterus Urechis Thalassema Astraea Megathura Mytilus Chaetopterus . . Halosydna Urechis 0 0 + + + + + + + + + 0 n + + 0 0 0 + + 0 0 + + 0 0 0 + + + + 0 0 + + + + + + + 0 0 + + + + + 0 0 + + 0 + + Thalassema .... Astraea Megathura Mytilus Ciona Ascidia. . . . that closely related species react alike to a particular fluid, but the closeness of rela- tionship required depends upon the particular heteroagglutinating fluid employed. In the present tests (Table I) we find that species that belong to the same class, in most instances, behave alike with respect to the ability or inability of their cells to be agglutinated by the body-fluids of all twelve of the species examined. In the sera of various mammals natural heteroagglutinins are found (cf. Thorn- sen, 1932; Wiener, 1943; Landsteiner. 1945) that are relatively species-specific and in human sera, as is well known, natural isoagglutinins are encountered that differ- entiate groups of individuals. In the various invertebrate body-fluids examined, agglutinins of such specificity have not, as yet, been found, the fluid of a particular species being inactive on cells of closely related species. However, from the ripe gametes, of many of these species natural agglutinins are obtained that act within the species. These consist in the fertilizins from eggs and antifertilizins from sperm described by Lillie (1913 et seq.), Just (1930), Frank (1939), Tyler (1939 et seq.), Hartman et al. (1939 et seq.), Runnstrom ct al. (1942 et seq.) and others. While these agents act on gametes of the opposite sex within the species, they also have 218 ALBERT TYLER been found to act on closely related species and in some instances the preparations act on remotely related species. Tims Arhacia egg water was found (Lillie, 1913) to agglutinate Nereis sperm, and sea-urchin eggs have been found (Runnstrom ct al., l'>42) to be agglutinated by sperm extracts of animals as remotely related as the salmon and the ox. In Lillie's experiment it was shown that absorption with Nereis sperm removed the cross-reacting substance from Arbacia egg water without dimin- ishing its agglutinating action on Arhacia sperm. Similar absorption experiments have not been reported in most of the cross-heteroagglutination reactions obtained by later workers with fertilizin and antifertilizin preparations. The need for such experiments is quite evident before any adequate determinations can be made of the specificity of these interacting substances obtained from the gametes. The present results may provide a helpful basis of procedure in such experiments. The bearing of the heteroagglutination reactions on phylogenetic questions has been previously (Tyler and Metz, 1945) discussed. The present results are con- sistent with the previously expressed view that the reactivity of the cells of a par- ticular species with various fluids is a characteristic of a group of related species and constitutes a group-specific trait in addition to the various group-specific morpho- logical and chemical features of animals. There is no reason, as yet, for considering this trait to be of more general significance than any other in any applications that might be made to phylogenetic problems. SUMMARY 1. The body-fluids of 12 species of invertebrates (including two ascidians) and the seminal fluids of 11 species were examined for agglutinating action on the sper- matozoa or blood cells of 34 species of animals. 2. All of the fluids were found to contain agglutinins for the cells of some of the species tested. Five of the fluids gave reactions with most of the species but none reacted with all of the species. 3. With one exception no heteroagglutination reactions were obtained with fluids and cells of animals belonging to the same taxonomic class. 4. Closely related (same class in most cases or same order in some) species were found to behave alike with respect to the ability or inability of their cells to react to the various fluids, and the fluids of closely related species exhibited similar reactivity. 5. The diluted seminal fluids gave reactions that in most cases paralleled those obtained with the body-fluids. 6. Absorption tests revealed the presence of at least four distinct heteroagglu- tinins in Patiria body-fluid, and indicated that each is characterized by a broad group-specificity similar to that previously reported for lobster-serum. 7. The general bearing of these results on fertilizin-antifertilizin reactions and on phylogenetic problems is briefly discussed. LITERATURE CITED I "RANK, J. A., 1939. Some properties of sperm extracts and their relationship to the fertilization reaction in Arbacia punctulata. Biol. Hull., 76: 190-216. (ii.ASKK. Orro, 1914. A qualitative analysis of the egg secretions and extracts of Arhacia and Asterias. Rial. Bull.. 26: 367-386. ( ioiu IAVSKI, H., 19.34. Nouvelles recherches sur I'heteroagglutination des spermatozoiides et sur 1'action d'extraits de cellules sexuelles d'esptn-s t'trani>eres. Arch, df Biologie, 45: 735-807. HETEROAGGLUTININS IX INVERTEBRATES 219 HARTMANN, M., AND SCHARTAU, 1939. Untersuchungen uber die Befruchtungsstoffe der Seeigel. I. Biol. Zcntralbl, 59 : 571-587. HARTMANN, M., O. SCHARTAU, AND K. WALLENFELS, 1940. Untersuchungen uber die Be- fruchtungsstoffe der Seeigel. II. Biol. Zentralblatt, 60: 398^23. HUFF, C. G., 1940. Immunity in invertebrates. Physiol. Rev., 20 : 68-88. JUST, E. E., 1919. The fertilization reaction in Echinarachnius parma II, the role of fertilizin in straight and cross-fertilization. Biol. Bull., 36: 11-38. JUST, E. E., 1930. The present status of the fertilizin theory of fertilization. Protoplasma, 10: 300-342. LANIJOIS, L., 1875. Die Transfusion dcs Blutes. Leipzig. LANDSTEINER, K., 1945. The specificity of scrological reactions. Harvard Univ. Press, Cam- bridge, Mass. LILLIE, F. R., 1913. Studies of fertilization. V. The behavior of the spermatozoa of Nereis and Arbacia with special reference to egg-extractives. Jour. E.vp. Zoo/., 14: 515-574. LILLIE, F. R., 1919. Problems of fertilization. Univ. Chicago Press, Chicago. PEARSE, A. S., 1936. Zoological names. A list of phyla, classes and orders. Prepared for sec- tion F, A. A. A. S. Duke University Press, Durham, N. C. RUNNSTROM, J., S. LINDVALL, AND A. TISELIUS, 1944. Gamones from the sperm of sea urchin and salmon. Nature, 153 : 285. RUNNSTROM, J., A. TISELIUS, AND S. LINDVALL, 1945. The action of androgamone III on the sea-urchin egg. Ark, f. Zool. (Stockholm), 36A, No. 22: 1-25. RUNNSTROM, J., A. TISELIUS, AND E. VASSEUR, 1942. Zur Kenntnis der Gamonwirkungen bei Psammechinus miliaris und Echinocardium cordatum. Ark. f. Kcmi (Stockholm) ISA. No. 16: 1-18. SAMPSON, M. M., 1922. Iso-agglutination and hetero-agglutination of spermatozoa. Biol. Bull., 43 : 267-284. THOMSEN, O., 1932. Serologie der Blutgruppen. Chap. 2 of P. Steffan's Handbuch der Blut- gruppcnkundc. J. F. Lehmanns Verlag, Munich. TYLER, A., 1939. Extraction of an egg membrane-lysin from sperm of the giant keyhole limpet (Megathura crenulata). Proc. Nat. Acad. Sci., 25: 317-323. TYLER, A., 1942. Specific interacting substances of eggs and sperm. Jl'cst. J. Surq. Obst. and Gyn., 50: 126-138. TYLER, A., 1942. A complement-release reaction; the neutralization of the anticomplementary action of sea-urchin fertilizin by antifertilizin. Proc. Nat. Acad. Sci., 28: 391-395. TYLER, A., AND C. B. METZ, 1945. Natural heteroagglutinins in the serum of the spiny lobster, Panulirus interruptus. I. Taxonomic range of activity, electrophoretic and immunizing properties. Jour. £.r/>. Zool., 100: 387-406. TYLER, A., AND B. T. SCHEER, 1945. Natural heteroagglutinins in the serum of the spiny lobster Panulirus interruptus. II. Chemical and antigenic relation to blood proteins. Biol. Bull.. 89: 193-200. WIENER, A., 1943. Blood groups and blood transfusion. 3rd ed. C. C. Thomas, Springfield. DISTRIBUTION AND PROPERTIES OF INTRACELLULAR ALKALINE PHOSPHATASES l EDITH JUDITH KRUGELIS -1 Department of Zoology, Columbia University, Neiv York The chemical and structural composition and the interrelationships of the com- ix uients of the cell have long been the concern of biologists, in hope of revealing the fundamental processes within the structural and functional unit of all living matter. The early studies of Miescher upon the chemical composition of the nucleus were con- temporary with the studies of the structure of the nucleus by van Beneden, Flem- ming, and others, which later culminated in the firm establishment of the nucleus as the site of phenomena of central importance in the mechanism of Mendelian heredity. The biochemical line of investigation, started by Miescher, has had no such tradition of continuous progress as have the studies of chromosomal structure ; but as a result of sporadic advances following the appearance of new methods and techniques, cer- tain facts about the chemical composition of the cell and the nucleus have become firmly established. Miescher noted that the nucleus had a high content of organically bound phosphate; later researches (Levene and Bass, 1931) have fully confirmed this, and have led to a fairly comprehensive knowledge of the chemical composition of nucleic acid, the substance in which all this nuclear phosphate is contained. The unit of nucleic acid is the mononucleotide, a phosphoric acid ester of pentose sugar in glucosidic linkage with one of the purine or pyrimidine heterocyclic bases. The nucleic acids, which have been intensively studied, have been found to be polynucleo- tides with four different bases, and in their native form they are very highly poly- merized, consisting in some cases of hundreds or thousands of nucleotides. There are two types of nucleic acid with respect to the sugar ; one contains a pentose which whenever identified has been found to be d-ribose, and the other a desoxypentose found to be d-ribodesose. The wide application of the Feulgen cytochemical test for desoxypentose indicates that nucleic acid with this sugar does not occur outside of the chromatin of the cell nucleus (that is, in the chromosomes), and actual analysis of isolated nuclei and chromosomes has confirmed the view that desoxypentose is the characteristic nucleic acid of chromatin (Mirsky and Pollister, 1942). The name chromonucleic acid has recently been proposed (Pollister and Mirsky, 1943) for desoxypentose nucleic acid as a convenient way of emphasizing this striking limi- tation in distribution. Nucleic acids can also be located within the cell by means of the intense specific absorption of ultra-violet light by purine and pyrimidine bases, and Caspersson ( 1936. 1940) has made use of a combination of microspectroscopy and the Feulgen nucleal reaction to determine the distribution of the Feulgen nega- tive pentose nucleic acids. This was found to be largely in the cell cytoplasm, the occurrence of pentose nucleic acid in the nucleus being confined to the plasmosome, 1 Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy, in the Faculty of Pure Science, Columbia University. 2 Present address: Department of Zoology, Vassar College, Poughkeepsie, N. Y. 220 INTRACELLULAR ALKALI XE PHOSPHATASES 221 or the non-chromatin nucleolus. The term plasmonucleic acid has been proposed to embody this information concerning the distribution of pentose nucleic acid. The nucleic acid makes up but a part, usually less than half, of the content of the nucleus. The remainder is apparently largely protein (Mirsky and Pollister, 1942). Much of this is of the basic type, either histone or protamine, but there is also evidence of protein of globulin type (Caspersson, 1936; Mazia and Jaeger, 1939). In view of the fact that the nuclear chromatin contains a large amount of phos- phate in desoxyribose nucleic acid, it is of considerable interest that definite evidence of phosphatase activity within chromatin has been found. Almost simultaneously the author (Krugelis, 1942) and Willmer (1942) reported the demonstration of alkaline phosphatase activity within chromosomes. This discovery stimulated the present study, the purpose of which is two-fold. The first purpose was to investigate the occurrence of alkaline phosphatase in all parts of the cell, and to investigate further the correlation of desoxyribose nucleic acid with alkaline phosphatase activity in the chromosomes. The material which seemed most promising for this first part of the study were the salivary gland chromosomes of Drosophila larvae, because of their peculiar structure which manifests itself in bands or regions rich in desoxyribose nucleic acid, alternating with interband regions poor or lacking in desoxyribose nucleic acid. This part of the investigation can be expressed in the form of a ques- tion : Is phosphatase activity located in the chromosomal regions rich in desoxyribose nucleic acid? The second part of the investigation followed from the first, and in- cluded the use as substrate of actual chemical substances that are found in the cell, in order to discover whether alkaline phosphatase is capable of hydrolyzing naturally- occurring substances with which its activity seems to be physically associated. If the activity of phosphatase is located in regions rich in desoxyribose nucleic acid, is it capable of hydrolyzing desoxyribose nucleic acid and other naturally occurring substances in the cell ? This part of the investigation it was hoped would lead to some suggestion as to the possible role of phosphatase activity within the cell and the chromosome. METHODS AND MATERIALS The histochemical method used was that developed by Gomori (1939, 1941); when using smeared insect material it was found necessary to make some modifica- tions of the preliminary steps. This process involved dissecting, and smearing the insect organs in isotonic saline, or Ringer's, made according to Buck and Melland (1942). After smearing, the material was fixed in 95 per cent ethyl alcohol vapors for one hour, then fixed in liquid 95 per cent ethyl alcohol for two hours. Material fixed as short a time as one hour and as long a time as 48 hours showed little appreci- able difference in the results. The coverslips fell off readily in the alcohol, and the smears on the slides were washed briefly in distilled water and then incubated in the substrate solution. The mammalian organs used, namely, the kidney, liver, testis, intestine, and pancreas, were removed after ether anesthesia, fixed in 95 per cent ethyl alcohol for 24 hours, dehydrated in 100 per cent ethyl alcohol, cleared in ben- zene, and embedded in paraffin in the usual manner. Sections of 5 microns thickness were cut. To avoid any variation in mounting or incubating, sections of the five organs were mounted together on one slide. After deparaffmization, hydration, and washing, sections were ready for immersion and incubation in the substrate solution. EUITH JL'DITH KRUGELIS The process from this point on was the same for both the sectioned and the smeared materials. The substrate solution was of the following composition : 2 parts of 0.1 M sodium glycerophosphate (or 0.1 M other organic phosphate ester) 2 parts of 0.1 M calcium nitrate 1 part of 0.1 M veronal buffer at pH 9.4 5 parts of distilled water. The pH was checked with thymol blue and adjusted to pH 9.0 to 9.2 with NaOI I. The incubation was carried out at 25-28° C. and lasted from 12 to 24 hours. Fol- lowing the incubation in the substrate, the slides containing the smears or the sec- tions were immersed in 0.02 M calcium nitrate for one minute, then into 0.1 M co- baltous nitrate for two minutes, then dilute potassium sulfide for two minutes. After this visualization treatment, slides were washed, dehydrated, cleared, and mounted by the usual histological technique. The results were checked, in some cases, by parallel experiments in which the silver nitrate visualization method (Gomori, 1939) was used. When no activity was obtained in the substrate-incubated slides, the re- action was allowed to proceed for 72 hours to verify the negative results. At the suggestion of Dr. A. E. Mirsky partially depolymerized nucleic acid was tried as the substrate. This depolymerization was accomplished by using desoxy- ribonuclease from beef pancreas, made according to the method of McCarty (1946). Desoxyribose nucleic acid is dissolved in distilled water and then placed in a nuclease medium of the following composition : 0.01 per cent magnesium sulfate 0.003 M gelatin 0.025 M veronal-HCl buffer at pH 7.5. The nuclease is allowed to react upon the desoxyribose nucleic acid for 24 hours in which time the solution becomes clear and loses its viscosity. This change is used as a criterion for the depolymerization of the nucleic acid. This depolymerization leaves the nucleic acid in only a less highly polymerized state ; and does not, for ex- ample, reduce it to an approximation of tetranucleotide. After the depolymerizing reaction, the solution is heated to 60° C. for fifteen minutes to destroy the activity of the nuclease on the desoxyribose nucleic acid, and then is used as the substrate source of phosphate ester. The biological materials used in this study were insect and mammalian organs. The salivary glands of the larvae of Drosophila were used for the first part of the experimentation. The larvae were raised at 17-18° C. and at room temperature. The glands were dissected out when the larvae were full grown and had left the food prior to pupation. The species of Drosophila investigated included the Bra- zilian species Drosophila pallidipcnnis obtained through the generosity of Professor T. Dobzhansky, D. sinndons, D. virilis, and D. anunussae obtained from the stocks at the Biological Laboratories, Cold Spring Harbor, X. Y. The major part of the investigation on Drosophila was done on Drosophila melanogaster. The mammalian organs used were those of mice obtained from the stocks main- tained at Columbia I'niversity by Professor L. C'. Dunn and Dr. S. Gluecksohn- Schoenheimer. The testes were used either after being smeared in the body fluid surrounding the tubules, or after being fixed and embedded as the other organs were. INTRACELLULAR ALKALINE PHOSPHATASES Yeast nucleic acid and adcnylic acid were obtained from the Schwartz Labora- tories, New York; thymonucleic (desoxyribose nucleic) acid and nuclease were gen- erously supplied by Dr. A. K. Mirsky of the Rockefeller Institute for Medical Re- search; guanylic acid, sodium guanylate, and cytidylic acid were obtained from the Levene collection of chemicals at the Rockefeller Institute, also through the kind- ness of Dr. A. E. Mirsky. Photographic materials were provided from a grant to Columbia University by The Rockefeller Foundation. EXPERIMENTAL DATA Intracellular occurrence and localization o] alkaline phosphatase activity in the sali- vary gland cells of Drosophila melanogaster Upon microscopic examination of smears of larval salivary glands of Drosophila, it is evident that in the preparations which have been incubated in the substrate, there is a much greater precipitate of cobaltous sulfide. indicating alkaline phos- phatase activity, than in the smears incubated in control solutions. A diffuse pre- cipitate occurs throughout the cytoplasm, but there is a much higher density within the nucleus (Fig. 1). This agrees with the observations of Moog (1944), on the tissues of the chick embryo that the nuclei were never less reactive than the cyto- plasm and that the nuclei, unlike the cytoplasm, were never negative in reaction. Dounce (1943) using nuclei isolated from rat liver, also found that alkaline phos- phatase activity is much greater in nuclei than in total liver. Within the nucleus the precipitate is sharply localized in the chromosomes and nucleolus. The nuclei of the control slides show slight, if any, precipitate, with the exception that there is always some cobaltous sulfide in the control nucleoli (Fig. 2), hut this is always considerably less in amount than that in nucleoli of substrate- incubated preparations. The most dense precipitate is found within the chromosomes. When smeared in Ringer's solution, which is necessary to preserve the activity of the enzyme, the details of the chromosome structure are byno means as distinct as in the acetic acid smears that have been used in mapping bands. Nevertheless one can be certain that the cobaltous sulfide within the chromosome is concentrated in transverse bands, alternating with regions of very little or no precipitate (Figs. 3 and 5). This obser- vation was reported earlier (Krugelis, 1945) and it has recently been confirmed by Danielli and Catcheside (1945) who compare the individual bands with specific re- gions of the cytogenetic chromosomal maps. In the preparations of Figures 3 and 5 the identity of the regions of phosphatase precipitate with the bands of acetic acid smears was established by a direct cytological observation. This method made use of the fact that the cobaltous sulfide precipitate is not altered by the standard Feulgen cytochemical procedure, that it becomes possible to superimpose the nucleal reaction on a phosphatase test. Figure 4 is from such a preparation ; and comparison with Figure 5, the same chromosome after the phosphatase reaction but before the Feul- gen test, shows that the bands in which the enzyme activity is localized correspond strictly with the Feulgen positive bands that form the basis of the familiar cytogenetic maps. The superposition of the second procedure merely changes the brown band to a purple brown, which is clearly a combination of cobaltous sulfide and basic fuch- sin. This combination of the two cytochemical tests shows, in a vivid manner, that the highest level of alkaline phosphatase activity within the salivary gland cell occurs 224 EDITH JUDITH Kkt'(;ELIS in exactly the region that contain* a very high concentration of organic phosphate in the form of desoxyrihose nucleic acid. Properties and dislrilnition o{ intracellular alkaline phosphatascs The occurrence ol the enzyme activity and the desoxyrihose nucleic acid within the same parts of the salivary gland chromosome suggests a functional relationship, which suggestion was put to test hy using other phosphate esters as suhstrates for the enzyme activity. In order to elucidate a more specific picture of possible en- zyme activity in vivo within the nucleus and the cell, esters were chosen which are normally present in the cell, and which the enzyme would consequently he expected to hydrolyze. The cellular substrates used, along with sodium glycerophosphate as the standard, were purine and pyrimidine nucleotides. ribose nucleic acid (presumably not highly polymerized), and desoxyribose nucleic acid (both polymerized and par- tially depolymerized). These substrates were tried on various mouse tissues for the presence of the enzyme activity in the cells. The testis was used for the chromo- somal localization, since the other tissues contained nuclei which were mainly in the resting stages. In the results the black-brown precipitate of cobaltous sulfide indi- cates the site of alkaline phosphatase activity on the specific organic substrate in the incubating solution, and the density of the precipitate may be regarded as a measure of the intensity of the enzyme activity. The photomicrographs (cf. Plates II, III, and IV) illustrate the results of these experiments ; and roughly quantitative estimates of the intensities of the reaction are summarized in Table I. With sodium glycerophosphate and the nucleotides as substrates, the precipitate is not sharply localized, though it is always more dense in the nucleus and chromosomes than in the cytoplasm. By contrast, the reactions with depolymerized nucleic acids result in precipitates which are restricted to par- ticular parts of the cell. On the slide in which depolymerized chromonucleic (des- oxyribose nucleic) acid has been used as-substrate the dark nuclei and chromosomes stand out very distinctly against a colorless cytoplasmic background, an appearance not unlike that of a Feulgen preparation. On the other hand, if plasmonucleic (ribose nucleic) acid is the substrate, the slide as a whole is intensely dark due to cytoplasmic precipitate, and the nuclei appear as light areas. The two types of slides compare with one another somewhat as do a photographic positive and nega- tive. The chromonucleic acid precipitate is not uniformly distributed throughout the EXPLANATION OF PI.ATK I The dark precipitate indicates the phosphatase activity. Photomicrographs made from smear preparations. Figures 1, 3-5 show phosphatase activity with sodium glycerophosphate as the substrate. FIGUKK 1. Alkaline phosphatase activity in salivary gland of Drosophila melanogaster larva. Magnification is 200 X. EIGURK 2. Salivary gland nucleus from control slide which had no substrate for the alkaline phosphatase activity. Magnification is 1000 X. l-'ua'KK 3. Salivary gland nucleus showing alkaline phosphatase activity. Magnification is 1000X. Fu.rkK 4. Salivary gland chromosomes with the Feulgen reaction applied over the phospha- t:iM' reaction precipitate. Magnification is 3000 X. Fi<;rKK 5. Salivary gland chromosomes (same chromosomes as Fig. 4) showing alkaline phosphataM- activity before Feulgen reaction applied over the precipitate. Magnification is 3000 X. INTRACELLULAR ALKALI XK PHOSPHATASES 225 PLATE I %f • r * - 1 - 226 KDITH JUDITH Kkl'CKl.lS nucleus. Instead it occurs only within the formed structures of the nucleus, the chromatin and the plasmosome nucleolus. In an attempt to find enzyme differences involved in producing such specific localizations of alkaline phosphatase activity, these reactions were subjected to vari- ous conditions which might alter the reaction environment. Exposures to tem- peratures of 55° C. or over for a period of 5 minutes or longer produced complete irreversible inactivation of all the enzyme reactions. Magnesium ions in the concen- tration of 0.01 M magnesium sulfate showed some activation with the cytoplasmic reaction on ribose nucleic acid, and little or no increase on the nuclear and the gen- eral reactions. This very slight activating effect is in agreement with Schmidt and Thannhauser's (1943) observation that there is but slight effect of magnesium on intestinal alkaline phosphatase activitv with sodium glycerophosphate as substrate. TABLE I Phosphatasc reaction with different substrates Intestine Testes Liver Pancreas Kidney Substrate C n c n c n c n c n Sodium glycerophosphate 1 2 1 2 0 2 ±1 2 3 2 Adenylic acid, guanylic 1 2 1 2 0 2 0 2 1 2 acid, cytidylic acid Desoxyribose nucleic acid 0 0 0 0 0 0 0 0 f) 0 Depolymerized desoxyri- 0 2 0 2 0 2 0 2 1 2 1 a >-r nucleic acid Ribose nucleic acid 2 ±1 2 1 1 ±1 2 ±1 2 1 Density of cobaltous sultkle precipitate recorded from visual estimates. 0, no reaction; ± doubtful; 1, definite reaction; 2, strong reaction; .}, very strong reaction. Each record based on microscopic examination of at least 20 slides. Column c, cymplasmic precipitate; column n, nuclear precipitate. Cyanide ions in the concentration of 0.01 M potassium cyanide completely inhibited all three types of reaction. . \rsenate ions in the concentration of 0.01 M sodium arsenate showed a more .selective effect on the different reactions than did the environmental influences men- tinned above. In the presence of arsenate, the alkaline phosphatase activity with the nucleic acids as substrates is completely suppressed. The glycerophosphate splitting, by contrast, in most case's proceeds at its normal rate in a medium which contains arsenate. A roughly quantitative estimation of the inhibition or suppression of ac- tivity by arsenatc ions is presented in Table II. I )iscrssio\ The differences in site of alkaline phosphatasc activity in the same tissues when Using different phosphate bearing substrates indicate localized reactions o! three INTRACELLULAR ALKALINE PHOSPHATASES PLATE II 227 < tt Photomicrographs from 5 micra thick sections. Figures 6-8 show alkaline phosphatase ac- tivity with sodium glycerophosphate as suhstrate. FIGURE 6. Small intestine of mouse showing activity of the enzyme' distributed in the mu- cosa and suhmucosa layers. Magnification is 1000 X. FIGURE 7. Testis of mouse. Magnification is 1000 X. FIGURE 8. Testis of mouse. M indicates chromosomes in metaphase. A indicates chromo- somes in anaphase. Magnification is 1000 X. FIGURE 9. Testis of mouse from a control slide with no suhstrate for enzyme activity ap- plied. .17 indicates chromosomes in metaphase. Magnification is 1000 X. 228 KMITII H'DITH KRUGELIS types, as lollows : tirst. a ^eiieral alkaline phosphatase reaction in both tin- cvtoplasm and the nucleus, as in the case where sodium glycerophosphate and nucleotides art- used as substrates; second, a definite nuclear reaction with little or no activity in the cytoplasm, as in the case of depolymerized desoxyribose nucleic acid as substrate; third, a definite cytoplasmic reaction with little reaction in the nucleus, as in the case of rihose nucleic acid as substrate. Since the enzymes cannot he isolated at this time, and since the nuclear and the cytoplasmic reactions are not absolutely specifi- cally nuclear or cytoplasmic, they will be listed as reaction types with reference only to the location of the activity observed. \Yith the different substrates used, the total reaction of phosphatase activity is somehow produced, and this is detected by the location of the cobaltous sultide precipitate. The final reactions produced under a TAMLK II Intensity of precipitate of phosphatase reaction in diffaent tissue?, using different substrates with nnil without arsennte ions Substrate Intestine Testes Liver Pancreas Kidney c n c n c n c n c n Sodium glycerophosphate Xo arsenate \»n> \\ ith arsenale ions 1 1 2 1*S1 1 2 2 0 0 2 0*S3 ±1 ±1 2 2*SO 3 2 2 ±1*S1 Depolymerized des<>\\ ribi^r n. acid No arsenate ions 0 0 2 0*S3 0 0 2 0*S3 (1 0 2 0*S3 0 0 2 ± 1 *S2 1 0 2 0*S3 With arsenate ions Rihose nucleic acid Xo arsenate ions 2 0 ±1 0*S3 2 2 1 0*S3 1 0 ±1 0*S3 2 0 ±1 0*S3 2 2 1 1*SO With arsenate ions. . . . : suppression estimates. 50 i-, no suppression. 51 is slight suppression. 52 is much suppression. 53 is complete suppression. Kach record represents observation on 8 experimented slides. variety of substrate conditions must be due to at least three different enzymes, work- ing in at least two different com] ilexes. The suggestion that the reactions observed are due to complexes of enzymes is based on the chemical structure of nucleic acids. The ribose and desoxyribose nu- cleic acids, which were applied as substrates, are polymers of mononuclebtides, which are considered to be linked to each other by an ester bond between the phosphate .ijroiip of one nucleotide and the sti^ar .^roiip of the nei^hbonn^ nucleotide. thus mak- ing the nucleic acids diesters of phosphoric acid all alon<^ the chain except in the terminal monoester of phosphoric acid. The total phosphatase action on the nucleic acids miidit lie considered as essentially due to a specific desoxyrihose nucleic acid phosphodiesterase, and a specific ribose nucleic acid phosphodiesterase action liberat- ing mononucleotides of the nucleic acids, which then are hydrolyxed by a phospho- IXTRACELLl'LAK ALKALINE PHOSPH AT \SES PLATK III . *k ^* ^ * N to X .* V i • ^ .,1 m 9 0 12 «^ •> 13 Photomicrographs from 5 micra sections. P'igures 10-13 show alkaline phosphatase activit\ with depolymerized desoxyribose nucleic acid as substrate. FIGURE 10. Small intestine of mouse showing the distribution of the enzyme activity in the mucosa layer. Magnification is 1000 X. FIGURE 11. Kidney of mouse showing enzyme activity in the proximal tubules seen in cross- section. Magnification is 1000 X. FIGURE 12. Pancreas of mouse showing distribution of enzyme activity. Magnification i.- 1000 X. FIGURE 13. Testis of mouse showing enzyme activity. Magnification is 1000 X. F.niTII Jl'MITIl KKl'l.KUS monoe.sterase to liberate the inorganic phosphate. Thus the precipitate formed when tlie nucleic acids are applied as substrates is flue to the activity of at least two dif- ferent enzvme complexes, as lollows: in the cytoplasm, first, a specific ph >spho- diesierase ("plasmonucleodiesterase") liherates ribonucleotides ; second, phospho- nioiioest erase liherates inorganic phosphates from these nucleotides; in the nucleus. first, a specific phosphodiesterase 3 ("chromonucleodiesterase") liherates desoxy- ribosenucleotides ; second, phosphomonesterase liherates inorganic phosphate from these nucleotides. This specificity of the diesterase follows from the fact that rihose nucleic acid will not serve as substrate for the nuclear diesterase activity, nor will desoxyribose nucleic acid serve for the cytoplasmic activity. \Yith regard to the monoesterase activity, however, no such specificity has been detected, for the products nf rihose nucleic acid hydrolysis serve equally well as substrates for phosphate pro- duction in either nucleus or cytoplasm. While the action of the phosphodiesterase in freeing mononucleotides from lower polynucleotides (depolymerized nucleic acid) presumably can only occur by attack upon the linkage between the phosphate of one nucleotide and the sugar of the adja- cent unit, it is evident that there is considerable restriction upon the exact nature of the bond which can be so attacked. If the enzyme were able to hydrolyze the bond at many points along a nucleic acid chain (consisting, let us say, of 2000 nucleotides) the diesterase should also function as a depolymerase. and a phosphate precipitate should be formed when the long, polynucleotide chains (polymerized nucleic acid) are used as substrate. 1C veil after 72 hours action, however, there is no visible pre- cipitate under these conditions, in contrast to the depolymerized nucleic acid experi- ments in which a dense precipitate is formed in a few hours. The type of bond which the diesterase can attack is evidently one which is enormously multiplied by a process of depolymerization. A most obvious view of enzyme specificity that would agree well with these facts is that the diesterase can attack only the bond between a ter- minal nucleotide and the penultimate nucleotide. Considering, for example, an ex- treme case, depolymerization of a 2000 unit polynucleotide chain to the minimum tetranucleotide should increase the number of bonds which such a specific diesterase can attack by a factor of 500. If specific terminal hydrolysis is the mechanism, the amount of mononucleotide that would become available by diesterase action upon the end of a highly polymerized nucleic acid chain would surely give an amount of phos- phate precipitate so slight that it would he cytologically undetectable. It is highlv important to the question of the functional significance of the distribu- tion of the phosphatases acting on the nucleic acids, that this actually coincides in a striking manner with the known locations of desoxyribose and rihose nucleic acids within the cell. The phosphatase activitv on depolymerized desoxyribose nucleic acid is restricted to the chromatin of the nucleus, the only part of the cell in which this type of nucleic acid is found. I'.y contrast, the riho.se nucleic acid phosphatase activity is in the cytoplasm, a region in which only rihose nucleic acid ha.s ever been demonstrated. ( 'I he only marked discrepancy is the occurrence of desoxyribose oncerning the existence <>t this nuclear diesterase, there is Mime possible supporting evi- dence from the work by Mazia and I'.allentine, reported by Maxia (In41), on an intranuclear enzyme troiii Arbaeia etiL's. Their enzyme, termed polynucleotidase, was active at a pi I ('JI and w;i oi reactin.L' on dcsoxyrihose nucleic acid still in a polymerized form. INTRACFLIA'LAK ALKALINE PHOSPH ATASF.S 231 16 17 Photomicrographs from 5 micra tliick sections. Figures 14-17 show alkaline phosphatase activity with ribose nucleic acid as the substrate. Fn.rkK 14. Small intestine of mouse showing the enzyme activity in tlie cells of the inucosa layer. Magnification is 1000 X. FiorKK 15. Small intestine of mouse showing goblet cells in the mucosa layer and the distri- bution of the enzyme activity. Magnification is 1000 X. Fioi'RK 16. Testis of mouse showing the enxyme activity. A denser precipitate is found in the nuclei of testis cells than in nuclei of other tissues under the conditions of the same substrate. Magnification is 1000 X. Fu;rKK 17. Kidney of mouse, showing enzyme activity in the proximal tubules in cross- section. Magnification is 1000 X. EDITH JUDITH KRUGELIS nucleic acid phosphatase activity in the nucleolus. which, since it is Feulgen negative. - ci m.sidered to contain ribose nucleic acid.) ' iVi'tain possible in -i-iro functions ol" these alkaline phosphatases are at once obvi- • in ."«. Xot only can the nuclear diesterasc split off terminal mononucleotides, as in these experiments, lint in the reverse direction, it niav conceivably catalyze the ier- •• j * ininal growth in the development of nucleotide chains. ( hie may easily picture the later stages of synthesis of a full length nucleic acid chain as involving the coopera- tion of two enzymes: the diesterase slowly builds up short chains, by terminal growth and this is followed bv the action of the depolymerase type of enzyme catalyzing the union of these short chains into the long complex which is such an important struc- tural component of a chromosome. While it is also obvious that catalysis of the synthesis of a mononucleotide from a nucleoside by phosphomonesterase is an essential step in nucleic acid synthesis, one's attention here tends rather to focus on the possibilities of dephosphorylation of nucleotide as a source of energy for nuclear and cytoplasmic reactions. Thus energy for synthesis of chromosomes and their products may or may not be available accord- ing to whether the nucleotides are structurally isolated from phosphomonesterase ac- tivity by being bound in nucleic acid chains, or whether as a result of a successive action of nuclear depolymerase, and "chromonucleodiesterase," there is mononucleo- tide available for dephosphorylation. Similarly the actual availability of energy lor such cyclic nuclear mechanical processes as chromosome coiling and mitotic move- ment may be dependent upon a cycle of binding and release of mononucleotide from :ts nucleic acid storehouse. SUM MARY 1. Using the histochemical test for alkaline phosphatase reaction in the larval salivary glands of several species of Drosophila, activity \vas found to be present in three main parts of the cell ; the cytoplasm, the nucleolus. and the chromosomes. 2. Phosphatase activity was found rather generally distributed in both the cyto- plasm and the nucleus. Within the larval salivary gland chromosomes, the enzyme activity was localized in those chromosomal regions which are Feulgen positive, and thus corresponds to the regions containing large concentrations ol desoxyribose nucleic acid. 3. Different naturally occurring phosphate bearing substances were used as phos- phatase substrates on mouse tissues, and resulted in demonstration ot three different types, of phosphatase reactions based on the localization of the enzyme activity. a) A general reaction with phosphatase activity located in both nucleus ( nucle- •ilus and chromosomes) and the cytoplasm was present when sodium glycerophos.- phatc and nucleotides were used as substrates. &) Xo phosphatase reaction occurred on polymerized desoxyribose nucleic acid, "nit a specific nuclear reaction (nucleolus and chromosomes) was present when nuclease-depolymerized desoxyribose nucleic acid was used as a substrate. \Vr do not actnallx know the location of the possible substrates for tin- diestenise activity, methods for looali/invj nucleic acids do not, in all likelihood, preserve- any hut the hisji and it may \\vll he that the lower polymers of the sol t used as substrates are not the -ame in distribution as the larger complexes. In such a difference may lie tin- explanation ot the liscrepancies above. 1NTRACELLULAR ALKALINE PHOSPHATASES r) A strong cytoplasmic reaction with slight reaction in the nucleus was present when ribosc nucleic acid was used as a substrate. 4. Subjection to several environmental variables produced little further evidence as to the differences among these three types of localized reactions. 5. The three types of alkaline phosphatase reactions observed were suggested to be due to at least two phosphodiesterases and a phosphomonesterase. 6. The nuclear phosphatase complex and the cytoplasmic phosphatase complex each probably consist of a specific phosphodiesterase, which splits the ester linkage between the phosphate of one nucleotide and the sugar of the neighboring nucleotide, and a phosphomonesterase which splits the second ester linkage and liberates in- organic phosphate. ACKNOWLEDGMENT I wish to express my appreciation and gratitude to Professor H. Burr Steinbach for originally bringing to my attention the cytochemical possibilities of the alkaline phosphatase test ; to Professor Arthur W. Pollister for his advice and encourage- ment throughout the course of the work and for his help in the preparation of the manuscript ; to Dr. Alfred E. Mirsky for advice and essential materials ; and to Dr. Robert Ballentine and Dr. Jack Schultz for many helpful suggestions. LITERATURE CITED BUCK, J., AND A. MELLAND, 1942. Methods for isolating, collecting, and orienting salivary gland chromosomes for diffraction analysis. Jour. Hered., 33 : 173-184. CASPERSSON, T., 1936. Uber den Chemischen Aufbau der Strukturen des Zellkernes. Skand. Arch. Phys., 73: Suppl. 1-151. CASPERSSON, T., 1940. Methods for the determination of the absorption spectra of cell structure. Jour. Roy. Micr. Soc., Ser. 3, 60 : 8-25. DANIELLI, J. F., AND D. G. CATCHESIDE, 1945. Phosphatase on chromosomes-. Nature, 156 : 294. DOUNCE, A., 1943. Enzyme studies on isolated cell nuclei of rat liver. Jour. Biol. Chem., 147 : 685-698. GOMORI, G., 1939. Microtechnical demonstration of phosphatase in tissue sections. Proc. Soc. Ex per. Biol. Med., 42: 23-26. GOMORI, G., 1941. The distribution of phosphatase in normal organs and tissues. Jour. Cell. Comp. Physiol, 17: 71-83. KRUGELIS, E. J., 1942. Cytological demonstration of phosphatase in chromosomes of mouse testes. Jour. Cell. Comp. Physiol., 19: 1-3. KRUGELIS, E. J., 1945. Alkaline phosphatase activity in the salivary gland chromosomes of Drosophila melanogaster. Abstract. Genetics, 30 : 12. LEVENE, P. A., AND L. W. BASS, 1931. Nucleic acids. Chemical Catalog Co., New York. MAZIA, D., 1941. Enzyme studies on chromosomes. Cold Spring Harbor S\m. Quant. Biol., 9: 40-46. MAZIA, D., AND L. JAEGER, 1939. Nuclease action, protease action, and histochemical tests on salivary gland chromosomes of Drosophila. Proc. Nat. Acad. Sci., 25: 456-461. McCARTv, M., 1946. Purification and properties of desoxyribonuclease isolated from beef pan- creas. Jour. Gen. Physiol, 29: 123-139. MIRSKV, A. E., AND A. W. POLLISTER, 1942. Nucleoproteins of cell nuclei. Proc. Nat. Acad. Sci., 28 : 344-352. MOOG, F., 1944. Localizations of alkaline and acid phosphatases in the early embryogenesis of the chick. Biol. Bull., 86: 51-80. POLLISTER, A. W., AND A. E. MIRSKY, 1943. Terminology of nucleic acids. Nature, 152: 692. SCHMIDT, G., AND S. T. THANNHAUSER, 1943. Intestinal phosphatase. Jour. Biol. Chem., 149: 369-385. WILLMER, E. N., 1942. The localization of phosphatase in tissue cultures. Jour. Exp. Biol., 19: 11-13. PHYSIOLOGY OF INSECT DIAPAUSE: THE ROLE OF THE BRAIN IN THE PRODUCTION AND TERMINATION OF PUPAL DORMANCY IN THE GIANT SILKWORM, PLATYSAMIA CECROPIA CARROLL M. WILLIAMS Society of Fellows, Harvard University, Cambridge, Massachusetts The phenomenon of insect diapause presents an exceptionally clear statement of one of the most important problems in biology ; to wit, the nature of the factors that preside over cellular growth and differentiation. For with the onset of diapause and through the workings of internal physiological mechanisms still to be elucidated, growth suddenly comes to a standstill and the animal for months thereafter persists in a genuine state of suspended development. With the termination of diapause the rapid tempo of cellular activity returns and metamorphosis continues where it had left off. Thus whatever may be the inner mechanism for the induction and termi- nation of diapause, it must have the capacity to turn morphogenesis off and on in a most striking way. The study of this phenomenon has not failed to claim the attention of a large array of investigators. For example, even in 1932, Cousin was able to cite 347 papers in a review of the literature. That this extensive literature has so imperfectly advanced our knowledge of diapause is apparently due to the fact that most investigations have been carried out either on muscoid flies, which have a most imperfect and complex diapause, or on the eggs of silkworms and grasshoppers, which are too small to per- mit extensive manipulations of the individual animal. In the present investigation these difficulties were minimized by working on species of insects that possess a wholly characteristic pupal diapause, and, by virtue of weighing up to 8 grams per individual, are among the very largest insects in America. MATERIALS AND METHODS Pupae of the giant silkworm, Platysamia cccropia, were used for the most part, approximately 1200 pupae being studied in a total of 690 experiments. These in- sects were reared from eggs obtained from fertile females ; a lesser number of pupae were secured from dealers. In my experience, this species has never failed to enter into diapause immediately after pupation, thus giving only one brood a year. If the pupae are maintained constantly at room temperature, diapause persists for not less than five months ; if they are placed immediately after pupation at a temperature of 3° to 5° C. and chilled for iy2 months or longer, adult moths emerge about 1 to I1/, months after being returned to room temperature. For this reason the stock of material was divided at the outset into two batches, one being placed and stored at -•> to 5° C. until needed ("chilled pupae"), and the other being maintained at room temperature where, as previously described, diapause persists for at least five months ("diapausing pupae"). 2.5-1 PHYSIOLOGY OF DIAPAUSE 235 In a number of experiments related species of saturniid pupae were used ; namely, Samia ivalkeri, Callosainia promcthea, and Telea polyphctnus. The most important factor facilitating the investigation was the discovery of a method of continuous anesthesia for insects during operative procedures. This method, utilizing carbon dioxide and described by Williams (1946), permitted ex- tensive and prolonged surgical manipulations without any loss of blood or apparent damage to the pupae. Other procedures will be described as encountered in the following discussion. PARABIOTIC EXPERIMENTS We have noted that diapausing pupae, after a period of exposure to low tem- perature, are rendered competent to develop when returned to room temperature, whereas, in contrast, pupae not subjected to chilling remain in diapause for at least five months. With these two types of animals at hand one is therefore in a position to test the fundamental nature of diapause by simply grafting one to the other so that they share the same blood. If diapause results from the presence of some factor inhibiting development, then such parabiotic combinations should fail to develop by virtue of the diapausing pupa distributing this inhibitory factor to the chilled indi- vidual. To the contrary, if diapause results from the absence of some necessary growth factor, both animals should develop, provided the chilled individual can sup- ply double the minimal amount needed by a single animal. In making these preparations a disc of pupal cuticle plus underlying hypodermis was cut from each pupa and the two animals placed together and held thus by the application of melted paraffin around the site. Most of the pupae were joined at the thoracic tergum (Figs. 1 and 3), but occasionally the junction was accomplished at the head or at the tip of the abdomen. Provided that the underlying heart is not injured and that no bubbles of air are trapped in either animal, such combinations are easily established and a high percentage survive. In order to demonstrate that the operation in itself is without effects on dor- mancy, a series of ten diapausing pupae were successfully grafted to diapausing part- ners. Diapause persisted in each of these animals, adult moths being produced only after a minimum of 5y2 months, the usual minimum length of time necessary for the spontaneous termination of diapause at 25° C. To the contrary, when diapausing pupae were joined to previously chilled individuals, the diapause in all viable prepa- rations was terminated. In a series of 15 such combinations the pairs emerged as fully formed, active moths in an average of 41 days. Metamorphosis was complete both externally and internally, the only defect being a failure of the wings to expand after emergence (Fig. 2). This activation was not species- or, indeed, genus- specific, for it was possibte to terminate the diapause of Platysamia cecropia by join- ing them to previously chilled pupae of Telea polyphemus (Figs. 3 and 4). Further- more, sexual differences were without significance, for male pupae had the capacity to induce development of females, and vice versa. A striking feature of all these parabiotic combinations is the fact that the animals invariably grow together so as to be connected by a pedicle, a phenomenon first noted by Crampton (1899) and later by \Vigglesworth (1936) and Bodenstein (1938) in grafting Rrocedures on insects. We shall have occasion subsequently to consider this union more fully, but in the present analysis suffice it to say that the 236 CARROLL M. WILLIAMS epithelial pedicle becomes externally chitinized and by way of its lumen permits a circulation of blood between tbe two animals. In the earlier preparations the blood of the diapausing and chilled pupae in para- biosis was daily propelled to and fro by pressing accordion-like on the abdomen of each pupa alternately. This was later found to be wholly unnecessary, since develop- ment begins just as promptly without such forced mixing. It may also be noted that the completion of adult formation in the previously chilled animal occurs about U/2 days earlier than in the diapausing partner (Fig. 5). This results from a corre- sponding delay in the initiation of development of the diapausing pupa. At all stages in adult differentiation the chilled pupa is therefore approximately IV-j days in ad- vance of the diapausing partner. Thus these initial parabiotic experiments indicate some of the essential features of diapause. In general, they support the proposition that diapause results from the absence of a non-species-specific growth factor that is able to pass in parabiotic preparations from the activated to the dormant individual and evoke the initiation of adult development in the latter also. BRAIN IMPLANTATION INTO DIAPAUSING PUPAE If the termination of diapause is, indeed, accomplished by the action within the previously chilled pupa of a factor necessary for adult development, then it should be possible to demonstrate the organ in which this factor arises. For this purpose, various tissues and organs were removed from chilled pupae and implanted singly into diapausing individuals. When experiments of this sort were carried out, it was found that only one organ in the chilled pupa has the power to evoke development of diapausing pupae and that this organ is the brain itself. When the brain is removed from a chilled pupa and implanted into the head, thorax, or abdomen of a diapausing pupa, the latter is invariably induced to undergo adult development. Furthermore, Platysamia cecropia can be activated by the brains of Samia walker i, Callosamia prorncthea, or Telca polyphemus — and, in fact, as far as the termination of diapause is concerned, there is a lack of species- and genus-specificity of brains among all these Lepidoptera tested. No other organ in the chilled pupa apparently possesses this power. This effect of chilled brains is in marked contrast to that of diapausing brains, EXPLANATION OF PLATE I Approximately Life Size FIGURE 1. Brainless, diapausing pupa of P. cecropia grafted to a chilled pupa of the same species. FIGURE 2. Animals in Figure 1, after adult formation. The two insects have grown to- gether and developed essentially simultaneously. FIGURE 3. Brainless, diapausing pupa of T. polyphemus grafted to a chilled pupa of /'. cecropia. FIGURE 4. Animals in Figure 3, after adult development. FIGURE 5. Parabiosis between two pupae of P. cecropia. removed from pupal cuticle before the completion of adult development. Development of the chilled pupa is IVi! days in advance of that of the brainless, diapausing pupa. FIGURE 6. Adult Cecropia moth produced by a brainless, dipausing Cecropia pupa, whose diapause was terminated by implantation of a brain from a chilled Polyphemus pupa. PHYSIOLOGY OF DIAPAUSE 237 PLATE I ^ V C \KKOLL M. WILLIAMS for tlu- latter fail to terminate dormancy even when as many as eight are implanted into a single pupa. 'Idle lack ot species-specificity of implanted, chilled hrains suggested the possi- bility that the effect might, conceivably, he mediated by an activation of the c-ndoge- nous brain of the diapausing, "host" pu]ia itself. For this reason the vast majority of subsequent experiments were carried out on diapausing pupae from which the brains had been removed. REMOVAL OF BRAIN The operation devised for this purpose is easily performed, as follows. The in- sect, anesthetized with carbon dioxide, is placed on its back and by means of a sharp scalpel a rectangular window of pupal cuticle excised from its face. The underlying. semi-transparent hypodermis is thus exposed. With microscissors the latter is trimmed away, along with the tracheae traversing the operative field. In this proce- dure the frontal ganglion is frequently removed, but this has been found to be incon- sequential. The brain now lies exposed and can be floated up toward the operator by pressing the pupal abdomen so that further dissection is performed in the pupal blood. With finely ground jeweler's forceps the nerves passing laterally to the site of the future adult eyes are grasped and broken. In similar fashion each brain hemisphere is carefully broken loose, in turn, from its tracheal supply, from the nerve passing posteriorly to the corpus allatum and corpus cardiacum and from the circum- esophageal connective. The brain can then be lifted free and examined in insect Kinger's solution.' Taking care to exclude all bubbles, the defect in the pupal chitin is then capped over with a small rectangle of thin, transparent plastic (cut from a plastic cover slip), which is sealed in place with melted paraffin. The operation was originally performed with due regard to surgical asepsis; this was later found to be of little importance since the pupae are apparently not affected by the usual contaminating organisms. The mortality from the procedure is low and one ends up with a brainless pupa possessing a transparent window at its ante- rior end. The defect in the hypodermis is rapidly repaired bv a deposit of blood cells, followed by an ingrowth of cells and of tracheoles along the under surface of the plastic slip. Simultaneously, an intervening, delicate, transparent, "chitinous" lamella is elaborated. These relations permit a detailed study- of the behavior of the hypodermis beneath the window, and, by means of an Ultropaque. cellular activity has been followed tinder the oil immersion objective. It may be noted that this local process of repair occurs just as promptly in diapausing as in previously chilled pupae. Notwithstanding this fact, the process of repair is without overall effects on dormancy. BRAIN IMPLANTATION INTO BRAINLESS DIAPATSINC, PUPAK The removal of the brain from diapausing pupae prior to using the animals ex- perimentally proved to be an exceptionally significant maneuver. For whereas, as 1 This physiological .solution was originally devised hy Kphnissi and IVadle ( In3(» for .stud- ies of I )roso])hila, hut it works equally well for the Lepidoptera used in the present experiments. I am indebted to Dietrich I'.odcnstein for calling my attention to its composition, which is as follows \"a( 1. 7.5 L'in.; l\('l. 0.35 »m. : and CaCl... 0.21 urn., per liter of water. PHYSIOLOGY OF DIAPAUSE 239 we have previously noted, diapausing pupae kept at 25° C. begin to escape spontane- ously from diapause after about five months, no such activation occurs if the brain is removed. Among approximately 400 such pupae there has not been a single case of spontaneous development. It is therefore apparent that by removing the brain the pupa is maintained in permanent diapause. Such pupae remain alive for up to a year and finally die of dessication. Yet at any time during this period the diapause can be terminated by implanting into the brainless pupa the brain of a previously chilled animal (Fig. 6). Data in regard to a series of such pupae are given in Ta- ble I. TABLE I Evocation of Adult Development of Brainless Diapausing Pupae by Implantation of Brains from Chilled Pupae Species of host Species of implanted brain Number of experiments Aver, time for adult emergence P. cecropia P. cecropia 16 35 days P. cecropia T. polyphemus 2 89 P. cecropia S. walkeri 3 72 P. cecropia C. promethea 2 63 T. polyphemus T. polyphemus 2 64 T. polyphemus S. walkeri 2 28 Manifestly, these experiments demonstrate that the termination of diapause re- quires the presence of an activated brain, in the absence of which adult development of these insects has not been observed. The conclusion is also self-evident that the termination of dormancy after diapausing pupae have been chilled results from the action of low temperature in rendering the brain able to evoke adult development. The other tissues in the diapausing pupa do not require such exposure to cold, for they are rapidly activated by implanting a brain which, alone, has been chilled. This fact can also be readily demonstrated by removing strips of integument from diapausing pupae and implanting them into previously chilled pupae. Such diapaus- ing tissues develop simultaneously with the host : the pupal cuticle is delaminated and a normally chitinized, adult cuticle, complete with scales and hairs, is found in the implant, in exactly the same fashion as described by Piepho (1938a and b) and Kiihn and Piepho (1940) in studies of other aspects of insect metamorphosis. Thus the effect of low temperatures in facilitating escape from diapause can be explained solely on the basis of its effect on the brain. PARABIOTIC EXPERIMENTS ON BRAINLESS DIAPAUSING PUPAE As soon as the brain was definitely shown to be the source of the factor terminat- ing diapause, ten more parabiotic combinations were prepared, but this time uniting brainless diapausing pupae with chilled individuals. Identical results were obtained : the two pupae in each combination grew" together by a chitinized, epithelial, blood- filled pedicle and after an average of 44 days emerged as normal, active, adult moths. BRAIN REMOVAL FROM PREVIOUSLY CHILLED PUPAE Further information concerning the action of the brain in terminating diapause can be gained from a consideration of the behavior of chilled pupae. We have previ- 240 CARROLL M. WILLIAMS ously noted that these animals undergo no apparent development as long as they are maintained at the low temperature, but that within 1 to U/2 months after return to 25° C. the adult moth has fully formed and emerges. It is therefore worthy of note that if the brain of such chilled pupae is removed as soon as the insect is returned to the warm temperature, adult development never occurs and, in the same fashion as described for brainless diapausing pupae, dormancy persists indefinitely until the animal finally dies of dessication. Yet development can at any time be evoked by merely implanting into the head, thorax, or abdomen a brain obtained from another chilled pupa. This phenomenon was studied more fully as follows. A series of thirty previ- ously chilled pupae was placed at 25° C. and every few days the brains from several of these insects were removed and implanted into brainless diapausing pupae. The results may be summarized most briefly by saying that when the brain is removed within approximately the first 11 days, the brainless donors never show any develop- ment ; such brains, in turn, evoke the development of brainless, diapausing pupae. In contrast, if the brain is removed from previously chilled pupae after approxi- mately 17 days at 25° C., development continues to produce normal, brainless adults and the removed brains are without effect in terminating the dormancy of brainless, diapausing pupae. These experiments have been repeated on a large scale, with special attention to the effect of brain removal during the critical period of 11 to 17 days. These more detailed studies were facilitated by establishing, at the outset, a transparent, plastic, facial window in each chilled pupa so that the behavior of the underlying hypodermis could be observed. It was at once apparent that the critical period, 11 to 17 days, was, in a sense, a statistical artifact, since, during this period, each individual achieves threshold activation during an extremely short interval, not exceeding a few hours. The actual critical period for each pupa is signaled by the initiation of hypodermal retraction from the overlying, facial chitin. Prior to this point, removal of the brain prevents development, and such brains evoke the development of brainless, diapaus- ing pupae after an additional latent period of approximately three weeks. The mo- ment hypodermal retraction is initiated, the brain can be dispensed with and such brains, when tested, are inactive. Thus it is apparent that diapause persists even in chilled pupae until the latter have been exposed to a developmental temperature for an average of two weeks. Consequently, the activation of the pupal brain during exposure to low temperature must be conceived in terms of some physical or chemical alteration in the brain sub- stance whereby the latter is rendered competent to produce or release its stimulating factor during subsequent exposure to a developmental temperature The brain's action is then exerted and, thereafter, metamorphosis can proceed independent of its further participation. ROLE OF THE CORPORA ALLATA All the evidence so far considered reveals the brain as the organ of paramount importance in engendering and terminating diapause. Thus diapause in these spe- cies appears to result from the absence of a factor necessary for adult development, rather than from the presence of an inhibitory factor. The possibility remained, however, that the failure of the brain to exert its effect and the consequent onset PHYSIOLOGY OF DIAPAUSE 241 of diapause might, in turn, be due to inhibition arising elsewhere in the organism. The corpora allata were deemed the most likely source of such hypothetical inhibi- tion and for this reason their significance in the production of diapause was studied. As originally demonstrated by Bounhiol (1938) and subsequently confirmed by Piepho (1940; 1941), the corpora allata of Lepidoptera specifically inhibit pupation during the larval instars and thus oppose the activation of the presumptive imaginal tissues. This finding seemed so significant that comparable experiments were car- ried out on the caterpillars of Platysamia cecropia and Tclea polyphcinns. Although the removal of the corpora allata from caterpillars is a difficult procedure, it was ac- complished in a sufficient number of immature (fourth instar) larvae to demonstrate that precocious pupation, indeed, results therefrom, the usual fifth instar being omitted. Furthermore, there is convincing evidence that the function of the corpora allata in inhibiting the imaginal discs disappears during the final larval instar and pupation then ensues. These findings have been considered in some detail for, al- though they concern pupation rather than adult differentiation, it is easy to see the importance of demonstrating whether, in potentially diapausing insects, the corpora allata once again inhibit the imaginal tissues, or the brain, and thus participate in the induction of diapause. A series of experiments was therefore performed in which the corpora allata were removed (by a frontal approach) from diapausing pupae. Such pupae 2 invariably continued to diapause normally, and the removed corpora allata when implanted into previously chilled pupae were without effect in inhibiting adult development. As many as six corpora allata have been implanted into a single chilled pupa without retarding metamorphosis. Similar negative results were obtained in regard to all other organs studied as a possible source of some inhibitory factor. For example, the diapausing brain itself is without inhibitory properties, for as many as six such brains have been implanted into a single chilled pupa without producing diapause. This is also true for the subesophageal ganglion, thoracic ganglion, gonads, imaginal discs, and strips of in- tegument. Although it cannot be denied that inhibitory factors may play a role in the pro- duction of diapause, the sum total of available evidence offers nothing to support this proposition. The brain itself remains the key to the production and termination of diapause in the species studied. DISCUSSION The role of the brain in terminating diapause, demonstrated for the first time in the present investigation, can to advantage be compared with its other functions in insect metamorphosis. Thus, in the bug, Rhodnius, the brain is necessary for moulting (Wigglesworth, 1940) and in the Lepidoptera it is how well established that the brain is also required for pupation (Kopec, 1922 ; Caspari and Plagge, 1935 ; - It may be noted that the allatadectomized pupae ultimately escaped from diapause after the usual minimum period of 5M> months at 25° C. The resulting moths were wholly normal in all respects and could be induced to mate and lay eggs, which, in turn, were fertile. These findings apparently deny a participation of the corpora allata in the egg production of these species, a function described for them in certain other Orders of insects (Wigglesworth, 1936; Pfeiffer, 1939). 242 CARROLL M. WILLIAMS Kuhn and Piepho, 1936; Bounhiol, 1938; Piepho, 1938a; Plagge, 1938). A sur- prising fact is that a role of the brain in imaginal differentiation has been specifically denied by all of these investigators of lepidopteran metamorphosis. The important point is that this conclusion was, without exception, based on studies of continu- ous, non-diapausing development. For such insects there can be little doubt that adult formation ensues even though the brain is removed from mature caterpillars in the last instar (i.e., after the "critical period" for pupation). The existence of this striking difference between continuous and diapausing development has been pointed out previously (Williams, 1942). In the present investigation we have seen that all the evidence supports the theory that diapause results from an interruption in the normal processes of adult development. This point of view suggests that the brain is also necessary for evoking adult development in non-diapausing pupae. In non-diapausing indi- viduals the brain may be viewed as having achieved its full developmental function precociously prior to pupation, whereas, in potentially diapausing animals, the brain first controls pupation and then months later after pupation it controls adult formation. The action of the brain in terminating diapause in these saturniid pupae poses an additional problem of even greater interest ; namely, the nature of the factor arising in the brain which so spectacularly evokes in dormant tissues a veritable flood of cellular activity. This problem will be considered in a subsequent communication. SUMMARY 1. The physiological control of pupal diapause has been studied on a total of ap- proximately 1200 pupae of the giant silkworms, Platysaniia cccrof>ia, Tclca poly- phcmus, Samia ivalkeri, and Callosaniia promethea. 2. The dormancy of diapausing pupae can be terminated readily by grafting them to activated (previously chilled) pupae. The two animals in each parabiotic com- bination grow together and some factor necessary for adult development passes from the activated to the dormant animal so that both develop simultaneously. This fac- tor is not species- or genus-specific. 3. By implantation experiments the source of the factor terminating diapause is shown to be the brain and in this function a lack of species- and genus-specificity of brains is demonstrated. 4. In these species the well-known action of low temperatures in facilitating es- cape from diapause results from the effect of cold in rendering the brain competent to terminate domancy. Actual termination of dormancy is accomplished only after the previously chilled brain has been exposed to a developmental temperature for an average of two weeks. The earliest indications of adult development then become evident and the brain, thereafter, is no longer required for the completion of metamorphosis. 5. Therefore, the effect of low temperatures on the brain must consist in some physical or chemical alteration in its substance whereby the latter is rendered com- petent to produce or release an imaginal-differentiation factor after return to a de- velopmental temperature. '>. No evidence was found to support the theory that diapause results from the PHYSIOLOGY OF DIAPAUSE 243 presence of inhibitory factors. In this regard, the functions <>f the corpora allata are considered in some detail. 7. It is concluded that diapause in these species results from an interruption in the normal processes of development by virtue of a failure of the brain to supply a non-species-specific factor necessary for adult differentiation. Diapause is ter- minated when this factor is provided. 8. The significance of the brain in the development of diapausing pupae is con- sidered in relation to its other functions, as reported in the literature. Notwith- standing a certain amount of evidence to the contrary, it is probable that even in the absence of diapause the brain plays a vital role in adult formation. LITERATURE CITED BODENSTEIN. D., 1938. Untersuchungen zum Metamorphose problem. II. Entwicklungsrela- tionen in verschmolzenen Puppenteilen. Arch. f. Entwmech. d. Oi'(/an., 137: 636-660. BOUXHIOL. J. J., 1938. Recherches experimentales sur le determinisme de la metamorphose chez les Lepidopteres. Bull, Biol. de I-'rancc ct dc Bclyiquc, Suppl., 24: 1-199. CASPARI, E., AND E. PLAGGE, 1935. Yersuche zur Physiologie der Verpuppung von Schmetter- lingsraupen ( \'orl. Mitt.). Naturwiss., 23 : 751. COUSIN. G., 1932. fetude experimental de la diapause des insectes. Bull. Bid/, dc France ct de Hchnquc. Suppl., 15: 1-341. CRAMPTON, H. E.. 1899. An experimental study upon Lepidoptera. Arch. /. Entwmech. d. Organ., 9: 293-318. EPHRUSSI, B., AND G. W. BEADLE, 1936. A technique of transplantation for Drosophila. Anicr. Nat., 70 : 218-225. KOPEC, S., 1922. Studies on the necessity of the brain for the inception of insect metamorphosis. Biol. Bull., 42 : 323-342. KivHX, A., AND H. PIEPHO, 1936. Ueber hormonale Wirkungen bei der Verpuppung der Schmet- terlinge. Ges. H'iss. Gottlnycn, Nachr. Biol., 2: 141-154. KUHN, A., AND H. PIEPHO, 1940. Uber die Ausbildung der Schuppen in Hauttransplantaten von Schmetterlingen. Biol. Zcnthl.. 60: 1-22. PFEIFFER, I. \Y.. 1939. Experimental study of the function of the corpora allata in the grass- hopper, Melanoplus differentialis. Jour. E.vp. Zool.. 82: 439-461. PIEPHO, H., 1938a. YVachstum uncl totale Metamorphose an Hautimplantaten bei der Wachs- motte Galleria mellonella L. Biol. Zcnthl., 58: 356-366. PIEPHO, H., 1938b. Uber die Auslosung der Raupenhautung, Verpuppung und Imaginalentwick- lung an Hautimplantaten von Schmetterlingen. Biol. Zcnthl., 58: 481-495. PIEPHO, H., 1940. Uber die Hemmung der Yerpuppung durch Corpora allata. Untersuchungen an der Wachsmotte Galleria mellonella L. Biol. Zcnthl., 60: 367-393. PIEPHO, H., 1941. Untersuchungen zur Entwicklungsphysiologie der Insektenmetamorphose. Uber die Puppenhaiitung der Wachsmotte Galleria mellonella L. Arch. f. Entwmech. d. Organ., 141 : 500-583. PLAGGE, E., 1938. Weitere Untersuchungen iiber das Verpuppungshormon bei Schmetterlingen. Biol. Zcntbl., 58 : 1-12. WIGGLESWORTH, V. B., 1936. The function of the corpus allatum in the growth and reproduction of Rhodnius prolixus (Hemiptera). Quart. Jour. Micr. Sci.. 79: 91-121. WIGGLESWORTH, V. B., 1940. The determination of characters at metamorphosis in Rhodnius prolixus (Hemiptera). Jour. E.rp. Biol.. 17: 201-222. WILLIAMS, C. M., 1942. The effects of temperature gradients on the pupal-adult transformation of silkworms. Biol. Bull.. 82 : 347-355. WILLIAMS. C. M., 1946. Continuous anesthesia for insects. Science, 103: 57. FEEDING OF OYSTERS IX RELATION TO TIDAL STAGES AND TO PERIODS OF LIGHT AND DARKNESS \ HTOR L. LOOSANOFF AND CHARLES A. NOMEJKO 1-isli and ll'ilillifc \rr:'/ per cent of the oysters collected during the ebbing tide possessed empty stomachs, while among the individuals dredged during the flood 10 per cent showed a complete absence of food. It is significant that among the oysters collected during the low water stage the per cent of the individuals containing a large quantity of food was higher, and that of the animals with empty stomachs lower than at many other stages TABLE I Relative quantities of food in stomachs of 1000 oysters collected at different tidal stages in Long Island Sound, June, July, and August 1945 Stage of tide Oysters examined Quantity of food Per cent Large Small Absent Large Small Absent Flood 1st hour 41 31 6 4 75 15 10 2nd hour 60 51 2 7 85 3 12 3rd hour 85 58 9 18 68 11 21 4th hour 75 66 5 4 88 7 5 5th hour 85 72 8 5 85 9 6 High water 110 91 9 10 83 8 9 Total 456 369 39 48 81 9 10 Ebb 1st hour 80 66 11 3 82 14 4 2nd hour 80 69 7 4 86 9 5 3rd hour 120 102 10 8 85 8 7 4th hour 90 76 10 4 85 11 4 5th hour 107 84 13 10 79 12 9 Low water 67 59 6 2 88 9 3 Total 544 456 57 31 84 10 6 Grand total 1000 825 96 79 82 10 8 of the tide. While this observation cannot be interpreted as definite proof that oysters feed most efficiently just prior and during the low water stage, it shows, nevertheless, that they do not cease, or noticeably decrease, their feeding activities during late ebb. Additional observations on the relative quantities of food in the stomachs of the oysters at the end of the flood and during the entire period of ebb were made in Long Island Sound on August 21, 1945. On that day a station was chosen in 20 feet of water on one of the beds planted with 4-year-old mollusks. The location of the station was designated by a special buoy. The samples were collected at hourly intervals beginning one hour prior to the high water stage (Table II). Al- together eight samples, each composed of 20 oysters, were collected and examined. FEEDING OF OYSTERS 247 The salinity of the water ranged from 27.31 p.p.t. soon after high water to 25.45 p.p.t. at low water. The results of the examination again showed that in the majority of the oysters the stomachs contained large quantities of food during all stages of the ebb (Table II). The observations also indicated that there was no decrease in the number of oysters with food-filled stomachs parallel with the falling of the tide. On the con- trary, the largest number of oysters containing large quantities of food was found only one hour prior to low water. It is significant that among the oysters examined during the last three hours of the ebb not a single individual was found with an empty stomach. While the pres- ence of food in the oysters dredged during -the early stages of the outgoing tide could be possibly explained by assuming that this food was ingested during the last TABLE II Relative quantities of food in stomachs of 160 oysters collected at hourly intervals during last hour of flood and during ebb from a station established in 20 feet of water in Long Island Sound, August 21, 1945. Each sample composed of 20 oysters. Quantity of food Stage of tide Temperature °C. Large Small Absent Flood 5th hour 21.2 11 8 1 High water 20.9 17 3 0 Ebb 1st hour 21.1 14 6 0 2nd hour 21.0 19 1 0 3rd hour 20.9 17 2 1 4th hour 21.1 19 1 0 5th hour 20.6 20 0 0 Low water 20.4 16 4 0 Total 133 25 2 hour of flood, such an explanation cannot be offered for the presence of food in the oysters examined from three to six hours after the high water stage. Our observa- tions performed under laboratory conditions on oysters kept in water of 20.0° C. showed that these mollusks pass the particles of food through their entire digestive system from 1 hour 20 minutes to approximately 2 hours and 30 minutes. There- fore, it seems rather improbable that the food found in the oysters during the latter part of the ebb was that ingested during the late stage of the flood, four to six hours prior to examination. In general, the observations made on 160 oysters on August 21, 1945, showed that the oysters of Long Island Sound fed very actively during the ebb, and that the relative quantities of food found in their stomachs during that period were at least equal to or even exceeding those recorded during the last hour of the preceding flood. Although observations in Long Island Sound have demonstrated that there was no correlation betwen the stages of the tide and the quantities of food found in the 248 \ It TOR 1, LOOSANOFF AND CHAKLKS A. XOMFJKO stomachs of the ovsters. it was desirable to supplement the data already available with additional observations on the oysters living under ecological conditions rather different than those of Long Island Sound proper. An area in Milford Harbor near the dock used for laboratory needs, was chosen for these observations. A large number of oysters living on the bottom near the dock made such an arrange- ment especially convenient. Milford Harbor was selected because it was a typical example of a small partially inclosed body of water where extensive natural beds had existed. In recent times many of the beds were destroyed by overfishing, and the profile of the bottom was markedly changed by the dredging of a wide and deep channel. Nevertheless, the oysters quickly reestablished themselves in more shallow sections of the Harbor, and at present are quite common. A good setting of oysters regularly occurs in the Harbor, indicating that the conditions are favorable for the propagation of these mollusks. Usually changes in the temperature and salinity of the Harbor water during the tidal cycle are more pronounced than in Long Island Sound proper, where both these factors remain very steady (Loosanoff and Engle, 1940). The observations consisted in examining the stomach content of the oysters at hourly intervals throughout a 24-hour period. The samples, each composed of six individuals, were suspended one day prior to the beginning of the examination in bas- kets made of 2-inch mesh poultry wire. All the baskets were kept at the same depth, namely, one foot below the mean low water line. They were separated from each other by a distance of approximately one foot and. therefore, the removal of any of the baskets did not disturb the oysters of the other containers. The experiment continued from 7:30 A.M. of July 27 until 8:00 A.M. of July 28, covering three low and two high water stages, and including periods of light and darkness (Table III). During this period the temperature of the water ranged from 22.0 to 25.0° C., the salinity fluctuated between 22.68 and 28.44 parts per thousand, and the pH from 7.7 to 8.7.' The data obtained indicated that during the two flood and two ebb periods cov- ered, the majority of the oysters contained large quantities of food. Of the total number of 150 oysters examined 86 per cent belonged to that group. This figure closely approached that for the Long Island Sound oysters where 82 per cent were found to possess large quantities of food (Table I). Only 4 per cent of the Milford Harbor oysters were found with empty stomachs, the figure being too low to suggest that large groups of the oyster population ceased feeding for appreciatively long inter- vals during the period of observation. More detailed studies of the data given in Table III do not offer sharply defined evidence which would lead to the conclusion that the oysters collected during the flood contained more food than those collected at ebb, or vice versa. Although it is true that all but one oyster collected during the second flood period, extending from S:40 i-.M. to 1 :58 A.M., contained large quantities of food, virtually the same observa- tions were made during the preceding period of ebb when 32 out of 36 oysters showed full .stomachs. In each case only one oyster with an empty stomach was found, ll the condition of the oysters during the first flood period (8:30 A.M. to 1 :30 P.M.) is compared with that of the oysters examined during the second ebb period (3:00 A.M. to 8:00 A.M.), it will be found that in each case 28 individuals showed large, and 6 showed small quantities of food, while 2 oysters were with empty stomachs. FEEDING OF OYSTERS 240 a •ta •S a a '•^ ^ f^-4 a "5 a VJ a — •§ ° — "XJ , *j e rt "« cfl OOOOOO O O <>1 T) 0 a D M l_i to OvO^O^OOO i/l >O ^ "I "5 "3 O (N o oo T t-.ir>ir)ir5iO"; uouo-^1^- — C c X 060606060006 060606060606 >, •*-> J C ^ O O "^ r~ o*^-'^-r<^r^-'~~ r^l^OOLOfN OfN"")"") tNCSfNCNfNr^l (^ICvlfSfNCNCS t, J 0 Loo\<-oin-HO O'^^otN-^1 0. C u 0 IH D 4-1 « »fir<3r» 3 ssssss ssssss 1^ c < 5 j fc&^a^fc'<< <<•<<<< toio'^>oiooo oooooo *-t<"t'"***-*'m OOOOOO t u OOO^O'-'fv]^^ rOT^lOvOf-OO < H (4 ( < 1 C ; j 3 j 10 d j Q ^ ^ -a -u j= j= .S«w « -^ -o j= j= | EO C U -M -w HH cnCu*-'*-'X ~HCNr'5T}-lH >— i CN ""I "* "3 >— 1 *t3 4—> O H 4J § U O [J t^ t^ O O O — CN r^r^rO 4- CO -*O^LO-*'iO'sO u^-HO\rocM'^5 >O t^iO^O-i'O'^ vO "O <^ >— i i/l •- 1 R C/] Q. d r-j t^vOt-~oor-^O t~-oO\OvO'^'* (N CNfNtNCNfNtN CNCNCNCSCNC-! 1-1 a u o CS OfN^O^--^ fNO'f'^O"!^ p a H U u a a •^i •*? rf) r**. rr> rf*. rf) rorO"0 a s ssssss ssssss t 3 o < <<<<£.— i. cu s. i. a. su O OOOOOO OO'OOOO f^s ^^r^r^r^r^; ^r^r^r^r^ ^ g f^ OO^O'-'fN^- CNr<-3'<1f1OvOf- V a u 0 u BO rt y) fe S S3 w^^, = U U Ui nl L- 53 *^ ** ** *i ^* w « "™ _— IZ BM *^ "^I"^-^"^?!! l|l||l|||llll JU-j'g-S-S-S|'J1S-§-S-S-S3 J ^^rsjr^-t-ioi-L, ^^cvirO'tLO-J 250 VICTOR L. LOOSANOFF AND CHARLES A. NOMEJKO Perhaps it should he emphasized that among the 18 animals collected at the three low water stages (Tahle III) all hut one oyster had large quantities of food in their stomachs. Studies of the same nature as those made on July 27 and 28 were performed again on August 9. However, in the latter case ohservations were made only during one flood and one ehb stage covering a period of ahout 12 hours. The results of the observations are incorporated in Tahle IV together with data on the water tempera- ture, salinity and pH recorded at the collection of each sample. The results showed that more oysters with large quantities of food were found in the group collected TABLE IV Relative quantities of food in oysters examined at hourly intervals during a 12-hour period in Milford Harbor on August 9, 1945. Each sample composed of 6 oysters. Temperature, salinity and pH are indicated for each stage of the tide. Quantity of food Stage of tide Time of day Temperature °C. Salinity p.p.t. pH Large Small Absent Flood 1st hour 7:55 A.M. 22.2 24.99 7.9 0 5 1 2nd hour 8:55 A.M. 21.9 25.72 7.9 1 5 0 3rd hour 9:55 A.M. 21.6 25.72 7.9 3 3 0 4th hour 10:55 A.M. 21.6 25.99 79 3 1 2 5th hour 11:55 A M 21.8 25.44 7.9 6 0 0 High water 12:55 P.M. 21.7 25.72 7.9 5 1 0 Total 18 15 3 Ebb 1st hour 1:55 P.M. 22.0 25.72 8.0 4 2 0 2nd hour 2:55 P.M. 21.8 25.72 8.1 3 2 1 3rd hour 3:55 P.M. 22.0 25.72 8.1 6 0 0 4th hour 4:55 P.M. 23.9 25.44 8.1 6 0 0 5th hour 5:55 P.M. 24.4 23.84 8.0 4 2 0 Low water 7:05 P.M. 25.1 21.44 8.0 6 0 0 Total 29 6 1 during the ebb than among the animals examined during the flood. However, no significant difference was noted between the two groups in the number of oysters with empty stomachs. In summarizing all our observations on the relative quantities of food present in the stomachs of the oysters during different tidal stages the conclusion may be formed that, as far as ingestion of food is concerned, the oysters of Long Island Sound and Milford Harbor do not show a definite preference either to ebb or to flood. During either stage the predominating majority of the oysters contained large quantities of food, whereas individuals with empty stomachs were found only occasionally. The data obtained fail to lend any support to the unqualified opinion that the American oyster, O. virf/inica. takes relatively little food on the ebb tide. FEEDING OF OYSTERS 251 Rale of zn'atcr pnmpiny in relation to tide To determine the rate of water pumping of oysters and, therefore, the efficiency of their feeding at different stages of the tide an apparatus was constructed which permitted the measurement of the actual quantities of water passing through an oyster (Fig. 1 ) . This apparatus was installed on the edge of the clock situated along the western shore of Milford Harbor where a swift tidal current flowed unobstruct- /////////////////////A SURFACE 77 BOTTOM FIGURE 1. Diagram of the apparatus used in determining the quantities of water pumped by the experimental oysters. Description in text. edly. The water supply was obtained through the hose, A, being pumped by the pump, B, into the storage barrel, C. The intake end of the hose, protected by a screen of large mesh always remained in the same position, six inches above the bottom. A constant water level in the barrel, C, was maintained by allowing the excess water to escape through the overflow outlet, D, to which a length of hose was attached. The capacity of the pump was such that the water in the storage barrel VKTOK L. LOOSANOFF AND CIIARLKS A. NOMEJKo was renewed every 4 or 5 minutes. Thus, the experimental oysters were receiving water, which was changing parallel with the changes of the tide. The water was fed to the experimental oysters through the tube, E, provided with a flow-adjusting cock, F. The constant level oyster chamber, G, contained the oys- ter, H , the excurrent side of which was covered with a rubber, cone-shaped apron which conducted the water pumped by the oyster into the smaller chamber, Gl. Moore (1908) and Nelson (1936) were the first to apply the rubber apron, while Galtsoff (1926) devised and began to use the chamber. A string glued to the upper shell of the oyster was attached to the counter-balanced lever, /, which recorded every movement of the shell on the kymograph, /. The water pumped by the oyster into the chamber, (7,, overflowed through the glass tube standpipe into the tripping vessel, K, of known capacity. When the vessel was filled with the water pumped by the oyster it tripped over, emptying its contents, and at the same time striking a string attached to the lever, /,, which made a mark- on the kymograph, /. Thus, each tripping was recorded, and because the capacity of the vessel was known, the quantity of water pumped by the oyster during certain intervals could be ascertained easily. The excess water entering the chamber, G, but not utilized by the oyster flowed out through the outlet, L. The part of the ex- perimental apparatus containing the oyster chambers and the kymograph was kept in a small shed to protect the oyster from the effects of the sun and from possible disturbances caused by the wind. The oysters were placed in the apparatus usually from one to two hours before high or low water. This time was allotted to the oysters to open their shells and to begin pumping water at a normal rate. This introductory period was not included in the analysis of the pumping activities of the oysters. At the end of the introductory period, which usually coincided with the high or low water stage, observations were carried on for 12 or 13 hours, covering one ebb and one flood stage. In this manner records of 27 oysters were obtained. Fortu- nately, in almost all cases the oysters remained open continuously throughout the period of exposure. The experiments were conducted at temperatures ranging from 19.1 to 25.8° C., a range considered very favorable for the pumping activities of oys- ters (Galtsoff, 1928). In analyzing the kymograph records obtained in the course of these studies it was found convenient to divide each flood or ebb period into six hourly subperiods. However, since the ebb and flood periods of Milford Harbor are not usually of ex- actly 6-hour duration, the data for the last hour of each stage had to be arrived at by determining the quantity of water pumped by the oyster from the end of the fifth hour until the change of the tide and then calculated for a 60-minute period. Analysis of the data showed that, in general, the rate of pumping of the oysters during the flood stages was somewhat slower than that during the ebbing periods (Table V). However, because of significant differences in the rate of pumping .shown by individual oysters within the same hour of a tidal period, and after con- sidering all the aspects of the data secured the opinion was formed that the oysters of Milford Harbor feed actively at all stages of the tide, and that the rate of feeding during the ebb is at least equal to and sometimes may be even more rapid than (hiring the flood. In Milford I I arbor the strongest tidal currents occur in the middle of the period between the high and low water stages. This period, therefore, corresponds to the FEEDING OF OYSTERS 253 third and fourth hours of each stage. There is no evidence, nevertheless, that during this period the pumping of the oysters was more energetic than during the preceding or successive periods (Table V). It is of interest to note, however, that during the two last hours of the ebb the rate of pumping was somewhat accelerated. The rapid rate at which many of the experimental oysters pumped water during ebb is well demonstrated in the photograph of the kymograph record showing the rate of pumping and shell movement of one of the experimental animals (Fig, 2). The period of observation lasted from 9:34 A.M. until 10:13 P.M., September 12, covering one complete flood and ebb period. Each vertical line of the lower record was made by the tripping vessel (Fig. 1), the capacity of which was 255 cc. Ex- amination of the lower part of the record, which, incidentally, was made with the help of a dissecting microscope, a method always employed when the pumping was rapid and the marks on the kymograph were made close to each other, provided the following information : the minimum quantity of water pumped by the oyster during a single hourly period of the flood was 4080 cc.. and the maximum, 20.655 cc. The TABLE V Mean of quantities of water (in cubic centimeters) pumped by oysters during each hour of the flood and ebb periods. The data are based on the kymograph records of 27 oysters, Milford Harbor, Summer of 1945. Stage of tide Mean Stage of tide Mean Flood Ebb 1st hour 15,952 ±1,2 14 1st hour 1 5,384 ± 787 2nd hour 14,411±1,578 2nd hour 15,962± 830 3rd hour 12,87S±1,427 3rd hour 16,186±1,021 4th hour 12.470± 1,389 4th hour 17,458±1,195 5th hour 13,438±1,130 5th hour 17,9 16 ±1,063 6th hour 14,131±1,040 6th hour 17,577± 944 average hourly rate of pumping for the entire flood period was 13,260 cc. For the ebb period these figures were 15,810, 23,715, and 20.244 cc. respectively. Examination of other records of the same series revealed that during the ebb some of the oysters averaged from 25 to 27 thousand cc. per hour, while the maxi- mum rate of pumping in some instances ranged from 31 to 34 thousand cc. per hour. Having in possession a large number of kymograph records of this nature one may well be inclined to disagree with the opinion that the American oyster is relatively inactive during the outgoing stage of the tide. In connection with this discussion an interesting deduction can be made concern- ing the efficiency of the pumping mechanism of oysters. The average weight of the oyster meat removed from the shell the length of which is 4 inches is approximately 20 grams and its volume is usually not more than 20 cc. This organism, neverthe- less, is capable of pumping 30 thousand cc. or more of sea water per hour. In other words, the volume of water passing through the oyster gills in one hour may be more than 1500 times greater than the volume of the oyster's body, a fact well attest- ing the efficacy of the feeding mechanism of this mollusk. 1 •! VM TOR L. LOOSANOFF AND CHARLES A. NOMEJKO 9. f rt f s'- • bxj "+J *tr' 1"°^ a| I M- "~ •„ O •— • 1^ u tj rt -*-1 • J > 2 £ ' "2^1 rt -g ~ ^ W "^ ^— •— iy~i rt 3 . ' O Q, as a; O > -a u a! rt O 00 rt 1§? 2 « be +-• a rsj ^ M-t W *O o P C be o a ON S r- -o ^ Ji"§-c3 •SW ~oo —- 1 <« G • & £ rt IT) O C 1- J2 u ^ ON en «*- — eg '-^ "5 'O *>. **"! ^ ^^§ t -00 c o rt • > o -o .; rs o £,' o w p w s •=• s^ O 'c« K - ^T3 FEEDING OF OYSTERS 259 •S ?P '?> ? O< 5 § "O ** t/1 f> OC QJ IS! T3 J^ -ass 2 «j o h "v a & 'i 5 " & 3 «*H r" o o h 260 VICTOR L. LOOS \\ol-F AND CHARLES A. NOMEJKO Tin- oysters we're kept in the apparatus for approximately *' hours ( Fig. 5). During this time the mollusks were exposed to two periods of light and two of dark- ness. However, the initial period, of approximately 3 hours, was not included in the analysis, as this time was given to the oysters to become accustomed to the experi- mental condition. To equalize the data half of the experiments began with a period of light, and the other with that of darkness. Altogether 30 records of 10 different oysters were obtained. Analysis of the data showed that the mean hourly rate of ] jumping during the periods of light was 5350 ± 445 cc. For the periods of dark- ness this figure was 5036 ± 385. Thus, the rate of pumping during periods of light closely resembles that during periods of darkness. Finally, the data on the same subject were taken from our 64 daily records of the shell movements of oysters. Analysis of these data showed that on an average the shells of the oysters remained open 94.4 per cent of the total time during daylight, and 93.8 per cent during the periods of darkness. The difference of less than one per cent is not considered as significant in this case, and, therefore, it may be con- cluded that in our experiments no correlation was found between the periods of closure of the shells and darkness. These results are in agreement with those of Galtsoff (1928) who in his article disagrees with Nelson's (1921, 1923) conclusion that the period of darkness, between 11 :00 P.M. and 4:30 A.M., should be considered as a time of rest for oysters. DISCUSSION The advocates of the opinion that oysters considerably reduce their feeding ac- tivities during the periods of ebb failed to suggest in their publications any factors which could be considered as responsible for the change in the behavior of the mol- lusks. It could be easily understood that in some areas, where the periods preced- ing and coinciding with the low water stage are accompanied by distinctly unfavor- able changes in the environment, the oysters would temporarily slow down or even cease feeding. For example, a sharp decrease in salinity could compel the oysters to be relatively inactive. However, according to Nelson (1921) the reduction in salinity was not the cause. This conclusion is based upon his statement that in Huey's Creek, where the experiments were conducted, the periods of complete ces- sation or of the commencement of feeding, although showing a definite correlation with the stage of the tide, occurred independently of the changes in the density of the water, because such changes were usually of small magnitude. Changes in the turbidity and temperature of the water were also considered as unimportant by Nelson (1923), who concluded that "The rate of filtration of water during any given period of time, as deducted from the rapidity and extent of ejec- tions of accumulated sediment from the mantle cavity, may vary widely independ- ently of the temperature and the turbidity of the water." In the same article Nel- son also stated that "No correlation could be shown between the food content of the water and the periods of inactivity of the oyster." All these conclusions were based upon the experimental data first reported by Nelson in 1921. Thus, accord- ing to that author, neither changes in salinity or temperature, nor changes in the turbidity or quantity of food present in the water affected the rate of feeding of oysters. Yet, because of some undetermined factors these mollusks fed much less actively during the ebb stages. FEEDING OF OYSTERS 261 In discussing Nelson's (1921) work it is necessary to mention that his experi- ments were devised to study the shell movements of the oysters, but not the rate at which these mollusks were filtering water through the gills, i.e., feeding. Only the shell movements of the oysters were recorded on the kymograph, while no data were obtained on the quantities of water pumped by the oysters during the different tidal stages. Obviously, no definite conclusions could be formed concerning the latter subject because of the almost complete lack of experimental evidence regarding this matter. Nelson's method of interpreting the data should also be mentioned. In analyz- ing his material on shell movement of the oysters Nelson (1921) takes into con- sideration only the numbers of openings and closures of the shells during the dif- ferent tidal periods. This method has already been criticized by Galtsoff (1928) who stated that "The examination of the number of closures and openings occurring during a given period of time does not convey a true idea of the activity or inactivity of the oyster. A better understanding can be gained by counting the number of hours the oyster was closed or open during a given period of a day." Obviously, Galtsoff's suggestion is well founded. Our observations and experiments supplied the evidence that the oysters of Long Island Sound and Milford Harbor fed actively during the flood and ebb pe- riods, and that during the ebb their feeding was often more energetic than during the flood. Observations of this nature could probably be made in many other bodies of water where changes in the tides are not accompanied by pronounced ecological changes. This conclusion appears to be logical because it is quite improbable that, if other conditions of the environment remain favorable, a change in the direction of the tidal current alone would affect the oysters. It is, to a certain extent, sup- ported by our experiments in which pairs of oysters, employed in our studies of the shell movements, were always placed so that the hinges of their shells pointed in opposite directions. Thus, while the gills of one oyster faced the flood, the gills of the other animal were turned away from the direction of the current. Yet, no dif- ference suggesting that an oyster in a certain position was more active on the flood or ebb was generally noted. No correlation between the rate of water pumping of the oysters and the time of day was demonstrated by our experiments. Neither was it found that the dura- tion of the opening and closure of the shells was affected by periods of light or dark- ness. In nature many oyster beds are located at a considerable depth. Very often the water flowing over the beds contains large quantities of suspended matter which stop the penetration of a large quantity of light before it reaches the bottom on which the oysters live. These oysters, therefore, normally exist in near-darkness even during very strong daylight. It is very doubtful that if other conditions re- main favorable, the slight change in the intensity of illumination caused by the ap- proach of night could have such a pronounced effect on the oysters that they would either begin to feed at a much slower rate or stop feeding entirely. If the rate of feeding of oysters were markedly decreased during the nights and during the ebb periods, the existence of these mollusks would be under a rather unfavorable condition. Because the periods of darkness are often followed by ebb. there would be times when the feeding activities of the oysters would be continu- ously depressed for a period of approximately 18 hours. This condition would occur in September and October when the nights become long. However, it is a 262 VICTOR L. LOOSANOFF AND CHARLES A. NOMEJKO well known fact that during these two months the oysters of our waters undergo very rapid improvement in condition storing large quantities of glycogen in their bodies. Xaturally, such an improvement could not be possible if, during this time of the year, the oysters had to exist under the conditions compelling them to be relatively in- active during approximately 12 hours of darkness and also during the 6 hours of ebb, a total of 18 hours per day. As may be seen from this discussion, our conclusions regarding the activities of oysters during ebb and during periods of darkness do not agree with those of Nelson and Orton (ref. cit.). However, as Dr. Nelson suggests in recent personal com- munication with the senior author, the cause of the divergence may be attributed to the marked reduction of the pH during ebb and at night in the waters where Nel- son's experiments were conducted. During the outgoing tide those areas received large quantities of swamp water which noticeably lowered the pH. Also, according to Nelson "These waters are but slightly buffered ; hence at night with the respira- tion of algae and of animals and decomposition the water may become acid by morn- ing." Such changes are indicated in one of Nelson's reports (1924). The condi- tions, however, are different in other basins, such as in many sections of Chesapeake Bay (Loosanoff, 1932) and Long Island Sound (Loosanoff and Engle, 1940) where the pH does not closely approach the neutral point. In general, our experiments have shown that under favorable conditions neither tidal changes nor changes in the time of day affect the rate of feeding of oysters of Milford Harbor and Long Island Sound. Although the differences in the behavior between the individual oysters are of considerable magnitude, these mollusks, never- theless, appeared to be feeding all or most of the time their shells remained open which, with a temperature range from 17.0 to 28.0° C, was approximately 94 per cent of the total time. In presenting the final conclusions it should be once more emphasized that we do not interpret our results as applicable to all oyster growing areas of this coast. While our observations hold true for the areas where the experiments were con- ducted, and also, probably, for the waters where the ecological conditions resemble ours, it is realized that in other basins, where during ebb the oysters are exposed to unfavorable environment, different conditions may prevail. Nevertheless, the material presented in this article clearly indicates that the conclusions of Nelson (ref. cit.), which, no doubt, are representative for Huey's Creek, should not have been generalized and presented as applicable to the American oyster as a species (Orton, 1929). SUMMARY 1. Examination of approximately 1400 oysters collected during the different tidal stages in Long Island Sound and Milford Harbor failed to show any definite period when the stomachs of these mollusks displayed absence of food. 2. During all hours of the flood and ebb, including the low water period, the predominating majority of the oysters contained large quantities of food, whereas individuals with empty stomachs were found only occasionally. 3. The relative quantities of food found in the oyster stomachs during the ebb period were at least equal to or sometimes even exceeded those recorded during the flood. FEEDING OF OYSTERS 263 4. Analysis of the kymograph records of the rate of water pumping by the oysters showed that they fed very actively at all stages of the tide, and that the rate of feeding during ebb was at least equal to or sometimes even more rapid than dur- ing the flood stage. 5. During the ebb some of the oysters pumped on an average of 25,000 to 27.000 cc. of water per hour, while the maximum rate of pumping in some instances ranged from 31,000 to 34,000 cc. per hour. '6. The efficiency of the feeding mechanism of an oyster may be well attested by the fact that the volume of water passed during one hour through the oyster gills may be more than 1 500 times greater than the volume of the oyster's body. 7. Within the temperature range of 17.0 to 28.0° C. the oysters remained open on an average of 22 hours and 39 minutes, or 94.3 per cent, and were closed 1 hour and 21 minutes, or 5.7 per cent of a 24-hour period. 8. During the periods of flood the shells of the oysters remained open on an average of 93.4 per cent of the time, whereas during the ebb periods the shells were open 95.2 per cent. 9. During the periods of darkness the percentage of oysters with full stomachs was comparable to that of the individuals examined during the day time. 10. During darkness the oysters were found feeding very actively. The average rate of pumping at night was not lowrer than during the daytime. 11. The shells of the oysters remained open 94.4 per cent of the total time dur- ing daylight, and 93.8 per cent during the period of darkness. No correlation was found between the periods of closure of the shells and darkness. 12. Under favorable conditions neither tidal changes nor changes in the time of day affect the rate of feeding of oysters of Milford Harbor. These mollusks were found to be feeding all or most of the time when their shells remained open. 13. The results of this investigation do not lend any support to the generally ac- cepted theory that the American oyster does not feed late at night and in the early morning, and is relatively inactive on the ebb tide. LITERATURE CITED GALTSOFF, P. S., 1926. New methods to measure the rate of flow produced hy the gills of oyster and other molluscs. Science, 63 : 233-234. GALTSOFF, P. S., 1928. Experimental study of the function of the oyster gills and its hearing on the problems of oyster culture and sanitary control of the oyster industry. Bull. U. S. Bur. Fish., 44 : 1-39. LOOSANOFF, V. L., 1932. Observations on propagation of oysters in James and Corrotoman Rivers and seaside of Virginia. Virginia Commission of Fisheries, Nezvport Neivs, Virginia, 1-46. LOOSANOFF, V. L., 1939. Effect of temperature upon shell movements of clams, Venus mercenaria (L.). Biol. Bull, 76: 171-182. LOOSANOFF, V. L., 1942. Shell movements of the edible mussel, Mytilus edulis (L.) in relation to temperature. Ecology, 23 : 231-234. LOOSANOFF, V. L., AND J. B. ENGLE, 1940. Spawning and setting of oysters in Long Island Sound in 1937, and discussion of the method for predicting the intensity and time of oyster setting. Bull. U. S. Bur. Fish., 49: 217-255. MOORE, H. F., 1908. Volumetric studies of the food and feeding of oysters. Bull. U. S. Bur. Fish.. 28: 1297-1308. NELSON, T. C.. 1921. Report of the Department of Biology of the Nezv Jersey Agricultural Col- lege Experiment Station for the year ending June 30, 1920: 317-349. 264 \ KTOK L. LOOSANOFF AND CHARLES A. NUMEJKO X i- 1. SON. T. C., 1923. On the feeding habits of oysters. Proc. Soc. Exp. Biol. and Mcd., 21: 90-91. Xi i SUN. T. C., 1924. Report of the Department of Biology of the AYtc1 Jersey .li/riciiltiiral Col- lege Experiment Station jor the year ending June 30, 1923: 194-209. X i i. SUN. T. C., 1936. Water filtration by the oyster and a new hormone effect upon the rate of flow. Proc. Soc. E.vp. Biol. and Med.'l4: 189-190. XKLSOX, T. C., 1938. The feeding mechanism of the oyster. I. On the pallium and the branchial chambers of Ostrea virginica, O. edulis and O. angulata, with comparisons with other species of the genus. Jour. M orph., 63 : 1-61. ORTOX, T. H.. 1929. Oyster and oyster culture. In Encyclopaedia Britaiuiica, 14th edition, p. ~1004. AUTOSOMAL ELIMINATION AND PREFERENTIAL SEGREGA- TION IN THE HARLEQUIN LOBE OF CERTAIN DISCO- CEPHALINI (HEMIPTERA) FRANZ SCHRADER Department of Zoology, Columbia University INTRODUCTION It has been known for some time that the various lobes in the testis of many species of pentatomid Hemiptera show constant differences in the size of their cells. Bowen (1922a and b) who investigated this condition most recently, concluded that it is attributable mainly to differences in the volume of cytoplasmic elements. That is a finding which I can only confirm ; if differences in the volume of chromatin exist they must be very small. However in some species there is a testicular lobe whose cells differ from those of other lobes not only in size but which has also evolved an entirely novel process of maturation. The main features of this mat- uration are in several instances almost fantastic in character, and the evolution and constant occurrence of such "harlequin" lobes is a matter of some interest. It should be emphasized that we are dealing here not with accidental or sporadic occurrences. In the species concerned the harlequin lobe is found in each testis of every male. Moreover for any given species it is always a certain and very defi- nite lobe that is thus characterized (in the Discocephalini here treated it is the fifth) and hence it is clear that its development involves conditions that are fundamentally and firmly established in the species as it is now constituted. Harlequin lobes have so far been encountered in three species of Loxa. a species of Mayrinia (Schrader. 1945a and b). and in a species of Brachystethus (Schrader. 1946). In the last named the departure from a normal meiosis lies primarly in the autosomes which are shunted out of the spindle in both divisions; in Loxa and Mayrinia the aberrancy takes the form of amitosis and fusion in the spermatocytes. resulting in a highly variable heteroploidy. The meiotic anomalies of Brachystethus and Loxa thus appear to be in no way related and yet it seems only natural to as- sume as a working hypothesis that the evolution of harlequin lobes involves similar basic conditions in all the species involved. It is likely that further investigations will discover that harlequin lobes are pres- ent in a great many species. To the five species mentioned above and the three taken up in the present paper may be added at least three further species which I have not as yet fully analyzed — a total of eleven. Taxonomically speaking, these species cover a wide range. Loxa and Mayrinia represent typical genera of the tribe of Pentatomini ; Brachystethus is so closely related to the Edessini as to fur- nish almost a "bridging" genus between that tribe and the Pentatomini ; and the Discocephalini constitute a tribe so distinct from the other pentatomid tribes that it has sometimes been elevated to the rank of a subfamily ( Lethierry and Severin, 1896). 265 266 [-RAN/ SCHRADER Conditions in the females of all these species are still unknown, except as they were used in all instances to check the identification of the sex chromosomes in the males. Cytologists need hardly be told that this gap in our knowledge is due mainly to the technical difficulties that render a study of meiosis in the egg so onerous a task. I should like to point out again as I have done in my study of Brachystethus, that the investigation of the harlequin lobe is made under the almost ideal condi- tions of a natural experiment. The adjoining lobes of the same testis are perfectly normal and serve continually as a control; frequently the normal and aberrant cells can be studied in one and the same field of the microscope. MATERIAL AND METHODS The Discocephalini investigated are : Mecistorhinus niclanolcuciis Westwood (one male from Panama) ; Mccistorliiiuis triptcrus Fabricius (four males and two females from Costa Rica) ; Mecistorhinus sepulcralis Fabricius (one testis each from eight different males and the ovaries from one female, all from Piracicaba, Brazil) ; N co- dine luacraspis Perty (five males from Costa Rica) ; and Platycarenus notulatus Stal (three males and one female from Costa Rica). The last named species has no harlequin lobe and is only briefly mentioned in the following pages. My thanks are due to the eminent hemipterist, Mr. H. G. Barber, who identi- fied all the species of Mecistorhinus. To Professor S. de Toledo Piza of the Uni- versity of Sao Paulo, Brazil, I am deeply indebted for the material of Mecistorhinus Fixation was made in either Bauer's convenient modification of Allen's Bouin or in Sanfelice. As in all my recent studies of mitosis, I have employed three staining methods. The Feulgen technique is indispensable as a test for chromatin ; gentian violet (in Smith's modification of Newton's method) is often very useful for a study of the detailed structure of the chromosomes but even when combined with erythro- cin is not an efficient stain for the spindle apparatus in pentatomids; whereas Heidenhain's hematoxylin remains beyond all comparison the best means for bring- ing out asters, centrioles and spindle fibers. As noted above, female material was studied only to check identification of the sex chromosomes in the male. In all the Discocephalini where a harlequin lobe occurs, it is the fifth of seven lobes in each testis. In every case it is two or three times as voluminous as any other lobe although its spermatocyte cells after the leptotene stage are smaller than those of the rest of the testis. The fourth and sixth lobes which flank it on either side carry exceptionally large but otherwise normal cells, whereas the remainder conform to more orthodox proportions. In the following pages the different species are taken up separately, the de- tailed analysis of Mecistorhinus mclanoleiicus being followed by briefer comparative accounts of the other forms. As far as possible, the interpretative treatment is relegated to the discussion that terminates the paper. MECISTORHINUS MELANOLEUCUS X annul lobes K. \cept for certain features which are pertinent to an analysis of the peculiarities of the harlequin lobe, no detailed account of the spermatogenesis in the six normal HARLEQUIN LOBE OF DTSCOCEPHALINI 267 lobes need be given. It conforms closely in its general course to that which has often been described in other pentatomids. The diploid set of fourteen chromosomes is marked by one exceptionally large pair of antosomes. This pair stands out almost as conspicuously as does the X chromosome of Protenor. Here however, the sex chromosomes are relatively small, the Y being the smallest member of the complement and the X little if any larger than the smallest of the autosomes (Fig. 2). One of the exceptional features lies in the heteropycnosis that marks not only the sex chromosomes but also certain of the autosomes. Already in the early generations of the spermatogonia there are from three to five heteropycnotic bodies that stand out prominently in the resting phase (Fig. 1). There is no prochromo- some stage intervening between the last. spermatogonia and the leptotene stage of meiosis. In the leptotene and synaptic stages the heteropycnotic bodies are usually aggregated in a single mass on the nuclear periphery and it is at this locus also that the fine leptotene threads come together in a bouquet formation (Fig. 3). But in some cells there are two heteropycnotic bodies during these stages, the second and smaller one usually lying at some distance from the first and not necessarily at the periphery. The advent of the pachytene and diplotene stages sees little change (Fig. 4) in these conditions of heteropycnosis and it is only when they in turn give way to the confused stage that the single heteropycnotic aggregate is dissociated again. In this peculiar phase when staining conditions and despiralization tempo- rarily convert most of the chromosomes into pale and flocculent threads there may again be three, four, or five heteropycnotic bodies (Fig. 5). This variation in number would seem to indicate that the mutual and nonspecific attraction that brings heteropycnotic chromosomal bodies together at certain stages is not very strong and it is likely that accidents of position determine these numbers to some extent. In early diakinesis, as the chromosome threads again become definite in outline, the topographic relationships are once more open to analysis. Now the great ma- jority of cells show only two heteropycnotic bodies, one of which becomes less and less conspicuous as the threads shorten and condense l while the larger one is seen to be intimately associated with the big bivalent (Fig. 6). Somewhat later, when the paired chromosomes have assumed the typical cross and ring formations of late diakinesis this large heteropycnotic body has disappeared, but there are then two smaller bodies, one associated with each of the two spreading arms of the large bi- valent (Fig. 7). There is no doubt about the identity of these bodies. The larger is the X and the smaller the Y, the two together constituting the larger hetero- pycnotic body or chromosome nucleolus of earlier diakinesis. The dissociation of this single nucleolus into its two components is perhaps due not only to the strains that attend the separation of the arms of the large bivalent but may be a part of the regular cycle that in other species also sees the reappearance of the separate sex chromosomes at this stage. What is more remarkable is their persistent union with the arms of the autosomal bivalent, a union which is not broken until shortly before metaphase. The rather even peripheral distribution of the diakinetic bivalents disappears 1 The present state of our knowledge concerning the changes in the chromosome during a complete mitotic cycle is still unsatisfactory. Almost certainly both coiling and nucleination are involved, but the relative importance of these two factors remains undetermined. For that reason the terms "condensation" and "diffuseness" are here used in a purely descriptive sense. 268 FRANZ SCHRADER just before the disintegration of the unclear membrane. They then lie belter skelter in the nucleus and may even come into contact with each other. It is at this time when the chromosomes are in the final stages of condensation that the two sex chromosomes sever their connections with the large tetrad, though their former as- .. — v .:*, v.v .-:«• • Ht • rl an y-'^i-^'' •• SJ*^<*%. V '*'••'•' '•• "* 00 FIGURE 1. FIGURE 2. FIGURE 3. FIGURE 4. FIGURE 5. FIGURE 6. FIGURE 7. toxylin). FIGURE 8. FIGURE 9. FIGURE 10. FIGURE 11. FIGURE 12. FIGURE 13. Mecistorhinus melanoleucus — Xonnal Lobe Early prophase in spermatogonial cell; tliree chroniatin nucleoli (Feulgen). Spermatogonial metaphase ; Y is smallest of the 14 chromosomes (Feulgen). Leptotene stage (Feulgen). Diplotene stage (Feulgen). Confused stage ( Feulgen ) . Early diakinesis ; XY nucleolus attached to large tetrad (Feulgen). Late diakinesis; X and Y attached to separate arms of large tetrad (Hema- Prometaphase ; X and Y still close to large tetrad (Hematoxylin ). Metaphase I; polar view (Hematoxylin). Metaphase I ; side view (Feulgen). Anaphase I; large tetrad lagging (Gentian violet). Metaphase II; polar view (Hematoxylin). Telophase II ; 6 autosomes + Y ( Feulgen ). socialion is frequently indicated by their close proximity to it (Fig. 8). It is this stage also that is marked by an elongation of the nucleus as a whole in the polar axis, a change that plainlv involves interaction with the two centers located at the periphery of the cell. HARLEQUIN LOBE OF DISCOCEPHALINI 269 About the meiotic divisions themselves, little need be said. Metaphase I is quite typical in its conformation, \vith the now separated X and Y usually in the center of a ring of six tetrads (Fig. 9). It is however worthy of note that the large bivalent is somewhat slower than the rest of the autosomes in its condensation and side views of the first equatorial plate still show it as a cross tetrad (Fig. 10). At anaphase the X and Y divide equationally and arrive at the poles before the rest of the chromosomes, whereas the large bivalent (often showing the tertiary split) lags in its division and is distinctly slower than the other autosomes in its anaphasic progress (Fig. 11). The second division also witnesses some lagging on part of the large autosome, but this is not as striking as in the first division. The touch and go pairing of the X and Y occurs as usual, and in metaphase they line up in the spindle axis so that in polar views one is superimposed on the other (Fig. 12). They then separate to oposite poles and the spermatids receive the typical pentatomid complements of 6A + X and 6A +Y respectively (Fig. 13). The departures from the orthodox process of meiosis thus do not affect the results, which conform to the regular pentatomid scheme. Harlequin lobe Spermatogonia and meiotic prophases The spermatogonial stages in the harlequin lobe differ in no discernible way from those of the normal lobes. Here too there are, in Feulgen preparations, from three to five heteropycnotic bodies in the resting phase and the succeeding stages closely parallel the normal course of events. As in the normal lobes there is no prochromosome stage. The meiotic leptotene duplicates that of the other lobes in the number and disposition of heteropycnotic bodies, but the chromosome threads do not seem to be as finely drawn out and delicate as they normally are. This dif- ference however is too slight to furnish a secure basis for contrast (Fig. 14). Suc- ceeding this stage the developments follow a path that diverges widely from the usual one. There is neither a synapsis nor a pachytene stage. During the period in which these developments occur in the normal lobes, the chromosome threads of the harle- quin lobe merely abandon their bouquet orientation and undergo a progressive con- densation. As a result the nucleus then shows twelve somewhat loosely coiled autosomes and the two more condensed sex chromosomes, clear evidence that any sort of pairing that may have occurred unobserved prior to this time has now been abrogated (Fig. 15). The picture presented is a surprisingly close approximation of the prochromosome stage as it occurs normally in some Hemiptera, and in this respect has some resemblance to the conditions in the harlequin lobe of Loxa (Schrader, 1945b) where such a post-leptotene condensation is also encountered. But in Loxa there is a true prochromosome stage as well which occurs quite nor- mally prior to the evolution of the leptotene threads. Both there and in Mecisto- rhinns no confusion is possible for not only does the true prochromosome stage oc- cur much higher in the testis. but its nuclei are considerably smaller than are the ones here in question. This post-leptotene condensation culminates in shortened, fuzzy chromosomes that show an equational split (Fig. K>). undoubtedly a condition corresponding to 270 FRAXZ SCIIRADER 14 l * 15 16 $ o 17 ' - « 18 " * 19 " ' 2 ' 23 24 25 Mecistorhinns melanoleucus — Harlequin • 26 FIGURE 14. l.cptntcne st;i.uf ( ( icntian vinlrt). |-'H,ruK 15. 1'ust lc]it(itriH- i-ondnisatiun ; X and ^' heteropycnotic ( Fful.nm ). I II.IKI- If). Diplotene >ta,m- in univak-nts ( ( iciitian \'iuk't ). I- n, i ui 17. I .ate mnfiisi-d sta.ur ( I'Vulyvn ). iMiiKKi IcS. l;.arly diakiiH'sis ; sex chroniusoim's attaclu-d to >e]iarate lar.yv autr.soiiH- (Gen- tian \ inlet ) . i 1('. Mid-dialsinesis ; \1 unixalents aiiti IM nnes, with rach lar.yc antosoiiu- combined with . i hromosome ( ' ientiaii \ inlet ) . II \RLEQUIN LOBE OF DISCOCEPHALINI 271 the diplotene stage of the normal lobes. Throughout this period two pairs of chro- mosomes are readily recognizable; they are the two large autosomes and the two heteropycnotic sex chromosomes. Each large autosome has its ends united so as to form a split ring, a configuration that very probably arises from the mutual at- traction of its heteropycnotic terminal regions. The two heteropycnotic sex chromo- somes show no such attraction at this time and usually lie well separated, evidence that heteropycnotic attraction is confined to certain conditions of the heterochro- matin. The confused stage which now intervenes, temporarily halts a further close analysis of progressive chromosome changes. The autosomes once more become diffuse and uncoiled and at the height of the stage stain very lightly. Usually three heteropycnotic bodies are present at this time, but there may be as many as five ( Fig. 17). In the latter case the bodies are smaller, generally speaking, which would indicate that the variations in number are due to some vagaries in mutual at- traction and aggregation. The two sex chromosomes and the ends ot the two auto- somes would account for six such bodies which suggests that some aggregation is nearly always present. \Yith the termination of the confused stage and the beginning of diakinesis, the individual chromosomes once more appear as such. There are then three hetero- pycnotic bodies and two of these are seen to be associated with the two large auto- somes (Fig. 18). The third shows no such definite association and gradually dis- appears. In mid-diakinesis a more exact analysis of these conditions becomes possible. At this time there is a total of either eleven or twelve chromosomal bodies in every nucleus. When there are twelve, the two autosomes are quite independent of each other and may lie far apart. Each of them has a large chromatin nucleolus or heteropycnotic body attached to it at the place where the ends are still joined in ring formation (Fig. 19). When on the other hand there are only eleven bodies, these two chromatin nucleoli have come together, and through them the two large, ring formed autosomes have joined in a figure eight (Fig. 20). The two chroma- tin nucleoli represent the X and Y chromosomes and again, the most natural ex- planation of such configurations would seem to lie in the forces of heteropycnotic attraction; the heteropycnotic ends of the large autosomes are drawn together to form rings, and the heteropycnotic sex chromosomes later become attached to these regions and to each other for the same reason. It is rather strange that no case has been encountered in which both sex chromosomes have become joined to only one of the large autosomes, since nonspecific heteropycnotic attraction might be expected to give rise occasionally to such configurations. However, nuclei of this stage in which the chromosomes are open to a clear analysis are not common and FIGURE 20. Mid-diakinesis; both sex chromosomes and both large autosomes in one combi- nation (Gentian violet ). FIGURE 21. Equatorial ring side view (Heniatoxylin). FIGURE 22. Equatorial ring slightly later; polar view ( < lenlian violet). FIGURE 23. Autosomes in precocious return to diffuse condition; X and Y still heteropyc- notic (Gentian violet ). FIGURE 24. Formation of autosomal aggregate; X and Y heteropycnotic (Gentian violet). FIGURE 25. Dissociation of X and Y from autosomal aggregate (Gentian violet). FIGURE 26. Metaphase I; X and Y on middle spindle and autosomal aggregate displaced (Heniatoxylin). l-'UAXZ SOI KADI. R the fourteen e.\ani])les which have heen studied liardly constitute a sufficiently large number to justify the conclusion that they do not occur. Prometaphase Shortly before the breakdown of the nuclear membrane a significant reorienta- tion of the chromosomes takes place. This is at about the time that the nucleus rloiigates toward the peripherally located centers. The chromosomes, still not fully condensed, then are shifted to the middle region between the centers and since they remain in close proximity to the nuclear wall and have lost the property of mutual repulsion, they tend to form a more or less circular row or chain in the equator. Some of the components of such chains may be in actual omtact with each other, while others may be connected by Feulgen positive bridges or show no attachment at all (Figs. 21, 22, and 72). It is likely that such bridges are similar in nature to those seen later at metaphase (see for instance Ris, 1942). but whether they represent viscous connections that persist after a former contact or are indica- tive of a "reaching out" of chromosomes toward each other, it is impossible to decide. When the nuclear membrane finally disappears, this picture undergoes marked and sudden changes. The chain of chromosomes, now free of the influence of the membrane, seems to collapse inwardly, frequently forming a closed ring at first and then an irregular aggregate in the middle of the nuclear space. In the many cells seen at this and the following stages no instance of more than a single aggregate has ever been observed, a point of difference with the case of Brachystethus ( Schrader, 1946). Concurrently with these changes of orientation there occur alterations in the chromosomal structure. These are marked especially by a partial return to the diffuse condition in the autosomes, with an accentuation of the equational split. The two large autosomes do not seem to become quite as diffuse as the rest, but this difference is not a striking one at best. This return to a more diffuse state causes the autosomal aggregate to appear as a spongy and vacuolated mass in both gentian violet and Feulgen preparations and this condition is maintained for the major part of the first division (Fig. 23). Hematoxylin slides allow no such structural diagnosis for there the aggregate is nearly always homogeneously and in- tensely stained. The behavior of the sex chromosomes is remarkable during the prometaphase and the establishment of the metaphase itself. At the time of the equatorial ring formation, just prior to the disintegration of the nuclear membrane, they are still very close or even in contact with the large autosomes. Almost always they lie on the inner side of the ring and not in seriation with the rest of the chromosomes (Fig. 22). They seem to be almost fully condensed at this time and are recnguixable in most cells. When the autosomal chain collapses to form the irregular aggregate, this distinction becomes even more marked, for in contrast to the autosomes they then maintain their condensed state and in addition tend to protrude from the spongy mass of autosomes (Fig. 24). This protrusion seems to be an indication of inter- action with the two centers, for the sex chromosomes not only make their appear- ance on the side toward one ol the poles but begin to place their long axis in align- ment with the polar axis ( Fig. 25) which is a placement assumed also at the ensuing metaphase. HARLEQUIN LOBE OF DISCOCEPHALIXI The first division The clumping of the partially diffused autosomes brings about a rather anoma- lous situation and the establishment of the metaphase can be followed only through the behavior of the sex chromosomes. These appear to be quite normal in their further maneuvers. They finally become completely detached from the spongy ag- gregate of autosomes and take up an equatorial position side by side. During this movement several other developments occur simultaneously. Chromosomal fibers appear connecting the sex chromosomes as well as the aggregate with the poles, and at the same time the whole mass of autosomes is shunted out of the middle region toward the side of the cell. This shift must be rather sudden, for intermedi- ate stages are very rare. In extreme instances the displacement may bring the auto- somal aggregate very close to the side of the cell, though never touching it. and in every case it comes to lie farther from the polar axis than from the cell wall. AYhile in this position, two points are to be noted: the autosomal aggregate remains con- nected with the poles through definite chromosomal fibers, and even in its dis- placement it maintains an equal distance from both poles (Figs. 26, 73, and 74). The whole reaction is obviously closely akin to a similar one observed in the penta- tomid Brachystethus ( Schrader, 1946). The two sex chromosomes apparently are not affected by the anomalous be- havior of the autosomes. They lie side by side in a compact and narrow spindle of normal length and undergo an orthodox equational division. In some cells the chromatids of the Y separate faster than those of the X and may precede them to the poles. As soon as the anaphase movement of the sex chromosomes is initiated, the autosomal aggregate once more approaches the polar axis, and by mid-anaphase is usually close to or even in contact with the sex chromosome spindle. This return also occurs in the equatorial plane of the cell, and is correlated with a shortening of the chromosomal fibers as well as the lengthening of the interpolar distance and cell as a whole — both of which will of course bring the autosomal aggregate closer to the polar axis again (Figs. 27 to 29). Although the autosomes, aggregated as they are, pass through these maneuver.- as a unit, there is evidence from the beginning that one of them plays a special role. Already at the first trace of division in the X and Y chromosomes, a single large chromosome protrudes from the autosomal clump, showing a well formed chromo- somal fiber connection with one center and clearly oriented toward it. In such a position it appears more condensed than the rest of the autosomes, a condition which would be difficult to discern while it is still in the midst of the vacuolated. unevenly staining aggregate (Figs. 27 and 28). The reaction of this autosome to the pole is quite independent of the sex chromosomes, but it is obviously hindered in its move- ments— probably because of the "stickiness" that tends to hold all the autosomes to- gether. As a result the two sex chromosomes are well on their way toward the poles before this autosome has disengaged itself from the encumbrance (Figs. 29 and 75). Soon after it has left the aggregate, a second large autosome begins to dissociate itself from the rest of the autosomes. The extent to which it succeeds in this is highly variable in different cells, but in most cases it at least protrudes from the mass be- fore the division is finished (Fig. 30). Often, while the first autosome is still fairly 274 I-U \XX SCHRADER 27 28 29 • f •* 30 33 31 34 32 f I 35 I n,i RK 27. I;.arly aiiaphaM1 I; autoMnnal a, y.s; rebate l)fuiniiin.L; to return { ( ientian violet). FIGI RB _J«. Mid-anapliasc I; ((.entian violet). RE 2(>. I. ate anaphase I ((ientian violet). FIGURE 3D. Marlx telopliase I ((ientian violet). RB 31. Late telophax- I ( I leinatoxylin ). FIGURE •>-'. l-'arly anaphaM- 11 — lar.ue ei'll ((ientian viok't). GURH i-x Mid-aiiapliast- II — lar^e eell ( I leinatoxylin ). FlGl RE .1-4. Marly telopliase II — lar^e cell ( Heinatoxylin ) . I;K, i -KK ^5 Telophase II large cell ( I leinatoxylin ). HARLEQUIN LOBE OF DISCOCEPHALINI 275 close to the aggregate, it shows a Feulgen-positive connecting thread with the sec- ond one, although this is nsuallv severed very shortly (Fig. 29). The rest of the autosomal clump, which is by now very close to the polar axis, also shows some response to the mitotic forces and frequently undergoes some elongation in the polar direction. Its effective movement however is always op- posite in direction to that of the first autosome and hence it approaches the other pole. The second large autosome may sometimes almost reach the middle of the cell, hut in only a single case has it proceeded so far that the ensuing cleavage con- striction promises to include it in the same cell with the first autosome. Indeed, a count of 100 small second spermatocytes (those which do not receive the main part of the autosomal aggregate) has revealed no such accidental inclusion and it must he very rare. Further, in all ohserved cases (of which there are many dozens) this second autosome has rejoined the aggregate by the time that the second division is begun so that its mitotic motion must finally be reversed (Fig. 31). Possibly it initially follows the large autosome only because it is dragged along by the connect- ing thread. The final result of these maneuvers is that the large autosome and the rest of the autosomal aggregate always go to opposite poles. Since the former is the first to evince any reaction to the mitotic forces, one is almost forced to the hypothesis that it determines the direction of movement on the part of all the re- maining autosomes. These two autosomes that tend to disengage themselves from the aggregate are patently larger than any but the two largest chromosomes of the diploid set. At the same time they do not seem quite to reach the size of that large pair and it may therefore be that we are dealing with the two chromatids of only one of the latter. But admittedly the changes in the state of autosomal condensation during meiosis make it difficult to decide the matter on the basis of size alone. The later behavior of these exceptional chromosomes in the spermatids would however seem to support their identification as chromatids rather than whole chromosomes. The second spermatocytes resulting from this first division, anomalous though it may be. are thus very constant in composition. Half of them contain the two sex chromosomes and only one autosome ; the rest carry all the remaining autosomes as well as the sex chromosomes. The latter are much the larger cells of the two, as might be expected (Fig. 31). The second division Lan/c cell: The chromosomes carried by these second spermatocytes comprise the sex chromosomes as well as all of the autosomes except the single one in the smaller cell. There is no interkinesis. In its general aspects the division of this cell simu- lates the first division. As the X and Y take their position on the new spindle, chromosomal fibers are also formed between the two poles and the autosomal aggre- gate. Almost simultaneously, the latter is shifted toward the side of the cell, the displacement being very similar to that observed in the first division i Fig>. 32 and 76). However the aggregate now begins to lie less spongy in appearance and there are other indications that its autosomal constituents are once more undergoing con- densation (Fig. 33). During the anaphasic separation of the X and V, the aggre- gate is again drawn toward the middle of the cell. It may be noted that when it reaches the compact little sex chromosome spindle, its surface of contact with it be- comes smooth and concave, indicating that it is under vnne pressure (Fig. 33). 276 l-UAXZ SCHRADER 36 37 f § 38 I 39 % 40 41 * * .* »• <%• 42 43 44 Mccistnrliiinis uicluiinlcitcits — I lurlcqitin Lobe (except Fi.ys. 41 and 43 < Fi<;ripennati(ls — harle(|nin lobe ( Hematoxylin). HARLEQUIN LOBE OF DISCOCEPHALINI 277 Then as the anaphase progresses and the cell and spindle elongate, the aggregate is drawn out between the two poles and many of its component autosomes are thus strung out in a roughly linear order (Figs. 34 and 78). But the tendency to stick and adhere to each other remains strong so that even when several individual chro- mosomes have been pulled out of the aggregate they usually still show thick, Feulgen- positive connections with each other and with the remaining aggregate (Figs. 34 and 35). The division so far as the autosomes are concerned is therefore obviously a haphazard one and the aggregate is frequently distributed very unevenly to the two spermatid cells. The spermatids resulting from the division of the large second spermatocyte are therefore exceedingly variable in composition. Since the sex chromosomes segre- gate normally, all spermatids carry either an X or a Y, but the number of autosomes is largely a matter of chance. Nevertheless no cases have been observed in which the latter have all gone to one pole. Small cell: As is the case with the large second spermatocytes there is no inter- kinesis of any kind and the final stages of the first division merge directly into the beginnings of the second. Indeed before the anaphase movement of the first divi- sion has been completed, the two daughter centrioles at each pole (each centriole appears double already at metaphase) have separated and begin their migration to establish the polar axis of the second division, at right angles to the first (Fig. 31). In some cells the sex chromosomes respond and orient to these new poles before they have completed the anaphasic movement. But apparently this precocious movement is then reversed, for in slightly later phases when the new axis has been established, the two sex chromosomes are near each other or in actual contact in the middle of the cell. Here too now lies the large autosome which has been delayed in its arrival at the pole. It may be in contact with neither, either or both sex chromosomes, or it may be connected with either or both through Feulgen-positive bridges (Fig. 36). Chromosomal fibers are formed already before the three chromosomes have gath- ered at the midpoint, but so far as can be seen the tiny spindle is concerned only with the X and Y. No chromosomal fibers can with certainty be traced to the auto- some. The sex chromosomes behave just as they do in a normal cell and after meeting in the middle of the spindle they separate to opposite poles in a regular segregation. It is the behavior of the autosome that presents some puzzling aspects, as it al- ready has done in the first division. The outstanding feature of this behavior lies in the fact that it nearly always goes to the same pole with the X. In so doing it appears to have little or no independent mitotic movement, acting merely as a satel- lite of the sex chromosome. Its dependence on the latter is shown not only by the absence or poor development of chromosomal fibers already mentioned, but is indi- cated also by its behavior before and during anaphase. In the grouping prior to the division, the autosome sometimes lies between the X and the adjacent pole, but when the anaphase movement is under way it always trails the sex chromosome (Figs. 36 to 40, and 77). Since during most late anaphases the autosome is at- tached more or less closely to the X, its behavior might be attributed to a simple adhesion between the two were it not for the fact that many earlier anaphases show no such connection. The latter thus frequently seem to be established after the mi- totic movement is under way. Indeed, connections of this sort cannot be decisive in any case since prior to the division the autosome often lies in contact with both 278 FRANZ SCHRADER sex chromosome or shows a Feulgen-positive bridge to the Y and not to the X (Fig. 36). Evidently this is later broken so that the mitotic association with the X must involve some selective action not dependent on such physical bonds. What- ever the underlying mechani>m may be. the results of the division admit of no doubts concerning the constancy of the relationship between these two chromosomes. In 100 clear side views of late anaphases there were only six that admitted the possi- bility of an association of the autosome with the Y. In two of these the autosome is clearly much closer to the Y than the X and probably going to the same pole, whereas in the remainder the identification of the X and Y is not certain. In close to 95 per cent of the cases therefore, the autosome accompanies the X to the pole and not the Y. Generally speaking then, the small spermatids which come from this division are of two types: one. which carries the X and the autosome; and an- other which contains only the tiny Y. Spermatids and sperms The two anomalous spermatocyte divisions of the harlequin lobe thus give rise to four main types of spermatids : X + one large autosome ; Y ; X + a variable num- ber of autosomes ; Y + a variable number of autosomes. Since in the spermatid the autosomes scatter again and tend to be distributed peripherally on the new nu- clear membrane before becoming diffuse, rather dependable counts are often possible. Both of the small types of spermatids are readily recognizable and it is to be noted that the close association between the X and the autosome of the preceding telo- phase is maintained into an advanced spermatid stage (Fig. 42"). In the large rmatids the number of chromosomes may be as low as five or six and frequently higher than twenty. The latter counts constitute conclusive evidence that the chro- matids of the univalent autosnmes have separated from each other, since the full. normal haploid number of the species is only seven (compare the normal spermatids of Fig. 41 with those of the harlequin lobe in Fig. 42}. Since the autosome that is . — •dated with an X in one of the small spermatids undergoes no such separation into smaller units, there is strong presumptive evidence (to support that which was adduced earlier) that it represents a chromatid which has already separated from its - -:er chromatid — the second large autosome of the first division. Formation of sperms seems to proceed normally in all cells until the stage when the elongation nucleus shows the pointed apex which indicates the presence of the acrosome (compare Fig. 43 with Fig. 44 I. At about this time, when the chromo- somes have become very diffuse and the contents of the sperm head seem structure- less, the small types of sperms gradually stain more and more faintly. It finally becomes impossible to recognize them, whereas the larger types assume the attenu- ated, intensely stained form of the head that is also encountered in the normal lobes. It is more than probable therefore that the smaller sperms never attain maturity, whereas the large ones appear normal in every respect but size. They enter the -rm duct in a perfectly regular manner and there mingle with the sperms of the :nal lo' MECISTORHINUS TRIPTERI 3 The main points in which .Mcclsturliiniis triptcrus differ- from Mecistorhinus invohe the I - - of the autoximes and the behavior of the HARLEQUIN LOBE OF DISCOCEPHALIXI sex chromosomes. So far as the diploid complement of chromosomes is concerned, there is no perceptible difference, both having an outstandingly large pair of auto- somes and a Y that is the smallest of all the chromosomes. However, already in the spermatogonial resting phases there is a less extensive heteropycnosis in .\fecis- torhinus triptcrus, for then as well as later there are never more than two hetero- pycnotic bodies (instead of from three to five). The spermatogenesis in the normal lobes is entirely orthodox. The heteropycnosis in the large autosomes is here lack- ing entirelv and hence there is no association between them and the sex chromosomes f Fig. 45). The lack of autosomal heteropycnosis obtains also in the harlequin lobe (Figs. 46 and 47). As a consequence, when the equatorial ring is formed the two sex i 45 51 Mecistorhinus triptcrus — Harlequin Lobe (except Fig. 45) FIGURE 45. Normal diakinesis ; X and Y independent of autosomes (Feulgen). FIGURE 46. Early diplotene stage (Feulgen). FIGURE 47. Diakinesis; X and Y joined (Feulgen). FIGURE 48. Equatorial ring in formation; X and Y independent of autosomes (Feulgen). FIGURE 49. Anaphase I (Feulgen). FIGURE 50. Anaphase II — small cell (Feulgen). FIGURE 51. Anaphase II — large cell (Feulgen). chromosomes lie in the approximate middle of the nuclear space instead of being carried to the periphery by the large autosomes (Fig. 48). The X and Y are at this time still in more or less intimate contact with each other and remain so until the first anaphase has begun. Another difference lies in the size relations between the X and the Y. In 3/V torhinus mclanolciicus the X is markedly larger than the Y, but this difference less pronounced in Mecistorhinus tripterus. In the latter also, as in most other pen- tatomids. the sex chromosomes become less sharply outlined in the second division and therefore it then becomes difficult at times to distinguish between then: 50). As a consequence, in about 25 per cent of fifty cells it is not possible to decide whether the autosome follows the X and there is at least the possibility that in such 280 1'RAXX SCIIKADER instances it accompanies the Y. Although there is thus no question that in this race too the association is between the X and the autosome in the great majority of cells, the case is not as clear-cut as it is in Mecistorhinus melanoleucus. Mention should also he made of the fact that already in the first division, the first large autosome usually shows a median furrow (Fig. 49). If, as in Mccistorhinns melanoleucus this chromosome represents only one chromatid of one of the large uni- valents, the furrow is equivalent to a tertiary split. Whether that he correct or not. this split is not consummated even in the early spermatid, where the large autosome still maintains this appearance. But in essence, the meiotic divisions of the two species are very much alike (Figs. 49-51, and 78). The difference in heteropycnosis docs not affect the results and Mecistorhinus trif>tcrns merely furnishes less decisive evidence anent the association between the X and the large autosome. MECISTORHINUS SEPULCRALIS The rather extensive material of this species from Brazil conforms closely in its general cytology to the Central American Mecistorhinus species. The diploid set is characterized by a very large pair of autosomes and the Y is again the smallest member of the set. The meiosis in the normal lobes follows an orthodox course al- though there are indications of heteropycnosis in a very restricted region of the two large autosomes. This heteropycnosis is very much alike in normal and harlequin lobes. The spermatogonial resting phases may show as many as three or four very small hetero- pycnotic regions but more often only one larger one. In the meiotic prophase fol- lowing the leptotene stage and up to diakinesis, it is difficult to recognize any heteropycnosis in the autosomes whereas both sex chromosomes are, as usual, hetero- pycnotic throughout. In the confused stage there are generally two heteropycnotic bodies, but whether these represent only the sex chromosomes or also certain auto- somal regions it is not possible to say (Fig. 52). Only in mid-diakinesis do condi- tions allow a closer analysis. In the harlequin lobe, where we are dealing with uni- valents, both large autosomes then show heteropycnotic terminal regions, and these are joined so as to form closed rings as in Mecistorhinus melanoleucus (Fig. 53). Apparently this heteropycnosis is less extensive in the present species and this may account for the fact that the mutual attraction of such regions does not bring the two large univalents together in the "figure eight" formations that occur so often in the other forms. Probably for the same reason the association between the sex chromosomes and these autosomes is highly variable. Thus in some cells only one sex chromosome is joined to the heteropycnotic region of one univalent autosome while the other sex chromosome is free (Fig. 53). In late diakinesis such a condi- tion is rare because, as in Mecistorhinus triptcnis, there is then a strong tendency for the X and Y to come together. Hence the free sex chromosome often joins its attached mate and as a result both are carried into the equatorial ring of autosomes. just prior to the breakdown of the nuclear membrane (Fig. 54). In other cases the sex chromosomes may be entirely free and will then take a more or less central position in the nucleus at the time of the equatorial orientation of the autosomes. But these various maneuvers that involve heteropycnotic attraction do not affect the course of the actual divisions in the harlequin lobe any more than in the normal HARLEQUIN LOBE OF DISCOCEPHALINI 281 lobes. In the former the sex chromosomes at both metaphases become detached from the autosomal aggregate, which then is displaced toward the side of the cell just as in Mccistorhinus triptcrus (Figs. 55 and 57). In Mccistorhinus sepulcralis how- ever, the first anaphase shows a definitely greater tendency for other chromosomes to follow the first large autosome out of the aggregate. Although the second large atitosome is included as rarely in the smaller second spermatocyte as in the preceding species, a smaller autosome sometimes succeeds in following the first autosome into this cell (in four out of forty cells) (Fig. 56). Apparently both these autosomes Mccistorhinus sepulcralis— Harlequin Lobe FIGURE 52. Confused stage (Feulgen). FIGURE 53. Diakinesis stage (Feulgen). FIGURE 54. Equatorial ring; X and Y associated with large autosomes (Feulgen). FIGURE 55. Metaphase I (Hematoxylin). FIGURE 56. Late anaphase I (Hematoxylin). FIGURE 57. Metaphase II — large cell (Hematoxylin). FIGURE 58. Anaphase II — small cell (Hematoxylin). retain the "sticky" condition of the earlier aggregate, for when they reach the pole the smaller joins the larger one and adheres to it in almost any position in random fashion. In the second division, both follow the X to the pole in the majority of cases (Fig. 58). But here too, as in the second division of Mecistorhinus tripterus the X and Y often show only a small size difference In perfect side views they can nearly always be distinguished, but if the cell is viewed at a slight angle it at once becomes difficult to do so. Hence, though Mecistorhinus sepulcralis also shows a definite preferential association between the X and the large autosome (as well as the smaller, if present) !< KI'HALIXI 289 ! 73 75 I 5 I <0 o o 0 76 77 78 290 FR ANY. S(. IIRADER LITERATURE CITED BOWEN, R. H., 1922a. Notes on the occurrence of abnormal mitosis in spermatogenesis. Biol. Bull., 43 : 184-203. BOWEN, R. H.. 1922h. Studies on insect spermatogenesis. I \'. I'roc. .\m. ./;-/. Sci.. 57: 391- 423. LETHIERRY. J. AND G. SEVERIN, 1896. Catalof/uc general dcs Hcmiptcrcs. Britxcllcs. 838 pages. Ris, H., 1942. A cytological and experimental analysis of the meiotic behavior of the univalent X chromosome in the bearberry aphid Tamalia (Phyllaphis) coweni (Ckll). Jour. E.rp. Zool.. 90: 267-330. SCHRADER, F., 1941. The spermatogenesis of the earwig Anisolahis maritima Bon. with refer- ence to the mechanism of chromosomal movement. Jour. Morph.. 68: 123-148. SCHRADER, F., 1945a. Regular occurrence of heteroploidy in a group of Pentatomidae ( Hemip- tera). Riol. Bull.. 88: 63-70. SCHRADER, F., 1945b. The cytology of regular heteroploidy in the genus Loxa ( Pentatomidae— Hemiptera). Jour. Morph.. 76: 157-177. SCHRADER, F., 1946. The elimination of chromosomes in the metiotic divisions of Brachystethus rubromaculatus Dallas. Biol. Bull.. 90: 19-31. THE EFFECT OF THE ADULT ANTERIOR PITUITARY HORMONE ON THE TADPOLES AND THE IMMATURE MALE FROGS OF THE BULLFROG, RANA CATESBIANA ROBERTS RUGH Department of Biology, H'asliington Square College oj Arts and Science, \rew York University, New York City, and Marine Biological Laboratory, Woods Hole, Mass. INTRODUCTION The interrelationship of the anterior pituitary hormone and the gonads of verte- brates has been well established. There is evidence that the amphibian gonads, under the influence of the anterior pituitary hormone stimulation, will either respond immediately (in mature animals) by the liberation of mature gametes or will dif- ferentiate (if immature) in much the same manner as they do during the breeding season, with the exception that these artificially stimulated changes are accomplished at a much faster rate. This statement is supported by the work of Riddle and Flemion (1928) on the clove; Houssay and Lascano-Gonzalez (1929) on the South American toad; Herre and Raiviel (1939) on larvae of Triton; Burns (1930) and Burns and Buyse (1934) on Amblystoma; Evans (1935) on the lizard; Forbes (1937) on the alligator; Rugh (1937, 1939, and 1941) on frogs of different genera; Puckett (1939) on the bullfrog tadpole, and Glass and Rugh (1944) on the mature frog, Rana pipiens. In this latter paper it was shown that while there are seasonal changes in the frog testis, indicating cyclical maturation, the gonad will respond at any time of the year to frog anterior pituitary hormone stimulation by the liberation of any mature spermatozoa which are present, and that frequently (especially during active maturation of August) the seminiferous tubules will also contain other matura- tion stages. Recently Schreiber and Rugh (1945) have studied the response of the testes of recently metamorphosed Rana pipiens to the pituitary hormone and to estradiol-benzoate. It was the purpose of this investigation to determine the effect of the anterior pituitary hormone of the adult frog on the gonad of the pre- and the post- metamorphosing male bullfrog. The testis can be differentiated from the ovary tnacroscopically even before tail absorption in the tadpole of Rana catcsbiana. MATERIALS AND METHODS The tadpoles of Rana catcsbiana were collected in the vicinity of \Yoods Hole, Mass., and were used immediately. During the month of August all stages of pre- and post-metamorphosis are available in great abundance. These tadpoles are in their second year of growth, and generally go through the stages of metamorphosis rather rapidly, depending upon the water temperature and the amount of food avail- able. Since these tadpoles were well fed in their natural habitat, there was no at- tempt to feed them during the short period of the experiment, three to four clays. 291 ROBERTS RUGH Tadpoles we're selected in five stages: n. Pre-metamorphic stage where there were four well-developed appendages and a full-length tail fin. /'. Post-metamorphic stage where the body length was less than 45 mm. and the tail had just been absorbed, but could be identified as a stub. These frogs had just shifted to lung breathing. c. Frogs of body length of 65 mm. Oil2r"J POST PITUITARY STIMULATION 35H&&2,/; m^ &:&,* ^$^2^^*%*g& •; -^^s^aite^ '-^•^•..;^-'A • . • PRE- METAMORPHOSIS POST -METAMORPHOSIS 65 MILLIMETER BODY LENGTH 8 95 MILLIMETER BODY LENGTH PLATE I. Effect of anterior pituitary hormone on testis of immature bullfrog, R. catcslnaua. 2«)4 KOlll-KTS ULT.H n lent membrane of the tubule-. There are al.so oval-shaped darkly staining nuclei of the spermatogonia, and occasionally there may he seen (toward the region of the future lumen ) some secondary spermatocytes. The average length of these testes is 2.18 mm. and the average diameter is 0.80 mm. When the pre-metamorphic tadpole is subjected to the anterior pituitary hormone (by coelomic injection) the most obvious change in the testis is the vacuolization of the forming seminiferous tubules (Figs. 2 and 10) so that the entire gonad takes on a sponge-like appearance. The peripheral region of the gonad is always the first to be affected, and the most vacuolatecl. Under high power magnification it is noted that the only cells remaining within the seminiferous tubules are those attached to or an integral part of the basement membrane. The other cells which had previ- ously occluded the lumen will be seen in the collecting tubules and the vasa efferentia, indicating that they were enroute through the reproductive tract at the time the gonad was fixed. ( )ne gets the impression that the gonad was flushed out of all loosely attached cellular components. The size variations of the control and the experimental gonads were not appreci- ably different. j Post-metamorphic static, hociv Icin/tli less than 4? nun. In this stage the gonad itself has enlarged considerably so that the maximum length of the testis is about 2/i mm. and the maximum diameter is about 1.8 mm. The shape of the gonad is also changed, since its length increased about 40 per cent while its diameter increased about 125 per cent. In the control testis ( Figs. 3 and 1 1 ) there is some evidence of the forming semi- niferous tubules, and maturation stages from spermatogonia to mature spermatozoa are seen. This indicates that coincidental with the drastic changes of metamorphosis the bullfrog tadpole's testis first acquires structurally mature spermatozoa. Whether or not these spermatozoa are functional has not yet been determined. Clusters of cells are seen, all in the same stage of maturation, with the progressively advanced stages more nearly in the center of the seminiferous tubules. Here again, when the testis is subjected to the anterior pituitary hormone of the adult, there is a liberation from the tubules of all freely associated cells (Figs. 4 and 12) so that the post-stimulated gonad has much the same appearance as that of the experimental totes of the pre-metamorphic stage (Figs. 2 and 10). 'There re- main only the spermatogonia, primary spermatocvtes. and the few interstitial cells. The majority of cells found within the collecting tubules and the vasa efferentia (Fig. 17) are secondary spermatocytes. 6? nun. buclv I an/ 111 static By this stage of development the testis has acquired its adult proportions and measures about 3.35 mm. in length and 2.0 mm. in diameter. The surface has become lobulated due to the further development of the seminiferous tubules about the peripherv. Xests of maturing cells are seen ( Figs. 5 and 13) representing all stages from the primary spermatocytes to the spcrmatids and mature spermatozoa.' early maturation figures are in great abundance with the pachytene groups the most prominent (Fig. 13). It is at this stage that the cysts or nesting groups of cells become distinct from each other, demarking those clusters of cells, all of which originate from the same spermatogonium. With the enlargement of the seminiferous KFFFAT OF PITl'ITARY HOKMoXK ON BULLFROGS 295 CONTROL POST PITUITARY STIMULATION Mr> • j& ;«•'<**. v i V*«-Vf YH&. * A W?^¥f i^Sv* ****, ^^V?*T pr.^ftfef ? 9 Ik POST- METAMORPHOSIS :CTINC TUBULES PLATE II. Effect of anterior pituitary hormone on seminiferous and collecting tubules R. cutcshinna. 296 ROBERTS Kl'CH till ink' there appears to be a compensatory reduction in the amount of interstitial tissue, due, no doubt, to the stretching of this tissue in the general growth of the gonad. I "pon stimulation of this testis hy the anterior pituitary hormone of the same spe- cies, there is a loss from the gonad of all or most of its freely associated cells and cell groups ( Figs. 6 and 14). The gonad seems to have its peripheral tubules emptied the more completely, and. upon closer examination, it is noted that often there remain only the scattered primary spermatocytes and spermatogonia. The primary sper- matocytes are fewer in relative number than in the previous stage studied. This is ]>rol')ably due to the fact that some of these cells have been utilized in the production of the maturation nests of cells, some of which were liberated under hormone stimu- lation. The maturation stages which are missing from the seminiferous tubules are found within the collecting tubules (Fig. IS) and the vasa efferentia leading from the testis to the kidney. f}j nun. hody lein/tli sta . / «t SPERMATIDS* SECONDARY SPERMATOCYTE PRIMARY SPERMATOCYTE flft€c.._, POST PITUITARY STIMULATION PI.ATK III. Seminiferous tubules of adult I\'UIHI catesbiana. * ROBERTS RUGE Steps in maturation. but, in contrast with tin- younger Frogs there are relatively few of the early maturation stages, i.e., spermatogonia. rp("i stimulation by anterior pituitary hormone the picture becomes >omewhat clearer, for the seminiferous tubules are emptied of the mature spermatozoa and an occasional spermatid, and consequently the other stages of maturation within the >eminifen>us tubules become visible (Fig. 21). The collecting tubules and the re- lated vasa efferentia, however, contain almost exclusively the mature spermatozoa indicating a resistance to liberation of the earlier maturation stages. This resistance was not in evidence with the gonads ot younger stages. Sperm release is inde- pendent of any overt sex behavior on the part of the male, being a direct response to the anterior pituitary hormone stimulation. DISCUSSION The presence of smooth muscle fibers in the ovaries of the frog has been demon- strated (Hugh, 1935) and the contraction of the lobes of the ovary has been seen in highly magnified moving pictures. It has further been demonstrated that the ante- rior pituitary hormone acts not directly upon the gonad but by way of either or both the nervous and the circulatory systems (Rugh, 1935 and Samartino and Rugh, 1945). This hormone, further, has been shown to stimulate other secondary sexual characters such as the thumb pad of the adult (Glass and Rugh. l'M-4) and the Mul- lerian ducts of the immature Rana pi picas (Schreiber and Rugh, 1945). In view of these and other findings, it is now clear that upon injection of the anterior pituitary hormone of the frog into the body cavity of the frog, supplementing its own glandular secretions, both the primary and secondary sexual characters are stimulated and the ovary (at least) utilizes its smooth muscle elements to free its follicles of their enclosed eggs. Puckett (1939) and previously Swingle (1921 and 1()26) found a race of bull- frog tadpoles in which there were so-called undifferentiated gonads, particularly in the male, known as protestes. These protestes reacted to daily injections of various mammalian hormones (pituitary and gonadal ) by showing precocious development and differentiation, particularly when male sex and pituitary hormones were injected simultaneously. There was no influence on the differentiation of the gonads with regards to sex, but rather the pituitary treatment determined the rate at which the gonads grew. The variety of Rana catcsbiana used in the experiments reported in this paper did not show the stage of undifferentiated gonads described by Puckett, but they did show a decided response to the amphibian pituitary alone. Adult fe- male frogs have been found to be almost totally resistant to mammalian pituitary ex- tracts ( K'ugh, 1()35 and 1942), while the closely related toads and salamanders would respond readily. Recently Schreiber and Rugh (1945) treated metamorphosing Rana pipiens with the adult frog anterior pituitary hormone and also with the female hormone CM radio] ben/oate. There hormones were used independently and it wa> demon- strated that the pituitary hormone acted on the gonad while the est radio! acted upon the gonaducts (oviducts and Mullerian ducts). It is therefore evident that the gonad, whether immature or mature, will respond to anterior pituitary hormone stimulation if the frog is treated \vith adult amphibian pituitary glands. The period of this experiment, i.e.. 4S hours, was too short to allow demonstra- EFFECT OF PITUITARY HnkMoXK ON BULLFROGS 299 tion (if any growth response of the gonad> lint there was immediate and drastic re- sponse at all stages, whether spermatozoa were present or not, by the elimination of cellular elements from the forming seminiferous tubules. Since there is experi- mental proof ol smooth muscle presence and activity in the frog ovary, and since the post-stimulation picture of the frog testis indicates a greater response at the periph- ery, it is suggested that here too there may be pituitary stimulation of smooth muscle elements within and surrounding the testis which causes the evacuation of cellular elements. Even if there were evidence of selective action of the pituitary hormone on a specific cell type such an hypothesis would have no real precedent. \\'e are dealing with phases in the maturation of one cell type and, in the immature gonad. all steps in maturation are found within the collecting tubules and the vasa efferentia. SUMMARY AND CONCLUSIONS 1. Bullfrog tadpoles and immature male frogs of body lengths of 45. 65. 95. and mature bullfrogs of 120 mm. body lengths, possessing normally differentiating gonads. were treated with anterior pituitary glands from adult frogs of the same species. 2. Scattered spermatozoa normally appear shortly after metamorphosis but In- die stage of 65 mm. body length there are relatively abundant structurally mature spermatozoa. 3. The short duration of the experiment, i.e., 72-96 hours, did not allow for any demonstrable growth response of the testis to the anterior pituitary hormone. 4. Kven before mature spermatozoa appear, or there is any evidence of a lumen within the seminiferous tubule, the testis of the immature bullfrog reacts to the anuran anterior pituitary hormone by evacuation of cellular elements of the forming seminiferous tubule, and by the formation of exaggerated lumina. 5. Cellular elements, found after pituitary stimulation within the collecting tu- bules and the vasa efferentia, consist of all stages in maturation except the sper- matogonia. Whole cysts of similar cells have been seen, indicating rather violent evacuation of the gonad. 6. The more mature the testis the more resistant are the earlier maturation stage- to anterior pituitary hormone release from the gonad, so that in the collecting tubule- of the newly metamorphosed bullfrog testis there may be found all stages 1rom sper- matocytes to spermatozoa but in the tubules of sexually mature (adult) males there are rarely any cells except spermatozoa. LITERATURE CITED BI-KNS. R. K. JR.. 1930. Effects of hypophyseal hormones upon Amblystoma larvae lollouin- transplantation or injection with special reference to gonads. I'roc. .Vor. /:.r/\ Biol. and Med., 27 : 836. BURNS, R. K. JR., AND BUVSK, A., 1934. Effect of extract of mammalian hypophysis on the re- productive system of immature male salamanders after metamorphosis. Jour. /:.r/\ Zool, 67: 115. EVAXS. L. T.. 1935. The effect of Antuitrin S on the male lizard. Attat. Rcc.. 62: 213. FORBES, T. R.. 1937. Studies on the reproductive system of the alligator. I. The effects of prolonged injections of pituitary whole gland extract in the immature alligator, .liuit. Rcc.. 70: 113. GLASS. F. M., AND RUGH, R., 1944. Seasonal study of the normal and pituitary-stimulated Erog ( Rana pipiens). Jour. Morph.. 74: 409. ROI'.FRTS Rl'OH \\ '.. AMI RAIVIKL. I'.. 1939. [mplantation of pituitary glands into larvae of various spe- cies < it Triton. I. F.ffect on *ex organs. I\ou.v' Arch., 139: 86. II. i SSAY, B. A., AMI LASCANE-GONZALEZ, \. M., 1929. Relationcs entre la hyponsis y el tcs- ticulo on el sapu. AY;1, dc la Soc. Ar. /on!., 80: 81. RUGH. R.. 1941. Experimental studies on the reproductive physiology of the male spring peeper, Hyla crucifer. Anicr. Philosophical Soc. 1'roc.. 84: 617. RUGH, R.. 1943. Experimental /:"»//>rv"/c.p. NOV., Hypocomides parva sp. nov., Hypo- comides kelliae sp. nov., and IiiMgnicoma venusta gen. nov., sp. nov., 200. KRIT.ELIS, EDITH JUDITH. Distribution and |)roperties of intraccllular alkaline phos- phatases, 220. J^EPTOSVXAP TA, swimming of, 93. Lithium chloride, effects on the fertilixed eggs of Nereis limbata, 188. Locus of action of DDT in flies (Drosophila), 148. LOOSAXOFF, VICTOR L. AND CHARLES A. NOMEJKO. Feeding of oysters in relation to tidal stages and to periods of light and darkness, 244. \ [ARINE Biological Laboratory Library, serial publications added to the, 88. Meiotic divisions of Brachystethus rubroma- culatus Dallas, elimination of chromosomes in the, 19. Method of describing the growth of animals, 141. MILLER, MILTON A. Toxic effects of copper on attachment and growth of Rugula neritina, 122. Myosin filaments in muscle, 32. limbata, effects of lithium chloride on the fertilized eggs of, 188. NOMEJKO, CHARLES A. See VICTOR L. Loos- ANOFF, 244. /~\YSTERS, feeding of, in relation to tidal stages and to periods of light and darkness, 244. pHYSIOLOGV of insect diapause: The role of the brain in the production and termina- tion of pupal dormancy in the giant silk- worm, IMatysamia cecropia, 234. Protoplasmic viscosity changes in different regions of the grasshopper neuroblast during mitosis, 109. I) ANA catesbiana, effect of the adult anterior pituitary hormone on the tadpoles and the immature male frogs of, 291. Ricii\Ki>s, A. GLENN AND LAURENCE K. CUT KOMI-. Correlation between the pos- >cssion of a chitinous cuticle and sensitiv- ity to DDT, 97. Ki '.ii, ROHI.KIS. The effect of the adtdt anterior pituitary hormone on the tad- poles and the immature male frogs of the bullfrog, Raua calesbiana, 291 . QALIVARY chromosome manipulation, 71. SCHMITT, F. O. See HALL, JAKUS, AND SCHMITT, 32. SCHRADER, FRANZ. Autosomal elimination and preferential segregation in the harle- quin lobe of certain Discocephalini (Hemip- tera), 265. SCHRADER, FRANZ. The elimination of chro- mosomes in the meiotic divisions of Brachystethus rubromaculatus Dallas, 19. Serial publications added to the Marine Bi- ological Laboratory Library and the Woods Hole Oceanographic Institute Library since February, 1943, 88. SHORT, ROBERT B. Observations on the giant amoeba, Amoeba carolinensis (Wilson, 1900), 8. SMITH F. G. WALTON. Effect of water cur- rents upon the attachment and growth of barnacles, 51. Striations and myosin in muscle, 32. Studies on ciliates of the family Ancistrocomi- dae Chatton and Lwoff (order Holotri- cha, suborder Thigmotricha). I. Hypoco- mina tagularum sp. nov. and Crebricoma gen. nov., 1. Studies on ciliates of the family Ancistrocomi- dae Chatton and Lwoff (order Holotri- cha, suborder Thigmotricha). II. Hypoco- mides niytili Chatton and Lwoff, Hypoco- mides botulae sp. nov., Hypocomides parva sp. nov., Hypocomides kelliae sp. nov., and Insignicoma venusta gen. nov., sp. nov., 200. 'PYLER, ALBERT. Natural heteroagglu tin- ins in the body-fluids and seminal fluids of various invertebrates, 213. \7TSCOSITY changes in grasshopper mitosis, 109. \K7ALFORD, LIONEL A. A new graphic method of describing growth of animals, 141. Water currents and growth of barnacles, 51. WILLIAMS, CARROLL M. Physiology of insect diapause: The role of the brain in the pro- duction and termination of pupal dor- mancy in the giant silkworm, Platysamia cecropia, 234. Woods Hole Oceanographic Institute Library, serial publications added since February, 1943, cS8. Volume 90 ?t Number 1 THE BIOLOGICAL BULLETIN PUBLISHED BY THE MARINE BIOLOGICAL LABORATORY Editorial Board E. G. CONKLIN, Princeton University E. N. HARVEY, Princeton University SELIG HECHT, Columbia University LEIGH HOADLEY, Harvard University L. IRVING, Swarthmore College M. H. JACOBS, University of Pennsylvania H. S. JENNINGS, Johns Hopkins University FRANK R. LILLIE, University of Chicago CARL R. MOORE, University of Chicago GEORGE T. MOORE, Missouri Botanical Garden G. H. PARKER, Harvard University A. C. REDFEELD, Harvard University F. SCHRADER, Columbia University DOUGLAS WHITAKER, Stanford University H. B. STEINBACH, Washington University Managing Editor Marine Biological Laboratory L I B R -A. H Y MAR 5 - 1946 WOODS HOLE, MASS. FEBRUARY, 1946 Printed and Issued by LANCASTER PRESS, Inc. PRINCE & LEMON STS. LANCASTER, PA. SERIAL LIST A. SERIAL list of the holdings of The Marine Biological Labora- tory has been published as a separately bound supplement to The Biological Bulletin. This supplement lists with cross references the titles of journals in the Library; additional titles and changes are published annually. A few extra copies of the original list are still available. Orders may be directed to The Marine Biological Laboratory. MICROFILM SERVICE 1 HE Library of The Marine Biological Laboratory is now pre- pared to supply microfilms of material from periodicals included in its extensive list. Through the generosity of Dr. Athertone Seidell, the essential equipment has been set up and put into operation. The Staff of The Marine Biological Laboratory Library is anxious to extend the Microfilm Service, particularly at this time when dis- tance makes the Library somewhat inaccessible to many who nor- mally use it. Investigators who wish films should send to the Li- brarian the name of the author of the paper, its title, and the name of the periodical in which it is printed, together with the volume and year of publication. The rates are as follows: $.30 for papers up to 25 pages, and $.10 for each additional 10 pages or fraction thereof. It is hoped that many investigators will avail themselves of this service. Your Biological News You would not go to the library to read the daily newspaper — probably you have it delivered at your home to be read at your leisure. Why, then, depend upon your library for your biological news ? Biological Abstracts is news nowadays. Abridgments of all the im- portant biological literature are published promptly — in many cases before the original articles are available in this country. Only by having your own copy of Biological Abstracts to read regularly can you be sure that you are missing none of the literature of particular interest to you. An abstract of one article alone, which otherwise you would not have seen, might far more than compensate you for the subscription price. Biological Abstracts is now published in seven low priced sections, as well as the complete edition, so that the biological literature may be avail- able to all individual biologists. Write for full information and ask for a copy of the section covering your field. BIOLOGICAL ABSTRACTS University of Pennsylvania Philadelphia, Pa. LANCASTER PRESS, Inc. LANCASTER, PA. THE EXPERIENCE we have gained from printing some sixty educational publica- tions has fitted us to meet the standards of customers who demand the best. We shall be happy to have workers at the MARINE BIOLOGICAL LABORATORY write for estimates on journals or monographs. Our prices are moderate. INSTRUCTIONS TO AUTHORS The Biological Bulletin accepts papers on a variety of subjects of biologi- cal interest. In general, a paper will appear within three months of the date of its acceptance. The Editorial Board requests that manuscripts conform to the requirements set below. Manuscripts. Manuscripts should be typed in double or triple spacing on one side of paper, 8% by 11 inches. Tables should be typewritten on separate sheets and placed in correct sequence in the text. Explanations of figures should be typed on a separate sheet and placed at the end of the text. Footnotes, numbered consecutively, may be placed on a separate sheet at the end of the paper. A condensed title or running page head of not more than thirty-five letters should be included. Figures. The dimensions of the printed page, 5 by 7% inches, should be kept in mind in preparing figures for publication. Illustrations should be large enough so that all details will be clear after appropriate reduction. Explana- tory matter should be included in legends as far as possible, not lettered on the illustrations. Figures should be prepared for reproduction as line cuts or half- tones; other methods will be used only at the author's expense. Figures to be reproduced as line cuts should be drawn in black ink on white paper or blue- lined co-ordinate paper; those to be reproduced as halftones should be mounted on Bristol board and any designating letters or numbers should be made di- rectly on the figures. The author's name should appear on the reverse side of all figures. The desired reduction should be specified on each figure. Literature cited. The list of literature cited should conform to the style set in this issue of The Biological Bulletin. Papers referred to in the manuscript should be listed on separate pages headed "Literature Cited." Mailing. Manuscripts should be packed flat. Large illustrations may be rolled in a mailing tube, but all illustrations larger than 9 by 12 inches must be accompanied by photographic reproductions or tracings that may be folded to page size. Reprints. Authors will be furnished, free of charge, one hundred reprints without covers. Additional copies may be obtained at cost; approximate figures will be furnished upon request. THE BIOLOGICAL BULLETIN THE BIOLOGICAL BULLETIN is issued six times a year at the Lancaster Press, Inc., Prince and Lemon Streets, Lancaster, Pennsylvania. Subscriptions and similar matter should be addressed to The Biologi- cal Bulletin, Marine Biological Laboratory, Woods Hole, Massachusetts. Agent for Great Britain : Wheldon and Wesley, Limited, 2, 3 and 4 Arthur Street, New Oxford Street, London, W. C. 2. Single numbers, $1.75. Subscription per volume (three issues), $4.50. Communications relative to manuscripts should be sent to the Manag- ing Editor, Marine Biological Laboratory, Woods Hole, Massachusetts, between July 1 and October 1 , and to the Department of Zoology, Wash- ington University, St. Louis, Missouri, during the remainder of the year. Entered as second-class matter May 17, 1930, at the post office at Lancaster, Pa.. under the Act of August 24, 1912. BIOLOGY MATERIALS The Supply Department of the Marine Biological Labora- tory has a complete stock of excellent plain preserved and injected materials, and would be pleased to quote prices on school needs. PRESERVED SPECIMENS for Zoology, Botany, Embryology, and Comparative Anatomy LIVING SPECIMENS for Zoology and Botany including Protozoan and Drosophila Cultures, and Animals for Experimental and Laboratory Use. MICROSCOPE SLIDES for x Zoology, Botany, Embryology, Histology, Bacteriology, and Parasitology. CATALOGUES SENT ON REQUEST Supply Department MARINE BIOLOGICAL LABORATORY Woods Hole, Massachusetts CONTENTS Page KOZLOFF, EUGENE N. Studies on ciliates of the family Ancistrocomidae Chatton and Lwoff (order Holotricha, suborder Thigmotricha). I. Hypocomina tegularum sp. nov. and Crebricoma gen. nov. ... 1 SHORT, ROBERT B. Observations on the giant amoeba, Amoeba carolinensis (Wilson, 1900) 8 SCHRADER, FRANZ The elimination of chromosomes in the meiotic divisions of Brachystethus rubromaculatus Dallas 19 HALL, C. E., M. A. JAKUS, AND F. O. SCHMITT An investigation of cross striations and myosin filaments in muscle 32 SMITH, F, G. WALTON Effect of water currents upon the attachment and growth of barnacles 51 D'ANGELO, ETHEL GLANCY Micrurgical studies on Chironomus salivary gland chromo- somes 71 SERIAL PUBLICATIONS ADDED TO THE MARINE BIOLOGICAL LABORATORY AND THE WOODS HOLE OCEANOGRAPHIC INSTITUTE LIBRARY SINCE FEBRUARY, 1943 88 Volume 90 Number 2 THE BIOLOGICAL BULLETIN PUBLISHED BY THE MARINE BIOLOGICAL LABORATORY Editorial Board E. G. CONKLIN, Princeton University E. N. HARVEY, Princeton University SELIG HECHT, Columbia University LEIGH HOADLEY, Harvard University L. IRVING, Swarthmore College M. H. JACOBS, University of Pennsylvania H. S. JENNINGS, Johns Hopkins University FRANK R. LILLIE, University of Chicago CARL R. MOORE, University of Chicago GEORGE T. MOORE, Missouri Botanical Garden G. H. PARKER, Harvard University A. C. REDFIELD, Harvard University F. SCHRADER, Columbia University DOUGLAS WmTAKER, Stanford University H. B. STEINBACH, Washington University Managing Editor APRIL, 1946 Marine Biological 3L.I B II A. K Y APR 291946 WOODS HOLE, MASS. Printed and Issued by LANCASTER PRESS, Inc. PRINCE &, LEMON STS. LANCASTER, PA. SERIAL LIST A SERIAL list of the holdings of The Marine Biological Labora- tory has been published as a separately bound supplement to The Biological Bulletin. This supplement lists with cross references the titles of journals in the Library; additional titles and changes are published annually. A few extra copies of the original list are still available. Orders may be directed to The Marine Biological Laboratory. MICROFILM SERVICE 1HE Library of The Marine Biological Laboratory is now pre- pared to supply microfilms of material from periodicals included in its extensive list. Through the generosity of Dr. Athertone Seidell, the essential equipment has been set up and put into operation. The Staff of The Marine Biological Laboratory Library is anxious to extend the Microfilm Service, particularly at this time when dis- tance makes the Library somewhat inaccessible to many who nor- mally use it. Investigators who wish films should send to the Li- brarian the name of the author of the paper, its title, and the name of the periodical in which it is printed, together with the volume and year of publication. The rates are as follows: $.30 for papers up to 25 pages, and $.10 for each additional 10 pages or fraction thereof. It is hoped that many investigators will avail themselves of this service. Your Biological News You would not go to the library to read the daily newspaper — probably you have it delivered at your home to be read at your leisure. Why, then, depend upon your library for your biological news ? Biological Abstracts is news nowadays. Abridgments of all the im- portant biological literature are published promptly — in many cases before the original articles are available in this country. Only by having your own copy of Biological Abstracts to read regularly can you be sure that you are missing none of the literature of particular interest to you. An abstract of one article alone, which otherwise you would not have seen, might far more than compensate you for the subscription price. Biological Abstracts is now published in seven low priced sections, as well as the complete edition, so that the biological literature may be avail- able to all individual biologists. Write for full information and ask for a copy of the section covering your field. BIOLOGICAL ABSTRACTS University of Pennsylvania Philadelphia, Pa. LANCASTER PRESS, Inc. LANCASTER, PA. THE EXPERIENCE we have gained from printing some sixty educational publica- tions has fitted us to meet the standards of customers who demand the best. We shall be happy to have workers at the MARINE BIOLOGICAL LABORATORY write for estimates on journals or monographs. Our prices are moderate. INSTRUCTIONS TO AUTHORS The Biological Bulletin accepts papers on a variety of subjects of biologi- cal interest. In general, a paper will appear within three months of the date of its acceptance. The Editorial Board requests that manuscripts conform to the requirements set below. Manuscripts. Manuscripts should be typed in double or triple spacing on one side of paper, S1/^ by 11 inches. Tables should be typewritten on separate sheets and placed in correct sequence in the text. Explanations of figures should be typed on a separate sheet and placed at the end of the text. Footnotes, numbered consecutively, may be placed on a separate sheet at the end of the paper. A condensed title or running page head of not more than thirty-five letters should be included. Figures. The dimensions of the printed page, 5 by 7% inches, should be kept in mind in preparing figures for publication. Illustrations should be large enough so that all details will be clear after appropriate reduction. Explana- tory matter should be included in legends as far as possible, not lettered on the illustrations. Figures should be prepared for reproduction as line cuts or half- tones; other methods will be used only at the author's expense. Figures to be reproduced as line cuts should be drawn in black ink on white paper or blue- lined co-ordinate paper; those to be reproduced as halftones should be mounted on Bristol board and any designating letters or numbers should be made di- rectly on the figures. The author's name should appear on the reverse side of all figures. The desired reduction should be specified on each figure. Literature cited. The list of literature cited should conform to the style set in this issue of The Biological Bulletin. Papers referred to in the manuscript should be listed on separate pages headed "Literature Cited." Mailing. Manuscripts should be packed flat. Large illustrations may be rolled in a mailing tube, but all illustrations larger than 9 by 12 inches must be accompanied by photographic reproductions or tracings that may be folded to page size. Reprints. Authors will be furnished, free of charge, one hundred reprints without covers. Additional copies may be obtained at cost; approximate figures will be furnished upon request. THE BIOLOGICAL BULLETIN THE BIOLOGICAL BULLETIN is issued six times a year at the Lancaster Press, Inc., Prince and Lemon Streets, Lancaster, Pennsylvania. Subscriptions and similar matter should be addressed to The Biologi- cal Bulletin, Marine Biological Laboratory, Woods Hole, Massachusetts. Agent for Great Britain: Wheldon and Wesley, Limited, 2, 3 and 4 Arthur Street, New Oxford Street, London, W. C. 2. Single numbers, $1.75. Subscription per volume (three issues), $4.50. Communications relative to manuscripts should be sent to the Manag- ing Editor, Marine Biological Laboratory, Woods Hole, Massachusetts, between July 1 and October 1 , and to the Department of Zoology, Wash- ington University, St. Louis, Missouri, during the remainder of the year. Entered as second-class matter May 17, 1930, at the post office at Lancaster, Pa., under the Act of August 24, 1912. BIOLOGY MATERIALS The Supply Department of the Marine Biological Labora- tory has a complete stock of excellent plain preserved and injected materials, and would be pleased to quote prices on school needs. PRESERVED SPECIMENS for Zoology, Botany, Embryology, and Comparative Anatomy LIVING SPECIMENS for Zoology and Botany including Protozoan and Drosophila Cultures, and Animals for Experimental and Laboratory Use. MICROSCOPE SLIDES for Zoology, Botany, Embryology, Histology, Bacteriology, and Parasitology. CATALOGUES SENT ON REQUEST Supply Department MARINE BIOLOGICAL LABORATORY Woods Hole, Massachusetts CONTENTS Page COSTELLO, DONALD PAUL The swimming of Leptosynapta 93 RICHARDS, A. GLENN AND LAURENCE K. CUTKOMP Correlation between the possession of a chitinous cuticle and sensitivity to DDT 97 CARLSON, J. GORDON Protoplasmic viscosity changes in different regions of the grasshopper neuroblast during mitosis 109 MILLER, MILTON A. Toxic effects of copper on attachment and growth of Bugula neritina 122 WALFORD, LIONEL A. A new graphic method of describing the growth of animals. . 141 BODENSTEIN, DIETRICH Investigation on the locus of action of DDT in flies (Droso- phila) 148 HOPKINS, DWIGHT L. The contractile vacuole and the adjustment to changing con- centration in fresh water amoebae . 158 Volume 90 Number 3 THE BIOLOGICAL BULLETIN PUBLISHED BY THE MARINE BIOLOGICAL LABORATORY Editorial Board E. G. CONKLIN, Princeton University E. N. HARVEY, Princeton University SELIG HECHT, Columbia University LEIGH HOADLEY, Harvard University L. IRVING, Swarthmore College M. H. JACOBS, University of Pennsylvania H. S. JENNINGS, Johns Hopkins University FRANK R. LILLIE, University of Chicago CARL R. MOORE, University of Chicago GEORGE T. MOORE, Missouri Botanical Garden G. H. PARKER, Harvard University A. C. REDFIELD, Harvard University F. SCHRADER, Columbia University DOUGLAS WmTAKER, Stanford University H. B. STEIN8ACH, Washington University Managing Editor JUNE, 1946 Marine Biological l;»o)i.to«y IL, I B R, .A. H Y JUL 1 WOODS HOLE, MASS. Printed and Issued by LANCASTER PRESS, Inc. PRINCE 8C LEMON STS. LANCASTER, PA. NEW! JUST OUT! THE LAST WORD ON Human Embryology By BRADLEY M. PATTEN Professor of Anatomy University of Michigan Medical School 1366 Drawings and Photographs, arranged in groups of 446 illustrations, 53 in Colors. 776 Pages. $7.00 Here is a book that stands out — un- paralleled— as a teaching text. Pat- ten's HUMAN EMBRYOLOGY offers a distinctly new approach to the sub- ject, gives a stimulating account of each aspect of the subject and shows the student how to use his knowledge of embryology in practical medicine. The style is clear, straightforward and impelling, such as characterized the author's popular earlier books on the Chick and the Pig. The book tells the complete story of human embryology in illustrations as well as in text — utilizing 1366 special drawings, which have been praised for their accuracy in detail, freshness of approach and beauty of execution. Fifty-three of them are in color. These unusual illustrations make the book an instrument of teaching un- equalled in its field. Written to facilitate the study of hu- man embryology, this new text will get better results for the teacher, and be prized by the student for its com- pleteness and clarity. THE BLAKISTON COMPANY PHILADELPHIA 5, PENNA. Use this convenient order-form THE BLAKISTON COMPANY. Philadelphia 5, Penna. Bioi. Bun. 6-46 You may send me a copy of the new Patten's HUMAN EMBRYOLOGY. $7.00 Name Address If a teacher, please state school here Your Biological News You would not go to the library to read the daily newspaper — probably you have it delivered at your home to be read at your leisure. Why, then, depend upon your library for your biological news ? Biological Abstracts is news nowadays. Abridgments of all the im- portant biological literature are published promptly — in many cases before the original articles are available in this country. Only by having your own copy of Biological Abstracts to read regularly can you be sure that you are missing none of the literature of particular interest to you. An abstract of one article alone, which otherwise you would not have seen, might far more than compensate you for the subscription price. Biological Abstracts is now published in seven low priced sections, -as well as the complete edition, so that the biological literature raay be avail- able to all individual biologists. Write for full information and ask for a copy of the section covering your field. BIOLOGICAL ABSTRACTS University of Pennsylvania Philadelphia, Pa. MICROFILM SERVICE The Library of The Marine Biological Laboratory can supply microfilms of ma- terial from periodicals in- cluded in its list. Requests should include the title of the paper, the author, peri- odical, volume and date of publication. Rates are as follows: $.30 for papers up to 25 pages, and $.10 for each additional 10 pages or fraction thereof. LANCASTER PRESS, Inc. LANCASTER, PA. THE EXPERIENCE we have gained from printing some sixty educational publica- tions has fitted us to meet the standards of customers who demand the best. We shall be happy to have workers at the MARINE BIOLOGICAL LABORATORY write for estimates on journals or monographs. Our prices are moderate. INSTRUCTIONS TO AUTHORS The Biological Bulletin accepts papers on a variety of subjects of biologi- cal interest. In general, a paper will appear within three months of the date of its acceptance. The Editorial Board requests that manuscripts conform to the requirements set below. Manuscripts. Manuscripts should he typed in double or triple spacing on one side of paper, 8!£ by 11 inches. Tables should be typewritten on separate sheets and placed in correct sequence in the text. Explanations of figures should be typed on a separate sheet and placed at the end of the text. Footnotes, numbered consecutively, may be placed on a separate sheet at the end of the paper. A condensed title or running page head of not more than thirty-five letters should be included. Figures. The dimensions of the printed page, 5 by 7% inches, should be kept in mind in preparing figures for publication. Illustrations should be large enough so that all details will be clear after appropriate reduction. Explana- tory matter should be included in legends as far as possible, not lettered on the illustrations. Figures should be prepared for reproduction as line cuts or half- tones; other methods will be used only at the author's expense. Figures to be reproduced as line cuts should be drawn in black ink on white paper or blue- lined co-ordinate paper; those to be reproduced as halftones should be mounted on Bristol board and any designating letters 'or numbers should be made di- rectly on the figures. The author's name should appear on the reverse side of all figures. The desired reduction should be specified on each figure. Literature died. The list of literature cited should conform to the style set in this issue of The Biological Bulletin. Papers referred to in the manuscript should be listed on separate pages headed "Literature Cited." Mailing. Manuscripts should be packed flat. Large illustrations may be rolled in a mailing tube, but all illustrations larger than 9 by 12 inches must be accompanied by photographic reproductions or tracings that may be folded to page size. Reprints. Authors will be furnished, free of charge, one hundred reprints without covers. Additional copies may be obtained at cost; approximate figures will be furnished upon request. THE BIOLOGICAL BULLETIN THE BIOLOGICAL BULLETIN is issued six times a year at the Lancaster Press, Inc., Prince and Lemon Streets, Lancaster, Pennsylvania. Subscriptions and similar matter should be addressed to The Biologi- cal Bulletin, Marine Biological Laboratory, Woods Hole, Massachusetts. Agent for Great Britain : Wheldon and Wesley, Limited, 2, 3 and 4 Arthur Street, New Oxford Street, London, W. C. 2. Single numbers, $1.75. Subscription per volume (three issues), $4.50. Communications relative to. manuscripts should be sent to the Manag- ing Editor, Marine Biological Laboratory, Woods Hole, Massachusetts, between July 1 and September 1, and to the Department of Zoology, Washington University, St. Louis, Missouri, during the remainder of the year. Entered as second-class matter May 17, 1930, at the post office at Lancaster, Pa., under the Act of August 24, 1912. BOOKS-WAR VICTIMS 'URING THE WAR, the libraries of lialf the world were destroyed in the fires of battle and in the fires of hate and fanaticism. Where they were spared physical damage, they were impoverished by isolation. There is an urgent need — now — for the printed materials which are basic to the reconstruction of dev- astated areas and which can help to remove the intellectual blackout of Europe and the Orient. There is need for a pooling of resources, for coordinated action in order that the devastated libraries of the world may be restocked as far as possible with needed American publications. The American Book Center for War Devastated Libraries, Inc., has come into being to meet this need. It is a program that is born of the combined interests of library and educational organizations, of gov- ernment agencies, and of many other official and non-official bodies in the United States. The American Book Center is collecting and is shipping abroad scholarly books and periodicals which will be useful in research and necessary in the physical, economic, social and industrial rehabilitation and reconstruction of Europe and the Far East. The Center cannot purchase books and periodicals ; it must depend upon gifts from individuals, institutions, and organizations. Each state will be organized to participate in the program through the leadership of a state chairman. Other chairmen will organize interest in the principal subject fields. Cooperation with these leaders or direct individual contributions are welcomed. WHAT IS NEEDED: Shipping facilities are precious and demand that all materials be carefully selected. Emphasis is placed upon publications issued during the past decade, upon scholarly books which are important contributions to their fields, upon periodicals (even incomplete volumes) of significance, upon fiction and non-fiction of distinction. All subjects — history, the social sciences, music, fine arts, literature, and especially the sciences and technologies — are wanted. WHAT IS NOT NEEDED: Textbooks, out-dated monographs, recrea- tional reading, books for children and young people, light fiction, materials of purely local interest, popular magazines such as Time, Life, National Geographic, etc., popular non-fiction of little enduring significance such as Gunther's Inside Europe, Haliburton's Royal Road to Romance, etc. Only carefully selected federal and local documents are needed, and donors are requested to write directly to the Center with regard to specific documents. HOW TO SHIP: All shipments should be sent PREPAID via the cheapest means of transportation to THE AMERICAN BOOK CENTER, C/O THE LIBRARY OF CONGRESS, WASHINGTON 25, D. C. Al- though the Center hopes that donors will assume the costs of transportation of their materials to Washington, when this is not possible reimbursement will be made upon notification by card or letter of the amount due. THE CENTER CANNOT ACCEPT MATERIAL WHICH IS SENT COLLECT. Reim- bursement cannot be made for packing or other charges beyond actual transpor- tation. When possible, periodicals should be tied together by volume. It will be helpful if missing issues are noted on incomplete volumes. BIOL o o Y MATE R IA L s T !, , C ,,'- ' .,,.-.,, ..— •, , I tory has a complete stock of excellent plain preferred and 1 materials, and would be plea* price* on PRESERVED SPECIMEN /or Z/x>l^,; Bo tan;. L and Co mp-arat :••-': LIVING SPECIMENS /or Zoology and Botany including P r '. : '. z o a n and Drosophila Culture*. i.nd AnimaU : - . "-. i ;. -: : " : ' - - - L ^ . , . -^ . - . ^' , - "^ MICROSCOPE SLIDES /or Botany, 1 ~ ::;-..; Paras: -:: . £.--r ; •• .- S u p p : ;- 0 e ; . MARINE BIOLOGICAL LABORATORY CONTENTS Page HAYASHI, TERU Dilution medium and survival of the spermatozoa of Arbacia punctulata. II. Effect of the medium on respiration 177 HENLEY, CATHERINE The effects of lithium chloride on the fertilized eggs of Nereis limbata 188 KOZLOFF, EUGENE N. Studies on ciliates of the family Ancistrocomidae Chatton and Lwoff (order Holotricha, suborder Thigmotricha). II. Hypo- comides mytili Chatton and Lwoff, Hypocomides botulae sp. nov., Hypocomides parva sp. nov., Hypocomides kelliae sp. nov., and Insignicoma venusta gen. nov., sp. nov 200 TYLER, ALBERT Natural heteroagglutinins in the body-fluids and seminal fluids of various invertebrates 213 KRUGELIS, EDITH JUDITH Distribution and properties of intracellular alkaline phos- phatases 220 WILLIAMS, CARROLL M. Physiology of insect diapause: The role of the brain in the production and termination of pupal dormancy in the giant silkworm, Platysamia cecropia 234 LOOSANOFF, VICTOR L. AND CHARLES A. NOMEJKO Feeding of oysters in relation to tidal stages and to periods of light and darkness -. 244 SCHRADER, FRANZ Autosomal elimination and preferential segregation in the harlequin lobe of certain Discocephalini (Hemiptera) 265 RUGH, ROBERTS The effect of the adult anterior pituitary hormone on and the immature male frogs of the bullfrog, Rana catesbiana 291 MBL/WHOI LIBRARY UH 17 JT X